I joined Robert Craven on the GYDA podcast, and we worked to a title that doesn’t leave much room for small talk: “I left my own agency in 2024. I left because of mental health.”
Aira was part of a programme that Robert ran with Google around 2016/17, back when we were one of a table of ten agencies all trying to figure out how to grow. So the conversation starts there, with what actually took us from 10 to 50 people, before getting into the harder