Early on in my career, I sat on a hiring panel where this was a question asked by a senior exec. As one of the more junior members of the panel, I thought to myself, "The answer is pretty obvious", but I didn't speak up.

Let me explain.

I was in a team of all male ex-consultants. Aside from me, that is (no background in consulting). Unsurprisingly, the consensus was that hiring ex-consultants from a specific group of "elite" consultancies was the right thing to do.

I'd often think: we have enough consultants, why not bring in more diverse perspectives?

But, too inexperienced to have found my voice and definitely in the minority, I held back from saying what I thought: Maybe consider a world where not all great hires need to come from an 'elite' consultancy?

One thing I learnt from this experience: you can have all the diversity training you want, unless you have a diverse panel, good luck navigating your blind spots – you likely end up with "group think" like the hiring panel I sat on all those years ago.

Meri Williams has been a real advocate for diversity in hiring. They are one of the leading voices in tech, not only because they are a rarity: a non-binary technical leader, but also because they support women and non-binary people in accessing STEM education through their charity 'One Goes Up'.

We cover:

1. AI in recruiting

2. Meri's frameworks and advice for hiring and developing teams

3. Tactical tips for companies looking to invest in DEI

4. Tactical tips for candidates from underrepresented backgrounds

AI in recruiting

I find myself conflicted when it comes to AI. On the one hand, I've never been this excited about work in my life: I wake up every morning wondering "what's possible today that hasn't been possible until recently".

But in recruiting, most AI startups remind me of the quote: "Just because you can, doesn't mean you should".

Meri articulates precisely why I feel this way:

“
It's reinforcing biases that already existed. Researchers train these models on past decisions. Since those are biased, we can only expect the same outcome.

Even if you erased the "first-line bias" by taking people's personal details from their resume, the way people from diverse backgrounds represent themselves is different because they're so used to being doubted and viewed as less competent.

Taking a process that was already pretty broken and sprinkling AI on top doesn't help. So what does Meri do instead?

“
We're combating these biases by doing more outreach, actively searching for candidates to make sure we don't miss out on great people. We're also making sure that our job specs are even-handed and use the right mix of masculine and feminine terminology.

We have a program for associate engineers, and we welcome people from different backgrounds. For example, we hired someone in customer ops who taught themselves to code.

Meri's frameworks and advice for hiring and developing teams

So what does it look like to successfully run outreach to ensure a diverse talent pool?

“
It helps if whoever does outbound looks similar to the target audience. I end up sending outreach messages myself because it's so unusual for someone perceived as a woman to be a CTO that it gets noticed.

They then break down the three questions that people ask when deciding whether to join somewhere, stay somewhere, say yes to a promotion or leave:

Tactical tips for companies looking to invest in DEI

Meri then breaks each of these down into more tactical suggestions.

• Am I expected here? For example, providing details about what you should expect before an interview. At Monzo, they'd say "the nearest tube station is noisy, so bring noise-cancelling headphones. Small gestures go a long way. Also, ensure that the employees on your website represent the diversity within your company.

They share a personal story about having a physical disability that means they are occasionally on crutches. They've never interviewed anywhere, which lets them know ahead of an interview whether they'd be able to get around on the day, for example.

• Am I respected here? It's important to give your existing team unconscious bias training, so they do not put off candidates at the interview stage or otherwise.

I cringe to think about how low the bar is. LinkedIn is full of jobs calling for “rockstars". Rockstars?! They throw up on people and break things, not exactly the traits you'd want in a colleague (Meri's gets all the credit for that sentence; I laughed a lot).

• Can I be myself here? The more you can do to show, for example, that you have a diverse leadership team, the better. If your leadership team is wearing matching underwear to go with their matching haircuts and ties…

It also comes down to the smaller details, like what kind of benefits you offer. If your benefits are foosball Wednesdays and beer Fridays, is it that surprising that people with kids aren't interested?

Tactical tips for candidates from underrepresented backgrounds

“
Find those who look like you and find out where they work. Underrepresented folks are like the future; we're here, but not evenly distributed. Some places are better for us and you'll see folks from diverse backgrounds congregating.

Great advice. If companies care about DEI, they'll walk the walk: leadership will be diverse. And even better, you can reach out to those folks, as mentioned below:

“
When people have reached out and I've gotten to know them a bit, I'm happy to advocate for them. Getting to know them is important, though, since you're putting your reputation on the line for them.

I usually prefer when candidates are willing to invest a bit of time before asking for a straight referral.

Some candidates refuse to reach out for fear of rejection. Others reach out too freely and end up spamming. You want to be somewhere in the middle, being deliberate about who you target and crafting a thoughtful outreach message.

“
Show people how you think. Go to meetups and code bars. Write a Substack. Show potential employers how your mind works.

As we've heard from countless experts in this newsletter: show, don't tell.

Wrap up

Whether you're an employer trying to hire from diverse backgrounds or a candidate from an underrepresented background, it’s not as simple as you might think.

Meri makes a final point that sticks with me:

“
You can't hire someone from a diverse background and expect them to change your company for the better.

It's not diversity then inclusion, it's inclusion then diversity.

So if you're an employer reading this, thinking about how you can hire from a diverse background, get to work and remember: there are no shortcuts.

Welcome back to The Inside Scoop 🍨 — real talk with the hiring managers you actually want to work for.

This week, we sat down with George Dilthey, Head of Support at Clay — the fast-growing workspace tool that’s quietly become a favourite among operators and technical teams alike.

Since joining, George has scaled the support org from 3 to nearly 20 people, helped build an internal AI-powered support stack, and launched a rotational program that’s redefining how early-career folks break into tech.

We got the scoop on what makes someone stand out (hint: it's not your resume), how Clay thinks about hiring for potential, and why some of their best teammates didn’t apply the traditional way.

Read to the end for George’s tips on getting hired — and how to skip the front door entirely.

Listen on 🎧

⚡ Role Snapshot:

• Role: Product Support Specialist

• Where: New York or London (in-office preferred)

• Team Size: ~20 and growing

• Apply: get introduced to George (more on that below)

The real reason this job is different

For job seekers looking for a foot in the door at a great company, support might not be the first thing that comes to mind.

But at Clay, it should be. Because here, support is more than solving tickets — it's a crash course in the product, the customers, and how the company works.

George is building a team of what he calls "expert generalists" — people who get a deep understanding of the full business. They may start in support and then go on to shape product, growth, education, and more.

"Support gives you exposure to everything. You talk to every team, see how customers think, and build product context fast."

That exposure pays off. Clay has a strong track record of people joining support and then stepping into roles across product, engineering, education, and go-to-market.

"I really feel like support is the best way to learn as much as possible... You work with pretty much every department."

So if you're early in your career, trying to break into tech, or just love the idea of learning the ins and outs of a fast-growing product company, this might be your best bet. You won't be siloed. You won't be stuck. You will learn faster here than almost anywhere else.

What you'll actually be doing

This isn’t a job where you can hide behind macros. In fact, George explicitly argues against them.

"I exported all our macros and asked, 'Why are we saying these things a hundred times a week?' If a question keeps coming up, it means we need to fix something upstream."

That’s the mindset at Clay: don’t patch, improve. You’ll spend your days solving real customer issues, but just as often, you’ll be redesigning processes, spotting patterns, and working with engineers to fix the root cause.

Expect to:

• Troubleshoot complex issues and edge cases that make you think

• Collaborate directly with product and engineering

• Identify ticket trends and turn them into feature or documentation improvements

• Rebuild parts of the support system when they’re no longer working

And if something goes sideways? No panic.

George shared a recent case where a new routing workflow he launched accidentally reopened over 800 old tickets. Total mess. But the team jumped in immediately, cleaned up the queue, and rolled with it.

"It was a rough Monday, but I told the team: there are better changes on the other side of this."

Things break, experiments fail, and you're trusted to learn and keep building. That’s the job.

What makes someone stand out

Clay isn’t impressed by polished resumes or buzzwords. They want to see curiosity, initiative, and actual product interest. Here’s what George looks for:

→ You’ve used Clay

People who’ve built something in Clay or helped others in the community instantly stand out. Even a simple workflow shows you’re engaged.

→ You ask smart questions

Not just "how do I reset my password" — more like "why does this workflow behave this way?"

"We’ve hired people who built something cool in Clay and shared it. They didn’t even apply through the regular process."

→ You think like a product person

Spotting recurring issues, debugging workflows, escalating with context — it all shows that you understand how tech products work and evolve.

→ You use AI to level up, not check out

They don’t mind if ChatGPT helps you move faster, as long as you can explain the thinking behind your answer.

The Wheel: Clay’s hidden gem for new grads

Breaking into tech is rough right now. Most entry-level jobs require experience you don’t have yet. So Clay created “The Wheel”.

It’s a rotational program built on a simple idea: start in support, learn everything, then branch out. You’ll spend your first three months going deep on support and the product. Then 30% of your time shifts into other areas — marketing, product, GTM, education.

"We weren’t even targeting grads at first. But we realized this was the perfect entry point, and started hiring for it intentionally."

Their first Wheel cohort featured students from all over the country, including NYU, Stanford and Cornell.

They’re not just open to early-career folks. They’re investing in them.

Ready to apply?

George is looking for people who are proactive, thoughtful, and not afraid to dive into the messy stuff. You don’t need to have it all figured out — but if you’re curious, technically minded, and ready to get your hands dirty, they want to hear from you.

"We're not looking for perfect backgrounds — we're looking for people who want to learn fast and help us build better systems."

🎯 Bonus points if you show up in their support queue before you apply.

What they offer:

• Competitive salary, equity, and a real chance to grow with the company (6x YoY growth)

• A fast-moving, low-ego team

• A supportive culture (customers apply to work at Clay!)

• A beautiful office in NYC (with a growing London hub)

• Resources to level up: tools, training, and opportunities to rotate into other teams

If you want to work somewhere that values curiosity over credentials, welcomes early-career folks, and lets you grow into the kind of operator startups dream of hiring, this is it.

Think you’d be a good fit? Send us a note and if you’re a fit we’ll introduce you to George.

🤝 Send us a note

Welcome back to The Inside Scoop 🍨 — real talk with the hiring managers you actually want to work for.

This week’s episode features Gualter Augusto, Head of Support Engineering at n8n — the open-source automation tool that recently raised its €90M Series B.

He joined in late 2024, scaled the team 3x, and is still managing to keep things thoughtful, human, and high-performing.

We got the inside scoop on exactly how to stand out when applying to n8n, why showing up in the community matters more than polishing your resume, and his specific advice for landing a role on his team.

Read to the end for Gualter's top tips for getting hired at n8n.

Listen on 🎧

The secret to getting noticed by n8n

Here's what we learned: the best candidates usually show up before they apply.

They're already in the community forums helping people, sharing workflows, and asking thoughtful questions. Some even make it to in-person meetups.

"The best candidates are already helping in our community forums. That speaks louder than a resume ever could."

This completely changes how Gualter evaluates people. Instead of polished resumes and rehearsed interview answers, he gets to see how people actually think and work.

Are they helpful? Do they ask good questions? Can they explain complex ideas clearly?

The person helping a stranger troubleshoot their workflow at 11 PM on a Tuesday? That's probably a great hire.

What n8n's culture actually looks like (and why it matters for your application)

The team is remote-first, async by default, and allergic to meetings without agendas. There's real ownership here — even the most junior hires end up leading entire projects. No micromanagement, no babysitting 😅

What's interesting is how productive they are remotely. Gualter mentioned that when they meet in person, productivity actually declines because everyone gets along so well and just wants to hang out.

If you love remote-first environments and want to work somewhere that actually trusts you to own your work, this is worth paying attention to.

Scaling support 3x (without burnout)

When Gualter joined n8n in December 2024, the Support team was just four people. Now it's twelve, spread across Europe and the U.S., and somehow, they're not drowning in chaos.

What struck us during our conversation was how intentional Gualter is about this. As Head of Support, he's not just chasing faster response times or higher ticket volumes.

He's thinking about the long game — how do you scale without breaking people?

His approach is pretty simple. He checks in with his team constantly, ensuring they have room to breathe and actually own their work, rather than just grinding through tickets.

"I'm always asking if people need time away from the queue. Even if it's inconvenient, I care more about their state of mind."

This tells you something important about what kind of manager he is and what kind of workplace n8n is building.

Gualter's top tips for getting hired at n8n

Here's the actionable tips.

This is what Gualter actually looks for when hiring and how you can position yourself as the ideal candidate:

What will make you stand out

Show ownership mindset: In your application and interviews, share examples of times you identified problems and took initiative to solve them.

Prove you're a team player: Share specific examples of how you've helped colleagues or made remote collaboration easier. Show you understand what it takes to empower others and work effectively in async environments.

Demonstrate curiosity over credentials: Don't worry if you don't have the "perfect" background. Focus on showing your growth mindset and resilience.

"I've hired people with no tech background who are now software engineers. It's about curiosity and resilience, not titles."

Express genuine product interest: Before applying, spend time with n8n's platform. Come prepared with ideas for workflows you'd want to automate or problems you think it could solve. If you can't think of a single use case, this probably isn't the right fit.

Resume tips that actually matter

Tailor it — no generic PDFs. Show you understand what n8n does and why you want to work there specifically.

Show real impact with metrics — "Cut response time from 6 hours to 2 hours" beats "managed customer inquiries" every time.

Explain what you did with tools — don't just list them. What problems did you solve? What was the outcome?

Share your community handle — if you're active in relevant communities, make it easy for them to find your contributions.

On using AI for applications

Fine to use, but don't let it do all the work. If you can't explain your AI-generated answers during the interview, it shows. Use AI as a tool, not a crutch.

“AI should make you more efficient, not do the job for you. If you can’t explain the answer, that’s a red flag.”

How to actually show you're interested

Join the n8n community — start helping people, sharing workflows, asking thoughtful questions. This is your best shot at getting noticed.

Attend meetups — if there's an n8n meetup in your area, show up. Meeting the team in person shows real commitment.

Share something meaningful — create a workflow, write about automation, contribute to discussions. Show, don't just tell.

Why this approach works

What we found interesting is how n8n is thinking about hiring. They don't expect you to be a power user walking in the door, but Gualter is definitely looking for people who actually care about what n8n does.

Someone who's already jumping into forums to help other users isn't just looking for any job — they're genuinely interested in the problem n8n solves. That kind of intrinsic motivation is hard to fake in an interview, but it shows up naturally in how people engage with the community.

Ready to apply?

n8n is hiring Support Engineers and Community Engineers in Europe and the U.S. If you're into automation, love helping people, and want to work on a product you'd actually use — this might be for you.

What they offer:

• Competitive pay, equity, and vacation

• A €1K yearly learning budget

• A team with a 94 eNPS (🔥)

• Real ownership and growth opportunities

Think you’d be a good fit? Send us a note and if you’re a fit we’ll introduce you to Gualter.

🤝 Send us a note

Listen on 🎧

Welcome to The Inside Scoop 🍨, where we spotlight the human stories behind hiring.

In our debut episode, we sat down with Simon Rohrbach, CEO of Plain (a $15M Series-A startup rebuilding B2B support tools), who is hiring a founding Support Engineer (and other roles across GTM, engineering and design) to join his team in London or San Francisco.

Catch the highlights below ↓

From Designer to Support Evangelist

Simon cofounded Plain after watching support teams grow to nearly half the org at his last company, yet still operate with the worst tools and the quietest voice in the room.

“Support teams are your link to customers. Why are they the ones with the clunkiest tools?”

He and his co-founder, Matt Vagni, didn’t just want to build better software; they wanted to build it with empathy baked in.

If you're interested in a job at Plain, skip to the "Interested in joining Plain" section below ↓

The Fake Customers Were Real

Instead of shipping features and hoping for the best, Simon hired a support consultant and an actor to simulate real customer calls.

Then he put the whole team — engineers, ops, designers — through live ticket drills.

“You’re on the phone with an angry customer. The queue’s growing. You don’t know the answer. And by the end of the day? The backlog hasn’t moved.”

That feeling of doing everything right and still feeling behind is now built into every product decision.

Support Is the Culture

At Plain, support is at the centre of the company's operating system. Everyone does it, everyone learns from it, and it informs every product decision.

Now they’re hiring a founding Support Engineer to keep that loop strong — working closely with high-growth startups like Cursor, Granola, and Stitch.

Why it matters: Every feature gets filtered through the reality of being overwhelmed, understaffed, and expected to do the impossible.

What Simon Looks for in a Candidate

Simon didn’t start with skills or titles. He looks for:

• Intellectual humility – curiosity > ego

• Ambition – not for status, but for building something meaningful

• Emotional intelligence – the ability to read a room and give feedback

“If you’ve used AI in your take-home assignment, great — just tell me how. Show me clear thinking and purpose.”

What It’s Like to Work at Plain

Simon was honest about the challenges:

• Priorities shift daily based on customer needs

• The team is split between London and SF (an 8-hour time zone gap)

• They’re going up against incumbents with a 10-year head start

But the tradeoff? A clear sense of purpose, a tight team, and a product that’s actually shaped by the people who use it.

What This Actually Means

Most support tools are built by people who’ve never done support. Most companies say they care about customers, but they build for scale, not understanding.

Simon chose a different path:

He turned his team’s biggest weakness — lack of domain expertise — into their greatest strength.

Plain just raised $15M and is hiring. But more importantly, they’re building something that might actually make support professionals' days a little better.

In a world full of productivity theatre, that feels quietly revolutionary.

Interested In Joining Plain?

Plain is looking for a Support Engineer who's tired of bad tools and who not only wants to work on support, but also help shape their product roadmap.

What they need: Someone with support experience who can communicate clearly with non-technical and technical customers. You should be curious about how things work under the hood, even if you're not a developer.

What you'll get: £50-85k (London) or $100-140k (SF) plus equity, real product influence, and the chance to establish support standards at a fast-growing startup.

Think you’d be a good fit? Send us a note and get introduced to Simon.

👋 Send us a note

The Inside Scoop 🍨 series will uncover more behind-the-scenes stories with founders and hiring managers who are approaching hiring differently

We’ve all seen the heartbreaking stories on LinkedIn: “It’s been six months, and I haven’t found anything,” “I’m stuck and don’t know what to do next,” etc.

In 2025, getting hired through a traditional application is nothing short of a miracle. But for those creative enough to try, there is another way…

Mindaugas Petrutis’ story is fascinating. His hustle is something that is both rare and contagious, a few highlights below:

• He got into recruitment with no experience: Sales experience was a must, so he took what he describes as “the lowest level sales job he could find”.

• He broke into recruitment and found an edge: In-person communities. He’d organise drinks, invite people to events and build deep relationships.

• His love for communities led him to InVision. A single email to the CEO, whom he knew from his community work, resulted in interviews and a job offer.

• He built a product with Lovable that ultimately led to a job: The Backchannel was a side-project that got him noticed and eventually hired by Lovable.

He is living proof of the famous quote:

“You can just do things”.

In this week’s edition, we discuss:

• How he was hired at Invision and Lovable

• The opportunity he sees with “building in public”

• Building Backchannel to help people get hired

How he was hired at Invision and Lovable

His approach is pretty simple: he plays the long game.

I want you to picture 2 scenarios:

1. Someone randomly messages you on LinkedIn. This is the first time you’re seeing their name. They’re asking for help getting hired at your current company.

2. Someone whose events you’ve attended, who you’ve spoken to several times and who you trust sends a message asking for the same thing.

This example might sound overly simplistic, but #1 is literally what I get contacted on LinkedIn for at least five times per week. #2 is rare, but when it does happen, you certainly want to help. Occasionally, it might even lead to a great hire.

It’s precisely the approach Mindaugas took to getting hired at InVision:

“

I’d met the CEO and one of the VPs several times, I’d co-hosted an event with Invision for my own startup. When I decided I wanted to get hired, I sent a short email saying “I’ve met a bunch of your team and everyone I speak to is amazing, helpful and talented. I’m looking for my next opportunity and InVision is a company I deeply admire”.

As luck would have it, InVision were building a team in Europe at that very moment. The CEO introduced him to the hiring manager. Long story short: he got a job offer.

You might be thinking, “Meh, got lucky”. You’d think that until you hear how he was hired at Lovable.

“
I was obsessed with Lovable. I built my own personal projects and shared them with the team, with product feedback. One of my projects, Backchannel (a product built to help job seekers get hired) got enough traction that the team asked if they could feature it on their blog. I was basically a superuser. I was even attending their events as a guest.

The next time an opportunity on Lovable’s growth team came up, guess who they called?

“
People at Lovable were viewing my LinkedIn profile, including the founder Anton. I got a message from Elena Verna (LinkedIn and Substack guru!) who leads growth, asking to chat. After a couple of calls and working sessions, I’d landed a job.

This is how you stand out. Submitting a generic application will only take you so far; ironically, Mindaugas had submitted one before and never heard back. This level of obsession is what gets you noticed and ultimately hired.

The opportunity he sees with “building in public”

Mindaugas created a win-win situation for himself. Yes, he was sending Lovable signals, and power users are always at the front of the line for job applications, but the alternative was that he was building in public.

“
I was literally working on random ideas and projects. And then I’d share them on LinkedIn. I would quickly prototype a product, post about it and see whether it generated any interest. The number of interesting people who reached out wanting to talk (both old and new acquaintances ) asking whether I wanted to partner, start something together or join an existing venture was amazing.

It's incredible to think that his “random ideas and projects” led to a job at one of the hottest companies in the world today. If you’re reading this and wondering how to increase your chances of finding a job, remember: “you can just do stuff.”

Many people cringe at the thought of posting on LinkedIn, writing on Substack, sharing their projects, etc. Usually, those people are taking themselves too seriously. If putting your work out into the world results in someone appreciating it, then great. If not, they’ll ignore you, and nobody will remember anyway. Win-win!

Building Backchannel to help people get hired

With a background as a recruiter, Mindaugas would frequently be asked for help navigating the job hunt. This led him to build Backchannel — ironically, the same product that got him hired at Lovable 💡

His approach isn’t to create yet another course or provide expensive coaching. He’s actively using AI to scale his knowledge, make it accessible to others, and bring them together through community.

The best part was that he built this single-handedly and validated that customers wanted it, as he built it with Lovable.

“

I had the idea on a Sunday night, and now it just generates money.

This is what the future of entrepreneurship looks like. Check out the Loom that led to his first paying customer ↓

Wrap up

AI is disrupting traditional hiring processes. Standing out from the sea of applicants is harder, but as Mindaugas’ story proves, it’s not impossible.

Right now, the best way to get hired is to show, not tell. Build a product, service, newsletter, whatever it takes to show a hiring manager that you’re a standout candidate from the sea of candidates sitting in their applicant tracking system.

Customer support is being disrupted, and most companies are wildly unprepared. They either dismiss AI as hype or go too far with job cuts.

Klarna is an example of the latter. In August 2024, they famously went "all in" with AI, announcing dramatic workforce reductions, only to reverse course in 2025 when they realised their AI-first approach compromised quality.

While they eventually rolled back their expectations for how much support volume they could delegate to AI, Klarna provides an excellent case study: AI can enhance CX, but humans remain essential.

This post explores what it means to set up an AI-first CX team.

The shifting CX landscape

You’ve likely seen new jobs for “AI PMs,” “AI engineers,” and “GTM engineers”—all roles that involve functional experts (in product, engineering and growth) using AI to scale themselves.

Customer support is experiencing this same shift, moving beyond traditional efficiency metrics toward something bigger.

For years, support was defined by responsiveness: quick replies, fast resolution times, and high satisfaction scores. But this model is becoming obsolete. AI now handles the bulk of simple queries, raising a critical question:

Do we still need humans in customer support?

Short answer: Yes.

Longer answer: Yes, but these roles look very different from today's CS roles.

We believe a foundational point will inform the future: humans like feeling understood. Neuroscientific research provides direct evidence that not feeling understood activates brain regions associated with negative affect and social pain1.

This is precisely the feeling after a bad support experience, one we’ve all encountered.

Feeling “understood” requires a human. Not every support conversation will need a human, but every flow will need one to validate that customers feel understood.

Companies that are overly focused on efficiency risk repeating Klarna's mistakes, damaging trust and customer loyalty.

This is what typically happens when you remove humans but leave behind the chaos:

Bad automation → User not understood → Frustrated user → Broken trust.

A customer’s trust is the most precious thing a company can protect. It’s earned slowly but easily destroyed.

Many AI startups are following the playbook of putting AI in front of a customer before being confident it can deliver. Unsurprisingly, this leads to churn.

The future of customer experience is one where a support ticket doesn’t need to be triggered in the first place; the magic will be in AI anticipating an issue before it becomes one. And when an issue does happen, systems should prioritise customer understanding by clearly defining:

• Who decides which issues get deflected and which get escalated?

• Who designs the conversation logic between humans and AI?

• Who defines what "good" looks like?

• What roles are essential in this new CX ecosystem?

Future CX job titles

We like to think of future CX teams like successful restaurants. Profitability and efficiency matter, but the customer’s overall experience is equally vital.

This is more than a fun comparison—just like in hospitality, delivering great customer experiences depends on thoughtful systems and attention to detail.

Here are some future-proof CX roles, with restaurant analogies included:

In a great restaurant, the behind-the-scenes roles — architects, chefs, designers, maître d’s — make the magic feel effortless to the customer.

In a world where every product clumsily puts AI at every customer touchpoint, this “magic” becomes the premium experience.

Continuing with the food analogy:

AI is processed fast food, and human support is the new “organic food”.

With the right people in place—people who understand systems, signal, and build with intention—support becomes an engine for growth. It’s no longer simply a reactive function. Instead, it helps shape the product and informs direction.

This transformation is more urgent than it seems. AI is already reinventing support as we know it, and without these roles, teams risk duct-taping automation onto broken systems.

Are you hiring? If you need help finding these future-proof CX candidates, comment below, and we’ll reach out.

Are you looking for a job? Submit your resume here, and we’ll contact you about job opportunities we find.

Something clicks. It makes sense. It’s intuitive to the point that a child knows how to work it (remember the first time you saw a baby swiping up on a smartphone?). Companies prioritising this feeling make it look easy, but it’s anything but.

Apple. Airbnb. Linear. These companies think from a user’s perspective and optimise for a feeling of “it just works.” The lingering question is whether AI can create or further elevate this feeling and the resulting implications for product designers.

This week, we speak to Filippos Protogeridis, head of product design at Manual and the author of Designary, a blog and course on product design.

Sign up for Filippos’ Blog and course, Designary.

We discuss:

• The bar is higher than ever for product designers

• How hiring changes for better and worse

• The opportunity ahead with AI as an unlock

The bar is higher than ever for product designers.

If you’ve spent time tinkering with v0, Loveable, etc. (tools that turn text prompts into software), you’ll have seen that the barrier to entry for creating a digital product has been drastically lowered. It’s pretty simple to go from a sketch in your notepad to a working product (or at least a prototype) you can put in front of customers.

This is a step change from where we were not too long ago, where taking an idea to a product designer meant explaining your idea in text or drawings (in my case, childlike 😅), hoping they were understood, and then passing them onto an engineer to build a first version. It’s incredible how far (and fast) you can now get independently, but there are downsides too.

With the bar to entry being lowered, the volume of products will dramatically increase. Standing out from the crowd will require effort.

“
We’ll have more clutter with tons of websites and micro-businesses/apps popping up. But that also raises the bar. When people started creating websites in the dot-com bubble, they looked terrible. The first websites that had a good UX and solved a problem prevailed. We’re facing a similar reset now. Everything we knew about the design industry is shifting. It’s harder than ever to create something that stands out.

Unsurprisingly, design founders are in vogue. YCombinator has called for design founders, and funds are being launched with the same focus.

The design job of the future will be founder. Designers already have so many of the skills needed to be great. Strong user empathy. A focus on solving problems. A high bar for quality. And taste. A must for any founding team

Aaron Epstein, General Partner, YCombinator

The opportunity is to rethink how we interact with products entirely. Plenty of startups are reimagining what’s possible with voice, for example. These startups don’t just build incrementally on top of existing paradigms; they develop entirely new ones that can delight their users. An LLM cannot look to previous examples for this. It requires creativity, experimentation, and craft.

“
Craft will become more important. In a world where everyone can create or copy a product quickly, the products and the teams that prevail are the ones that take something average and turn it into something truly remarkable. That's something that AI won't be able to do by itself, that's where you need human traits: deep passion and empathy. That’s the opportunity.

If craft is the ultimate currency, how do you hire for it?

How hiring changes for better and worse

“
There will be fewer roles, and the remaining ones will have a higher bar. The industry is flattening to a great degree. Companies that had an army of designers are probably going to keep half or even a third of them, because too many positions involved repetitive tasks that can now be automated (e.g., putting components in place, creating layouts, or replicating a simple interaction pattern).

This is consistent with what we’ve discussed in this newsletter before. The bar is higher, and jobs of the future will fundamentally change when humans are freed from mundane yet simple tasks. But that also creates an opportunity. For those who excel at their craft and have superior taste, there’s a huge opportunity to scale themselves to a point where the model of employee/employer fundamentally changes.

“
There will likely be a new generation of design founders, freelancers, and one-person businesses. I have done this myself before (freelanced and built businesses), and it was hard. With the tools we have now, I would be in a much better position to build a fantastic career working with multiple employers. People will have the opportunity to go out on their own and deliver incredible products/services without many resources.

The rise of the solopreneur and fractional worker is still underappreciated. I wrote about my journey with fractional work and can only see this trend increasing further.

The opportunity ahead with AI as an unlock

I like to finish each interview with a question: “What should I have asked you today that I haven’t?”. Occasionally, it reveals a gem from the interviewee:

“
I’ve been thinking about AI a lot recently. The past 6 months have been a learning experience, and even though it sounds scary and overwhelming, change brings opportunity. As an industry, we’ve spent too long focused on design systems and processes for the sake of velocity, and now that we have AI, we don’t need to be so rigid. We can see tools not as an obstacle but as an unlock for creativity.

I have spoken to many people about their careers recently, as a result of this newsletter. I can honestly say I’ve never witnessed a time where there is so much anxiety around the future of work, at the same time as there being so much opportunity.

As with most things, the key is to have a growth mindset and approach new ways of working with genuine curiosity.

“
I think we'll see a lot of creativity over the next couple of years. You just need to ride that wave with curiosity and optimism, get out of your comfort zone, and try new tools and approaches.

Wrap up

The only constant in these next couple of years will be change. While it might feel like there’s a lot of uncertainty about the future of work, there is also a lot of opportunity.

AI can create at speed and on a much larger scale than humans ever could.

We can delegate the parts of our work that don’t require creativity or experimentation to AI. As Filippos mentions, our time as humans can be spent pushing the boundaries of what’s possible, using it as an unlock for creativity.

The “feeling” I mention in the beginning is something I believe remains a human endeavour until now. And that’s where we can still raise the bar.

What’s the point of using “ex-Google” (or equivalent) on a resume or LinkedIn?

Asking this is a surefire way of stirring up debate on LinkedIn. Usually, people will jump in with comments like “get over your exes” or “this says nothing about an applicant”. While you can see a case for those comments, they ignore reality.

Whether we like it or not, “ex-Google” works pretty effectively. In the sea of noise that is LinkedIn and applicant tracking systems, it can be hard to find the “top” candidates.

As applicants, both of us have benefited from being “ex-Airbnb”. It hits the lizard brain pretty effectively (“if they worked at X, they must be decent”, or so the thinking goes). As hiring managers, we’ve learnt not to let the lizard brain take over.

The “ex-Google” trap and why it’s problematic:

Bias: When hiring from a pool of candidates who have only worked for prestigious brands, you’re inheriting the bias those brands have when hiring.

Fit: Just because someone was successful at Google doesn’t mean they’ll be successful at your company, especially if your startup has limited resources.

Depth: Having experience from a top brand doesn’t exclude a candidate from being good at their job, but it’s not enough to warrant hiring them. Interviews need depth.

The rise of AI makes this problem worse, not better.

If relying on prestige was a crutch before, it has the potential to become a full-blown dependency now. Resumes are more polished and keyword-optimised than ever before. Anyone can generate a beautifully written, applicant tracking system-friendly resume, regardless of whether the underlying experience backs it up.

That’s made the hiring process more overwhelming than ever.

When every resume starts to read well, hiring teams may fall back on what feels safe: recognisable names and logos. This could cause hiring teams to miss some of the best, high-slope candidates—the ones with steep growth curves or grit.

Here’s how AI may make the problem worse:

• It amplifies existing bias: Most AI resume screeners are trained on historical hiring decisions, which often favour prestige. If your model learns that “ex-Google” resumes tend to get interviews, it will keep surfacing more of the same.

• It encourages resume homogeneity: With tools like ChatGPT, resumes are starting to look identical and appear near perfect. This creates the illusion of competence, even when experience and skills are lacking.

• It overindexes on keywords: AI often ranks candidates based on specific terms, which rewards those who optimise for buzzwords, not those who can do the job. The risk is that you pass on someone great who used different keywords.

We need better signals.

Leonardo da Vinci is often credited with the first documented example of a resume—a short letter to the Duke of Milan highlighting his expertise in engineering and architecture. But he didn’t list all his skills and experience; he tailored his pitch to match what he believed the Duke was looking for.

In that sense, not much has changed since 1482. Today’s candidates take a similar approach by adjusting bullet points, reframing responsibilities, or using strategic keywords on their resumes. But unlike da Vinci’s time, a resume today isn’t a personal letter; it’s a standardised document that the hiring manager often doesn’t see.

Most resumes are filtered through layers of software or busy recruiters, and as a result, candidates who don’t speak the correct language, haven’t worked at the “right” companies, or come from non-traditional backgrounds, don’t stand a chance.

Some teams are already skipping the resume entirely. Amanda Zhu, co-founder of Recall.ai, has discussed building teams without looking at a CV.

We need a new signal that’s not a resume proxy or a brand name shortcut—something deeper than a resume. It should help hiring teams understand what a person has done, how they operate, and their potential, without hours of prep from the candidate.

AI can help, but we shouldn’t replicate the existing hiring system. We don’t need better filters for keyword-matching resumes—we need to rethink the entire model. We need tools that surface real potential by analysing work samples, writing style, career trajectory, and peer endorsements, which can then be presented to a human decision-maker.

An AI plus human-in-the-loop system could help surface candidates based on their experience and trajectory. That’s the opportunity.

Wrapping up

I’m working with a small group of individuals who believe this is a problem worth solving, and it’ll take more than just better tools to do it. If you’re currently hiring for roles in product, let us know, and we’ll share more. We’ll share candidates who not only look good on paper but who have the credentials that your ATS is probably missing.

If that sounds interesting, drop a comment below or shoot over a DM on LinkedIn or Substack to .

I’m running a hands-on Hiring Sprint with Zeshaan (ex-Onfido Director of Talent) on June 6th. If you’ve just raised money and need to make your first hires, we’ll help you build your hiring playbook—job specs, scorecards, interview flows, and onboarding plans—in a day.

AI is redefining what we expect from humans.

“Did you create that with AI?” might have once been considered an insult. These days, if you're not thinking about how your product can scale its users, you're falling behind.

AI engineers are at the forefront of this type of applied AI. Essentially, they use code to augment (and even replace) humans.

“

There are levels to AI engineering, but often you're trying to replace a human expert's work. It’s not just an LLM call, it’s a bunch of them feeding into one another. That system's complexity means you can no longer test it as you could with non-AI products. This is what creates the role of AI engineering.

Lawrence is a founding engineer at incident.io and has spent most of the last year building agentic-AI features into their incident-response product. We’re excited to speak to him because of his experience transitioning into AI engineering.

We discuss:

• Hiring AI engineers.

• How to build AI products that consistently deliver.

• The practicalities of building agentic products in 2025.

Hiring AI engineers

Finding candidates who are a good fit is an interesting challenge when a new role, like AI engineer, emerges. It’s not simply keyword matching for prior jobs and experience. You have to look for signals in the application and during the process that you wouldn’t usually have to.

“

I’ve been quite impressed when candidates have built a UI or a new way of communicating with an LLM as a side project. Sure, there’s the risk that it’s been vibe-coded into existence, but that’s relatively easy to detect in an interview process. When they’re excited about AI to the point they’ve been tinkering with it, that’s the signal I’m usually looking for.

Incident deeply cares about product thinking. Their engineers are expected to speak to customers and determine their needs. Their AI engineers are no different:

“
It’s things like ensuring that when an AI proposes specific actions in response to an incident, there’s a human in the loop when it matters. That’s critical for a product like ours, so we check for this understanding in interviews.

If you think about an engineer dealing with an incident (website downtime, critical bug, etc.), the presence and timing of the “human in the loop” are crucial.

Equally important is having a way of quantifying how well a large language model is performing.

This is typically done through an eval dataset containing input/output pairs. The outputs are compared against an expected result to validate for correctness. For example, inputs could be questions, outputs could be the answers, and the eval dataset would evaluate whether the expected answer is correct:

An example eval

(Q: Capital of Germany → Expected answer: Berlin → Answered: Berlin —> Correct).

The key to generating these eval datasets is setting up foundational AI infrastructure, which Incident has invested heavily in. They look for engineers with the mindset to thoroughly inspect the data and determine whether or not the AI is improving.

How to build AI products that consistently deliver

There are a lot of AI products that seem impressive in a demo. But, in practice, they fail to meet expectations. Incident’s product relies on them delivering value during incidents with little room for error. Lawrence talks about what it takes to provide this experience:

“

We have a scorecard, which we apply to different parts of the product. It contains several key metrics and a load of diagnostic information. As an AI engineer, someone would take a particular set of requests on which the AI performs poorly and iterate on the solution until the evals start improving. This type of work is very different from product engineering.

This is the crucial work required for an AI product to deliver value consistently.

In Incident’s case, this has been a key strategic priority for their product. They have tight feedback loops with customers and diligently evaluate their eval datasets.

It's not easy, but their ambition isn’t either: they aim to use agentic systems to 10x the productivity of an incident responder, actively aiming to replace some of the human work required in an incident. It’s less about eliminating the role and more about reimagining where someone spends time dealing with an incident effectively.

The practicalities of building agentic products in 2025.

In addition to the “human in the loop” and “treating AI as a human” points mentioned above, Lawrence discusses the need for citations to provide trust.

“

When an AI product makes a suggestion, users must unpack why the product came to that conclusion. They need to see the raw evidence, i.e. the inputs. We’re facing this challenge right now: how to package this information in a way the user understands. So it’s more than just getting the product to make the proper recommendations; it’s also about ensuring it leaves a trail the user can follow.

I’m curious to understand the role of a PM in an organisation with product engineers and AI engineers. We’ve previously written about whether AI will kill product management. – and ultimately concluded that the future would have more, not fewer, PMs.

”

In our case, our PMs are thinking further into the future. If our product engineers are focused on the here and now, our PMs typically think about where we might get to next year. How will we package up what we’re building now to customers in a year? The building is primarily on the AI engineers, but how to position and market the product is more of a PM’s remit.

Even in a “product engineering first” company, PMs are safe for now.

Wrap up

Building agentic products is not easy. You need to hire engineers with the right skillsets, set up the relevant infrastructure, and be prepared to put in the work to achieve reliability. There is no easy fix, but the uphill struggle is worthwhile.

Agentic products allow us to rethink what humans can focus their time on. For an incident responder, this means spending less time on menial and administrative tasks and more time dealing with a critical incident.

This is a good analogy for the wider workforce and a future we should be running towards, not away from. “Disrupt or be disrupted” is no longer a cliche for tech products; it’s how we should all think about our work.

It might seem like a fad – and let’s be honest, “AI” is being plastered everywhere at the moment – but it’s a fundamentally different role to that of a traditional software engineer, and an example of a role that is only growing in popularity.

Aurimas Griciūnas writes extensively about AI engineering in his newsletter SwirlAI. His content is among the best I’ve seen, going deep in a way that remains accessible to a non-technical audience; refreshing in the age of self-proclaimed AI gurus.

We talk to Aurimas about:

• What is an AI engineer?

• How can software engineers make the transition?

• Common pitfalls

• How to hire an AI engineer

What is an AI engineer?

“

From an implementation perspective, AI engineering is very similar to software engineering. From a system’s perspective, it’s more comparable to machine learning engineering, because of all the nondeterminism in the system and how you need to evaluate, experiment and evolve it.

When looking under the hood of an AI product, you’ll uncover nondeterminism wrapped up in complex systems. Where traditional software products are deterministic (“if X then Y”), AI products are nondeterministic (“If X then Y, Z, or K).

Combine this in a system of agents where large language models pass information to each other to reason, and a set of skills is required that neither software engineers nor machine learning engineers excel in: AI engineering.

“

Most software engineers are not trained for nondeterministic systems. They don't know how to evaluate them other than getting feedback from users (e.g., thumbs up or down and overall satisfaction). Because of this, an AI engineer needs a combination of different skill sets, which is why we need a new title to describe their role.

Aurimas breaks this down as a combination of AI researcher, ML engineer and software engineer.

How can engineers transition into AI engineering?

One interesting topic we cover is who is best placed to move into the role of AI engineer from the three roles mentioned above.

“

For early-stage startups looking to hire their first AI engineer, the best background is that of a machine learning engineer. It’s easier to learn code than it is to understand statistics. But it’s also important to note that they’ll still need to upskill.

They’ll still need to accustom themselves to distributing the type of compute necessary for microservice architectures and need to become skilled at implementing monitoring and observability into these systems.

Regardless of the discipline, we’re starting to see specialisms converge. It’s ideal if an individual is exceptionally skilled in one area, but the ability to 1) upskill in adjacent areas and 2) use AI to scale themselves will be necessary for everyone.

The same logic applies to marketing, which we discussed in our previous post: “Should you replace your marketing team with AI?” - and led to the conclusion that the era of the “T-shaped marketer” is over. Change is coming here…

“

This is a new field that is continually evolving. But you’ll need to decide which areas to dive deeper into. An AI engineer doesn’t need to train machine learning models. However, they should stay up to date with the latest news and research. There are also boot camps that give engineers a solid foundation to learn the skills needed for this role.

Aurimas is running one of these himself. Check out his Maven course, End-to-End AI Engineering Bootcamp. Given how insightful his newsletter is – it’s taught me a bunch already – I’d highly recommend checking it out.

I’m keen to dive into what he’s seen when building these systems in practice.

Common pitfalls in AI engineering

“

You need to be prepared for imperfect systems. You’ll need to cluster input questions into those that are answerable and those that are not, and spend a significant amount of time improving the answers for each cluster.

It’s not enough to feed your data into an LLM and expect it to work. Even when things do work, you’ll need to be prepared for them to stop working. This is why setting up tests and observability is so essential. Dealing with non-deterministic systems is not easy.

Blanket comments about engineers being eliminated are more hype than substance. If anything, AI products require more, not less, human expertise than non-AI products.

But the above also explains why there are so many sloppy “AI products”. We’ve all used them: they look amazing during a demo, but when you use them in a real-life scenario, they fall over at the first hurdle.

Maybe we need a new mantra for AI products, as launching an “embarrassing MVP” that loses trust only leads to high churn (typical for many AI products).

“

Another common pitfall is not involving a business stakeholder early enough. Business people are usually more intelligent than engineers might think, and because of their experience, they’ll probably be better placed to write prompts.

If you’re trying to automate a part of their job away, simply asking them to review your prompts can lead to solid gains.

Hiring in AI engineering

I’m especially interested in getting Aurimas’ thoughts on how to interview AI engineers:

“

I would focus primarily on system design. I’d present candidates with a hypothetical scenario in which we collaboratively build a system and ask how they would approach it.

We can start by identifying the problem and defining the metrics that we’d optimise for, not just building an LLM system. Ideally, you should be able to articulate a story about how you would initiate and continuously develop such a system.

What I like about this is that it’s not simply an output. It’s testing the thought process in a collaborative environment.

Again, different discipline, but quite similar to the theory outlined in a previous post about GTM hiring: the value in an interview process isn’t the outputs but rather the thinking and inputs that go into creating those outputs.

“

After the “proof of concept” stage, I’d ask about how a candidate would proceed.

How would they define their evals?

When would they start implementation?

How would they evaluate the end-to-end system as well as the smaller pieces?

How would they control the costs of all this testing whilst guaranteeing that the system evolves in the right direction?

Wrap up

It’s not all doom and gloom. AI will inevitably create new jobs. Once you start to understand how AI products work, you know that building them isn’t straightforward and will require a new set of skills.

The doomsters claiming the end of software engineering are clickbaiting.

Another role emerging is that of a full-stack AI engineer. Similarly to AI engineers, it encompasses several disciplines, requiring product thinking and rapid iteration.

In our next post, we'll talk to Lawrence Jones from incident.io about what this role requires in practice and get a peek into the day-to-day responsibilities.

“
We don’t need a copywriter anymore. We have ChatGPT.

I heard this from a technical founder last year. The quote “a little knowledge is a dangerous thing” came to mind. In the founder’s defence, I don’t think he’s alone in this thinking. And I don’t think it's unique to marketing.

There’s a common experience I’ve witnessed from people using AI

1. Try a new AI tool in a domain outside of their expertise

2. Get to the “aha” moment. Find it amazing and magical

3. Conclude that professionals in that domain are doomed

4. Take to social media to share the above (how fun those posts are to read 🙄)

Not much is shared about what happens next. But I’ve seen it happen time and time again. The inflated expectations are inevitably met with a trough of disillusionment as soon as it becomes evident that AI needs guardrails.

After going through the first three steps above (countless times), I’ve concluded that the third bullet needs adjusting. Concluding that average professionals in that domain are doomed is more accurate. Those in the top tier, who have a mastery that makes them seem like wizards to everyone else, will be the ones who will become superhuman when using the right tools.

Why? Because they know where to set the guardrails. They know when to zoom in and out. I’ll make a bold prediction here: The era of the T-shaped marketer is ending. Being OK in multiple marketing domains is something we can delegate to AI. What’s harder to delegate? True mastery. ChatGPT can’t replace a world-class copywriter. It can replace an average one.

Let’s take Joanna Wiebe, for example. Joanna is a world-class copywriter and the founder of Copyhackers. Last year, I attended one of her talks, “AI copywriting frameworks for startups,” where she gave a masterclass.

Here is how she uses AI to help her write copy:

• Analysis: She uses it to analyse and categorise customer data, such as datasets from customer surveys.

• Segmentation: She then uses the data analysis on her customer data to help her decide ‘what to say’ and ‘how to say it’, creating segments to explore.

• Ideation: She uses it to help her brainstorm examples for ads and create rough first drafts, which helps to get ideas flowing.

• Ordering: AI helps to order ideas based on specific copywriting frameworks that she’s created and uses herself.

• Checks: AI helps her figure out ad outlines & checks against Meta’s ad rules for compliance.

But she always writes the final copy. AI is her assistant, not her thinker. She’s still the one accountable for the final outcome.

Note how it’s not as simple as delegating to AI and calling it a day. 10x copywriters and content marketers use AI to increase the depth of their knowledge, deepen their insights and accelerate their output. But they wag AI’s tail, not the other way around.

Here’s five practical ways that 10x content marketers use to scale themselves with AI1:

Wrap up

10x content marketers are super close to the customer and deeply understand their pains, which they turn into relevant messages. Only then can they produce content and operate at scale with AI.

But they are not your single marketing machine. A 10x content marketer can be a master at copy but won’t be able to replicate this scale in design or UX: they don’t know the guardrails.

They dominate in their niche: content. They can leverage tools and scale themselves. But they are not a ‘one-person marketing band’.

To conclude, if you think you can replace your marketing team with AI based on what you’ve seen on social media, think again.

As much as I dislike the cliche, it’s fair to say that hiring has become more challenging, not easier, in the last few years. Recruiting remains one of the meatiest yet most underestimated problems. There’s a graveyard of startups that set out to solve the “simple” problem of recruiting, only to quickly discover that it’s much harder than it seems. Paddy Lambros is acutely aware of this phenomenon.

“
The reality is that recruitment is hard. There's a reason that the recruitment agency market is a $969 billion market, and it's not because it's easy. The majority of recruiters and recruitment tech companies fail. It's one of the more complex sales processes you can imagine because the candidate at the end can turn around and say no.

In this week’s edition, we discuss:

• His hiring advice for founders and hiring managers

• Why he believes most founders in recruiting are focused on the wrong problem

• His advice to candidates and his playbook for getting hired

His hiring advice for founders and hiring managers

I’m excited to speak to Paddy because he’s seen the hiring problem front and centre, through helping 100s of founders hire. Few people in recruiting have that depth and breadth of hiring experience.

“
Finding talent is one part of the equation, but you've got to understand what the company is trying to achieve. Nine times out of ten, when there's a problem in hiring, it’s an issue on the company side; they haven't scoped what they need or the proposition. They can't articulate the job to be done is, or why someone should want to do it. The most obvious example is when companies say, “We're cool, we do AI. " You are one of 50. That doesn’t make you special.

I’ve seen this story before: “We want the best” without thinking through their motivations (and how much they want to be compensated). This is one of many complexities recruiters handle when helping hiring managers navigate a search.

”
People use heuristics in hiring. "What I want is a Cambridge Comp Sci graduate”. That's not really what you want. You want a really good problem solver with a specific skill set. But heuristics help. No human recruiter has complete information, so talent is often overlooked. We hire someone from Google, because if they pass the process to get a job at Google, they must be good, right? And that may not be the case.

A recent post that went viral on Reddit discussed this phenomenon. Essentially, it addressed the opposite issue: candidates are overlooked because they have worked for specific companies. While heuristics can be shortcuts, hiring managers who obsess over them are lazy. They haven’t done the necessary work to determine who they need, so they use badges as a shortcut.

This is one of the problems I hope is solved in recruiting. For too long, recruiters and hiring managers have had to rely on “shallow” data to determine who advances to a first-round interview. However, most startups are not trying to address this issue.

Why he believes most founders in recruiting are focused on the wrong problem

“
Many view recruiting as a typical sales funnel, a lack of nuanced understanding leads people to automation, “How can we remove humans from the process” and you end up with automated messages trying to get candidates to do one way interviews with bots. Candidates respond in kind, “If they want to use AI to assess me, I’ll use AI to cheat”. We’re rapidly approaching a zero-trust environment and companies are overwhelmed. I know multiple companies that no longer advertise their roles because the inbound volumes are too much to wade through.

The reality is that a low-trust environment benefits only a privileged few: candidates with strong networks and those from elite schools. Companies that want to escape this trap will still need human recruiters, despite what many startups will claim!

“
We’ll see more companies using agency recruiters because they prefer to outsource. Hiring managers will eventually question what is real anymore. I'm certain someone will employ a Synthesia bot with Claude operating in the background without realising it.

Paddy predicts that this will increase the number of in-person interviews. Google is already doing so. One step forward, three steps back for the hiring process!

His advice to candidates and his playbook for getting hired

Paddy finishes off with a masterclass in how to get hired.

“
People ask me for advice, especially job seekers, and the advice I always give is to follow the process I have used in every career move I have made in the last 10 years. I map out:

Paddy’s playbook

1. What is my goal?

2. Ideally, what does my next role look like?

3. Who are the 50 companies that I find exciting?

4. Who is the hiring manager?

5. What will capture their attention?

He then crafts a personal email that outlines why he’s a top applicant. This is a simple five-step process, albeit one that takes time. Instead of trying to game the applicant tracking system or using tools to auto-apply to as many jobs as possible, this “artisanal” approach to job applications undoubtedly leads to better results.

“
Ultimately, choosing a job is your biggest purchase. It’s bigger than buying a house or a car. It determines where you spend your time, which is the only thing you can never get back. Putting in the work just makes sense.

Wrap up

Whilst Paddy isn’t yet ready to fully share what he’s building at Dex, it’s exciting to think that there are people with deep domain knowledge who are obsessing about creating a better way to hire.

I share Paddy’s concerns that applicants will face more challenges in the short term as founders and startups try to square AI into every circle-shaped problem they see (I believe we’re at #2 below).

But as with every hype cycle, there is light at the end of the tunnel. I can’t see a world where AI doesn’t improve things long-term. We will, however, have to endure annoying chatbots and AI interviewers for a while longer.

But that doesn’t mean people are not getting hired. Mandy Liu, a career coach and former international student who eventually landed jobs at Meta and Nextdoor, knows a thing or two about getting hired on hard mode.

She uses her lived experience to coach others on how to get hired into data science in big tech. The results speak for themselves: she’s gotten 10 clients hired, and on average they’ve secured an extra $20-50K in compensation as a result of her coaching.

We discuss:

• AI’s impact on hiring

• Practical tips for job seekers

• How to prepare for the future of work

AI’s impact on hiring

“
AI has ruined hiring. As soon as a job gets published, there’s a flood of AI-written resumes and cover letters in a recruiter’s inbox. I’ve heard of hiring managers getting up to 2000 applicants within a few days. Most candidates are trying to optimize their resumes for applicant tracking systems to the point where their resumes are no longer appealing to a human reader. The result is a recruiter being inundated with junk.

It's ironic. These tools promised to save time for both sides and make it easy to find a match. So far, they’ve done the opposite, creating a flood of noise that helps no one.

So if optimising for an ATS isn’t the answer, what is?

“
My advice is simple. Talk to people. Those in your chosen industry, those who work in companies you admire or have jobs you’d want. Understand their day-to-day, and use that to learn the skills you need to showcase on your resume.

Practical tips for job seekers

“
Mindset is the number one thing holding people back from getting hired. Given how hard the market is, it can be easy to fall into a rut. But when I speak to clients, the first thing I do is to change their mindset and get them to believe they can land that dream job.

It may seem simple, but helping people out of a rut is hugely valuable. Spending months applying with no results can be heartbreaking, so having someone walk candidates through what they could be doing better is invaluable. And no, it’s not something AI can do yet (human empathy is what’s making the difference here).

“
If your one-page resume doesn’t do you justice, everything else is unlikely to count. Make it enticing to a human and not an ATS. You get one shot.

I spend a lot of time advising startups on their growth and go-to-market. I help founders realise that their website is confusing to someone landing on it for the first time without context. The trick is to figure out what outcome website visitors are trying to achieve and how the product being advertised can help.

The exact same logic applies to a resume. A hiring manager will have a problem. They will be looking at a resume, trying to evaluate whether a candidate is the solution to that problem. They will scan it for 10-20 seconds and then move on to the next one.

Whether it’s a resume or a homepage, delegating this task to a large language model (LLM) will likely lead to a bad outcome. The trick is figuring out how to use the LLM to assist in coming up with an asset that’s appealing to a human reader.

How to prepare for the future of work

“
I encourage everyone, even employees, to think of themselves as businesses. When I started creating content, people reached out asking for coaching, advice, job offers, etc. When I started investing in myself through coaching, my business reached new heights. Nobody knows what the future holds with AI, but thinking of yourself as a business is a good way to prepare for an uncertain future.

I could not agree more. I believe the lines between full time employee and a services business will become blurry, so thinking of yourself as a business can be hugely beneficial (see our previous post on fractional work to read more on this topic).

As AI tools become ever more powerful, the bar for a human employee will be raised across the board. But for those at the top of their game? - they’ll be more in demand than ever. Those employees will have all the leverage, as Mandy discusses below.

“
When you’re looking for a job, it seems like the company has the upper hand because there are more applicants than jobs. But you need to know your worth if you’re the right candidate. The average employee at Meta creates an incremental $1M revenue per year. So if you’re a top applicant, they should be begging you to join, not vice versa. When my clients realise that they tend to become more confident in their interviews

Wrap up

To say that AI is raising the bar for white collar work is an understatement. The hiring game has changed forever. But that doesn’t mean candidates should despair. Working the network. Finding career coaches. Continually upskilling. These are things that candidates can do to stand out from the crowd.

The change might seem scary, but even scarier is the thought of candidates and hiring managers aiming to one up each other through crude uses of AI. I for one look forward to the day where “AI first” applicant tracking systems and AI generated resumes and cover letters are behind us.

But I’ve heard many misconceptions. Aren’t you just a consultant? Isn’t it simply part-time work? I’ve also heard many questions from those considering the move: where to start, how to source work, etc. As you’ll see below, this trend isn’t going away anytime soon. My bet? It will accelerate even further in the next few years.

I was lucky enough to talk to Karina, the founder of Fractionals United. She launched her community at the beginning of 2023, making her better placed than most to discuss fractional work.

“
I became a fractional COO accidentally, before it was a thing really. I’d spent two decades in the publishing industry and was tired of reorgs and layoffs, so I started freelancing and fell into a pattern of running small businesses part-time (ie fractional COO work). Years later, after having briefly gone full-time with one of my clients, I came back to a different fractional landscape and couldn’t find a community for us to connect, collaborate, and learn from each other—so I decided to build it myself.

We discuss:

• How Karina defines fractional workers (hint: they’re not consultants)

• The pros and cons of fractional work and when to do it

• How to get started with fractional work

How Karina defines fractional workers

A common question from sceptics is: Are you not simply a consultant? I’ve avoided this label for several reasons. Startups (my target market) usually see consultants as expensive line items that often deliver little aside from fancy slides. I prefer the term “operator” (even if slightly cringe). I’ve been there. I’ve done that. I can get clients there faster by doing rather than telling. This is what gets you hired at a startup.

“

Consultants are external and project-based, even if it’s a long-term project. Fractional workers are part-time embedded leaders. We are part of the team, with people reporting to us, and are long-term (ie, we don’t deliver and move on). The only difference between us and full-timer leaders is that we are part-time.

This is why I love interviewing people for this newsletter. That’s a better framing than I’ve ever been able to come up with. Most of my engagements have been long-term, even if I didn’t initially expect them to be.

“

I don’t do proposals; that’s for consultants. I do two-way job interviews. I want a job description and understanding of my responsibility, domain, and team. I also want to be sure that I can work with the CEO and that they’ll allow me to do the job he or she is hiring me to do.

Ironically, most fractional workers I’ve encountered who haven’t succeeded have taken the opposite approach—the hard sell. Overpromising a short-term outcome is a terrible way to start a working relationship (especially when it’s not delivered).

The pros and cons of fractional work and when to do it

“

After the pandemic, many people didn’t want to return to their previous ways of working. They want control over their time, which fractional work can provide. However, there are downsides: figuring out how to maintain a pipeline, paying your own taxes, and managing multiple contexts and clients simultaneously. It’s not easy.

I discovered this the hard way: I do not multitask as well as I thought I could. Managing multiple Slack workspaces, emails, and stakeholders? It quickly got to a point where I realised less is more. I now work with a maximum of 2 clients at a time.

We then move on to when someone should consider fractional work.

“

A fractional worker is a senior leader. We can work fractionally and still deliver value because we’ve done this job full-time and have the experience and skills. We know how to make the most of our resources, both time and team, and what mistakes to avoid. This takes time.

This experience is what allows fractional workers to save their clients time. Thinking back to how you can save your clients time is ideal framing for those considering the move to fractional work.

How to get started with fractional work

As with any product you’re selling, you need to define some basics before getting started. Figure out your edge, your buyer and where they’d usually buy.

In my case:

• Edge: I’ve led growth teams at marketplaces and helped several marketplaces with international expansion.

• Buyer: Early-stage marketplace founders find my services most valuable when they’ve found product-market fit and want to scale (geographically or otherwise).

• Where they buy: They usually seek help from experts focused on growth & marketplaces through accelerators and/or niche marketplace communities.

Looking at my service as a product, the first two define my product market fit, whilst the third is my distribution. As with a product, distribution is the biggest priority.

“

Networking and business development is critical, but so is knowing your target and niching down. The best way to get started is by alerting your immediate network and joining communities like Fractionals United. Both will allow you to make connections with fractionals adjacent to what you do, for your target, and refer work to each other.

I strongly believe in the power of community. I’m closely involved with two, Everything Marketplaces (as a community EIR) and SYSTM (as a growth coach). These are often great places to be “discovered” outside my own distribution. Not only that, but I’ve also met some extremely smart people who I’m lucky enough to call friends. As you’ll discover, having great people around you matters a lot in fractional work.

Wrap up

In summary, fractional employment can be great when it works. But it’s not easy. For those considering a move, I always recommend having comfortable reserves to get through the initial months before those first clients land.

But when it does work, it can be some of the most freeing work ever. If you’re considering the move: surround yourself with others that have been through the journey, join relevant communities and don’t be afraid to put yourself out there. And if I can ever help, please don’t hesitate to leave a comment below or reach out 👋

This was a good trade-off as eager grads could “learn the ropes” and get their foot in the door. But with grunt work delegated to AI, have we sleepwalked into a situation where that door has now closed shut?

In this week’s edition, I talk to Peter Wood, Founder and CEO of The Graduate Guide. After seeing this problem up close, he launched a community to help. In the UK alone, 54% of employers find it challenging to recruit for graduate roles1, and 40% of graduates cannot find high-skilled work2.

He believes the problem can be solved through the power of community.

We discuss:

• His advice to young people looking for their first job

• What universities need to change in the AI age

• How communities like the Graduate Guide can help

His advice to young people looking for their first job

Peter’s story is different from what his LinkedIn profile might suggest. He got into one of the best universities in the world, University College London (UCL). You might assume finding a job was easy for him, yet he feels that academia did not allow him to play to his strengths.

“

I wouldn't have gone far if I’d continued the academic route. My career started by simply reaching out to people and being curious. They’d wonder how I’d come across them, but most of the time, it was as simple as saying, “I like this poster; I wonder who created it,” and then following up to see if I could find out. In most cases, people were flattered that I wanted to speak to them. Many of these people eventually supported me in my career.

This advice is helpful for anyone, regardless of age. Reaching out to people whose work you admire can lead to hidden opportunities. Given its effectiveness, why is it still rare to see people doing this? Embarrassment? Fear of rejection? Societal norms?

“

It's a societal issue. Once people become successful, they’re often asked, “How did you do it?” The answer is pretty simple: doing the things that most people would find embarrassing, putting yourself out there. One thing you can do on a micro level is become comfortable with doing things others might find embarrassing and not take yourself too seriously. The worst that can happen is you fail, and there’s value in that, too.

Again, this is great advice regardless of age. It’s easy to idolise “overnight” success stories without seeing the grit, determination, and multiple failed attempts that preceded them. Yet it isn’t something taught in universities.

What universities need to change in the AI age

Exam-based assessments have always been flawed and are often useless in gauging an individual’s effectiveness in the workplace. Emotional intelligence usually ends up being much more helpful than IQ. Peter has a unique perspective:

“

I hope AI results in education being treated less as a box-ticking exercise. The largest virtual school in the UK (Minerva) is innovating in assessing students. It allows them to opt into different exam boards. One might focus more on writing and the other on oral communication. People should be able to lean into their strengths, and this is where AI can help to cater to the individual's needs, rather than a one-size-fits-all approach.

I’m also curious how universities have adapted their career advice over the last two decades (back when this old soul was a grad). It seems not a lot has changed:

“

Curiousness and willingness to reach out to others have not been incorporated into university education. Yet this is so important, and I’m trying to help normalise it through The Graduate Guide. We need to enhance young people’s ability to reach out to those with experience who are further along in their careers. In-person relationships will become even more crucial as AI floods online applications.

I’m bullish on communities like the Graduate Guide. AI is already flooding the open web with bland, mass-generated content. Closed communities can counter the noise and help create genuinely valuable connections—which is much more difficult on LinkedIn.

How communities like the Graduate Guide can help

“

Being on LinkedIn puts you in the top 5-10% of people with career consciousness. But there’s a whole universe of people not on there. We need to reach those people too.

Admittedly, I’m sceptical of social media in the long term. You’ve probably already seen AI-generated comments and posts on LinkedIn, and I think this will only grow as a trend. With its rage-baiting and click-bait videos, it’s getting harder to distinguish LinkedIn from TikTok.

This is where communities like the Graduate Guide can provide an alternative and foster deep connections that provide helpful guidance for those starting their careers.

“

I have a following on LinkedIn, and I care about reaching people because it’s a part of our mission. But I ensure I’m reaching the right people, either students or founders or anyone who can contribute to our mission of empowering the next generations to explore and build their careers. Since I started building this community, viewership has gotten much stronger, and I haven’t spent a lot of time trying to hack LinkedIn’s algorithm

Wrap up

Getting a foot on the career ladder has always been difficult. But when so much of the work that young graduates do is now replaced by AI, it creates a problem that affects them in the short term but affects all of us in the long term.

We should not sleepwalk into a situation where talent pipelines are starved of early-stage talent. This has grave implications for the future of work that are not being discussed enough.

I’m rooting for Peter and The Graduate Guide, and I believe many other upstarts can contribute to solving this problem. However, our educational institutions, governments, and those further along in our careers must also be part of the solution.

We’ll all feel the pain soon enough, if not.

If you pay too much attention to certain “thought leaders” on socials, you might conclude that AI has rendered the profession obsolete. You only need a subscription to Cursor, Replit, or any of the many AI tools available to conclude that writing code is a solved problem. Right?

Talk to anyone who has been in the details or extensively uses these tools, and you’ll hear a very different story. It’s true that – in the right hands – they offer large efficiency gains. But we are nowhere near eliminating the role of a software engineer.

Jordan Cutler is definitely in the details. He has extensive experience using AI for software engineering and advises others on this topic in his newsletter.

“
The landscape is changing very fast. Some might think: what’s the point of learning these skills in the first place? Rather than eliminating software engineers, these tools provide opportunities to up level them. Each layer of seniority will be elevated. Junior + AI = Mid-level. Mid-level + AI = Senior. Senior + AI = Staff. And so on and so forth. There will be increased efficiency across the board.

In this week’s edition, we cover:

1. The advice Jordan gives to entry-level software engineers

2. How to use AI to improve communication and design better systems

3. The engineering skills that companies will hire for in an AI future

The advice Jordan gives to entry-level software engineers

Jordan advises entry-level engineers to approach their skill set as they might approach managing a portfolio. Diversification is key.

“

Develop your people and AI skills (some recommendation in the footnote1), use AI to solve problems, and ensure you have the people skills to work with anyone.

He then breaks it down even further with a framework for entry-level engineers to use for evolving their careers.

The 3 Es for entry-level engineers:
Education: Evolve your understanding of the foundations, for example, embeddings and AI's probabilistic nature and randomness. There are AI courses on Udemy that can help.
Exposure: Curate your “information diet” and read content from credible experts, whether it’s newsletters or following them on LinkedIn. (Recommendations in the footnote2)
Experience: Start building products. AI can guide you to an output, and you’ll quickly understand whether your code has worked or not.

Using this framework to read credible content gives the quickest route to discovering the best AI tools. Learning how to master them is the best way for engineers to up skill themselves to a mid level role, so practice is key. We’ve listed some great resources to feed your brain at the bottom of the post!

How to use AI to improve communication and design better systems

We move on to how engineers can use AI to enhance their own communication. Being able to get a framework onto paper (even if digital!) for others to understand is a skill that’s useful for everyone, which Jordan discusses:

“

For engineers, being able to use tools like chatPRD and WriteEdge gives them a flexible way of writing precise product requirement documents (PRDs) by voice dictation without needing to start a document from scratch.

This moves the focus from building a well-formatted output to refining a proposal. We’ve all seen nicely formatted memos or slideshows that help to “sell” an idea that hasn’t been thought through. Time spent formatting these outputs could be spent refining the actual idea instead.

This is the power of AI. It minimises the time spent on formatting, so more time can be spent on the idea itself.

Jordan also highlights a similar benefit when working cross-functionally with designers, for example:

“

There are times when I’ll have an idea for a product feature. The options for communicating it to a design team are:
1) Explaining the feature verbally.

2) Putting low-fidelity designs together on a tool like Excalidraw.

3) Building a prototype.

Each is better at explaining an idea than the previous one but requires more effort. Tools like v0 can help with producing higher fidelity designs with minimal effort.

Using v0 you can turn a simple prompt into a working front end prototype, which often reveals missing details. It makes it easier to get an idea approved by a cross functional member of your team, because it’s more refined and considered.

The engineering skills that companies will hire for in an AI future

Finally, we discuss how engineers can prepare for the jobs of the future. Jordan advises reframing what an engineer should consider “their job”.

“

You shouldn’t view your job as simply taking tickets and executing them. That job is probably going to be eliminated in the next 5 years. Focus on building relationships with other cross-functional colleagues. Engineers should use AI to up skill in other areas, to collaborate with an iterated proposal (designs in Figma, PRDs, etc.).

This is a common theme from my conversations with domain experts.

The overly specialised worker that only does a single thing well? AI is likely to free up their time, which they can then use to take their work to the next level and work on things that were previously out of their scope.

Jordan also highlights the relative ease of getting started with learning a skill:

“

A simple example: I had a friend, a backend engineer, who wanted to get started with a new front end programming language. With a search engine he’d have to traverse multiple links with different answers that are difficult to navigate. With AI? Asking the right questions and prompting effectively, gets him there much faster

Wrap up

It’s easy to get caught up in the AI hype, but it’s clear that engineers are not going to be out of a job at any point soon. That’s not to say nothing has changed either, though. As with most desk based work, engineers will need to get comfortable with working outside their scope and constantly up skilling. The only constant is change, so being comfortable with it is no longer a nice to have.

In part 1, we discussed when and why founders should make their first CX hire and Susana’s framework for CX teams. This post discusses 10x employees, thoughts on future hiring, and the cold email to Brian Chesky that got Susana hired at Airbnb.

10x CX employees, and thoughts on future hiring

If AI will raise the bar for CX, what of the 10x CX employees?

Susana highlights traits like empathy and being passionate about helping solve customer problems as the key. She also mentions the importance of being a continuous learner and teacher. The latter might have you scratching your head.

Susana explains:

“
By teacher, I mean seeing an issue, confirming it as a trend and providing those insights to decision-makers with recommendations. Saying, “I’ve seen a customer trend, affecting Y% of users and moving CSAT by Z points. Here’s the documentation and write-up”. They are teachers to decision-makers about their customers. Notice how much value that adds compared to someone simply describing CSAT movements.

During our conversation, Susana repeatedly emphasized metrics as the key to a great CX hire. She also discussed how this hire would effectively advocate for the customer when prioritizing a product’s roadmap.

“
Instead of simply requesting a feature, it’s better to say: “Here’s a request and the data to back it up based on user feedback”. This focus on data allows CX teams to work effectively within their teams and cross-functionally. It allows them to influence others whilst representing the customer. AI changes the tools, but the ability to influence will continue to make the difference between good and great.

Susana uses her expertise to advise startups on hiring in CX and notes an interesting trend:

“
I’m speaking to startups that are changing the type of CX person they hire. They’re finding that most candidates are not yet following the trend. They’re stuck in the old way of doing things. With AI, there will be new skill requirements; it’s not enough to say you can do busy work. That will be minimised over time. CX hires must become more comfortable with analytics, understanding data, and improving processes.

This points to a similar trend across functions: the worker that scales themselves.

It’s not about eliminating humans. It's about using AI to eliminate tedious work and using the extra time to focus on delivering delight.

In CX, that means supporting customers on the front line and effectively representing them to internal teams. Being comfortable with data is the key to representing customers at scale, a skill Susana developed at Airbnb.

The cold email that got Susana hired at Airbnb

“
I was a 20-year-old in college with no idea what I wanted to do, but I admired Airbnb and saw they were hiring in CX. I applied online but didn’t hear back. So, I did some research and found that Brian Chesky was a RISD alumni (one of the first art and design schools in the US). I designed a resume, knowing this was a potential way to stand out. Brian responded. That resume got me a first-round interview, and ultimately, I was hired!

Susana would use the same tactic to secure her role at Loom. If there’s one thing to learn from this experience, a single email can be the difference to getting that dream job, customer, mentor, etc.

But it takes time and patience to uncover what someone cares about —even more time to craft an email that speaks to that. In Susana’s case, it’s been time well spent.

Wrap up

It dawns on me that Susana's skill in sending cold emails is also the skill she uses to run CX teams: empathy.

She puts herself in her counterpart’s shoes and figures out what delightful looks like. She then meticulously works back from that to ensure she stands out.

In a world where AI promises to eliminate so much of the work we dislike and find tedious, I’m hoping it leads us to focus on delivering delight. The empathy required to get there is something I predict (hope?) remains the focus of humans.

Should you ban job applicants from using AI? Hiring managers are worried. Some are even banning candidates from using AI in their job applications. That’s pretty ironic given how long companies have been using AI to filter candidates via keyword searching, but let’s leave that rant for another day.

Banning job applicants from using AI is problematic for a few reasons:

1. If they are bright enough, a hiring manager will not be able to tell.

2. They should use AI because they will need it in their jobs. Banning it would be like banning someone from using the Internet.

3. You can learn a lot about a candidate from how they use AI. More than you could from brainteasers or vague “tell me about a time when” questions.

The key point that most are missing? It’s not the final output that’s useful.

It’s the prompts.

Prompt engineering: the process of crafting precise instructions to guide generative AI models in producing desired outputs. An effective prompt engineer is a strong communicator who has critical thinking and subject matter expertise.

If the above sounds familiar, it’s because these are precisely the things that make for an effective GTM hire.

They must be effective communicators to talk to your customers and translate those conversations into appealing copy (i.e. language market fit)1.

They’ll need critical thinking to find the right channel to continually reach more customers and avoid the law of shitty click-through rates2 from slowing your growth.

They’ll need to be subject matter experts in your product, customers and industry to understand how to best position your product in the market.

Letting candidates use AI and assessing their prompts gives you insight into how effective they are at all of the above. It gives you an insight into their mind.

Here are 3 things you should be evaluating in a candidate’s prompts

1. How effectively do they communicate? Can they provide detailed context? - A good prompt should read like instructions you’d provide to a super bright but inexperienced intern with zero context. Would you be confident the intern could complete the task after reading the prompt?

2. Are they creative in their prompting? Do they know how to get the best out of an LLM? For example, do they provide example outputs, use chain-of-thought prompting or ask the LLM to take the time to think and provide clear roles (amongst other things suggested in Anthropic’s prompt engineering guide)?

3. Are they effective subject matter experts? After each output, you should see them correcting the LLM and guiding it to a better answer. For that to work, they’ll need to know what good looks like. Each iteration should provide clear guidance for how to improve the next output.

Wrap up

In short, rather than banning AI, hiring managers should embrace it for GTM hiring. We have an exciting opportunity to rethink how we work and hire, and the future belongs to those who figure out how to make these tools work in their favour.

AI will eliminate CX. Said literally no one that has any practical experience working in CX, ever. That’s not to say it won’t evolve and raise the bar for customer experience, though. Just not in the ways most people think.

It doesn’t get more practical than Susana. She joined Loom as its first customer-facing employee when it had only a handful of people and huge ambitions.

I had a remote onboarding with one of Loom’s founders (Joe). It was the morning for me but 1am his local time. He said, "OK, here’s the spreadsheet with passwords and logins. I’ll see you in eight hours.” I loved that. I was so happy to dig in and figure things out myself.

Susana would go on to build Loom’s customer experience function from a scrappy startup to the household name that eventually sold to Atlassian for close to a billion dollars. These days, she uses the lessons learned at Loom and Airbnb to advise founders on how to set their CX teams up for success, keeping the customer at the centre of everything they do. We discuss:

1. When and why founders should make their first CX hire

2. Susana’s framework for CX teams

3. 10x CX employees, and her thoughts on future hiring

4. The cold email that got Susana hired at Airbnb

When and why founders should make their first CX hire

This is a question Susana gets often. It’s especially pertinent given how much is changing about how we work and build companies in recent times. A founder may be tempted to outsource this responsibility or even automate it. But that would be a miss:

When founders feel overwhelmed managing support, they miss key feedback. I’m a huge believer in founders' spending time with users. It's one of the key things I learnt at Airbnb, where Brian Chesky didn't just meet with users, he'd live with them and would turn their feedback into a product roadmap1. I went on to see the same thing at Loom, where the founders were still dealing with support tickets even after I was hired. It’s so important. But if founders have their attention elsewhere, it can degrade the user’s experience. They should hire, and hire with care, so that users continue to get the time/ attention they deserve.

I like that she prioritises the customer's perspective rather than the founder's or team's. In the long term, capturing, measuring, and categorizing customer feedback can save teams considerable time, so it’s ultimately in their best interest anyway.

What impact does she think AI will have on this first hire?

AI will delay the first CX hire and impact how quickly a CX team needs to grow. But, the combination of humans and AI will be the winning one. CX as a function will certainly still be crucial, but AI will change how much a small team can get done.

As in most functions, AI will allow employees to scale themselves. It will reduce the tedious and manual work that must be done, which begs the question: where should they spend their newly discovered time?

Susana’s framework for CX teams

We discuss Susana's customer experience with British Airways (BA), as an example of how not to do CX.

I was flying to New York with a connecting flight from London. I was ready to leave for the airport and got a message saying my second flight was cancelled. I messaged support to ask about my options. A chatbot asked me to confirm my details and asked me what the issue was. After re-explaining the context back, I repeatedly got the response: “Looks like you want to cancel your flight? Please confirm”. Not only had it completely misunderstood the situation but it was ready to take action on it!

Anyone who has travelled with BA is probably unsurprised. They provide a great example of how not to insert technology into the customer flow. OK, rant over.

Susana then gives a masterclass on what BA should have done instead:

They should have proactively figured out that I had a cancelled flight. Confirmed the context of the situation with an agent. Then, that agent should have found a solution. They could have kept me in the loop as the situation was resolved. I shouldn’t have to contact BA, clunkily confirm my details, and the situation etc. They should figure out the context, anticipate the problem, and tell me they’re finding a solution.

Here are the key steps she recommends for CX teams to support customers:

1. Identify customer: Who is this person, and how important are they as a user?

2. Gather context: What’s their current situation? What do they need to do (if anything?)

3. Take action: Decide what to do and what needs to happen

4. Communicate: Contact the customer proactively before they need to reach out

This framework allows you to understand where AI can be most effective. Helping the support agent understand the customer and context while helping them decide on the best course of action seems perfect for AI. Notice how much better this seems than simply putting a chatbot before a customer.

The key is to have AI enhance what’s happening on the backend while preserving and enhancing the human-facing communication.

Susana sees the future of CX as one where a support ticket doesn’t need to be triggered by a customer issue; the magic will be in AI anticipating an issue before it becomes one. If and when there are issues, using AI to enhance the customer experience by keeping customers up to date. It should serve as a co-pilot to the support agents who are responsible for the solution (by creating context, surfacing information, etc.).

Part 2 of this post below, where we discuss 10x CX employees, Susana’s thoughts on future CX hiring, and the cold email that got her into Airbnb

How do you prepare for the future as an SDR? How about as an SDR or revenue leader? We’re in the most significant change to how we work we’ve ever witnessed. AI is tearing up the rulebook across every function in white-collar work.

This is especially pertinent for SDRs. It’s one of the roles that – done effectively – requires a lot of creativity, empathy and outside-of-the-box thinking. Intuitively, you might think this would mean it’s safe from disruption. But it’s not that simple.

We’ve all had bad SDRs clunkily sending us emails/DMs. After scanning these messages, you’re left scratching your head, trying to figure out how you even became a target. In many cases, it’s simply “spray and pray”, where the SDR hasn’t taken the time to identify an ideal customer profile (ICP), much less look through yours.

The opposite is also true. We’ve all received an outbound message that’s felt like magic. It doesn’t feel like sales. The best sellers act as sherpas or guides that add value by showing prospects how to do their job better. Sometimes, with the product or service they’re selling, other times by referring them elsewhere.

The clunky SDR will be eliminated. In a spray-and-pray approach, the wider the “spray”, the better. If the total addressable market (TAM) is large enough and the annual contract value (ACV) is low enough, the unit economics do not justify putting a human to work. This is where jobs will automated. An AI agent makes more sense and effectively reaches more prospects. See the chart below for illustration1.

What about the more sophisticated SDRs? Should they pack up and go home, too? Let’s break down the job of a “sophisticated SDR”. Broadly, there are 3 key stages.

1. Research: finding prospects that match an ICP, identifying where to find them and understanding their most pressing pain points. At this stage, the objective is to build a target list and gather useful information that will help in the next stage.

2. Outreach: involves writing an outreach message that gets a prospect's attention. It involves talking to the prospect about a pain point in their daily lives. Done effectively, this should not feel like selling.

3. Qualification: once an SDR has understood the prospect well enough, they are ready to pass them on as a qualified lead to an account exec or sales rep responsible for progressing them through the funnel.

In theory, AI could perform all of these tasks, but doing them well requires creativity and thoughtfulness, which AI doesn’t have (at least not yet!).

For sophisticated SDRs, rather than considering where an AI agent can replace them, it’s more helpful to think about where an AI agent can help to scale them. To find more prospects and qualify more customers.

Let’s explore each stage through this lens:

Research

This is where AI can help the most. Anyone who’s spent time using tools like perplexity will know how much of a step change it is for research. With the proper context, inputs, and tools, SDRs can increase their efficiency massively at this stage.

Outreach

It’s easy for AI to generate an outbound message. Generating one that stands out is much harder and will become even harder as AI agents are further adopted and inboxes are flooded. Human empathy is most valuable at this stage. Figuring out what a prospect cares about and how to provide value is not easy for humans, let alone AI.

That’s not to say AI can’t help with brainstorming, iterating on copy and executing on actually sending the message. However, human judgement must prevail at this stage to determine what a prospect will care about the most and how to get their attention.

Qualification

If the initial outbound resonates enough, a prospect will provide an SDR information that an AI agent would struggle to find. The SDR should be in the driver’s seat now. If a prospect thinks they’re talking to an AI, they will lose interest. As much as tools improve, human psychology stays the same: we prefer speaking to other humans, at least the ones we like.

AI can still help here, but it should be less customer-facing. It should help the SDR avoid wasting time on questions an AI agent can answer. Instead, the SDR should spend time gaining new information about the prospect’s current context.

Closing thoughts

Unsophisticated SDRs will need to upskill, and sophisticated SDRs will need to scale. For the latter, figuring out where to use AI and where a human touch is valuable will be the skill that differentiates them from the crowd and ensures they are not replaced by an SDR using AI. Broadly, the “customer-facing” touchpoints will need to feel more authentic and human. Everything else is up for discussion.

1

The Signal

Should You Buy an AI SDR?

If you were forwarded this newsletter, consider joining over 1,600 weekly readers—some of the smartest GTM founders and operators—by subscribing here…

Read more

a year ago · 4 likes · 1 comment · Brendan J Short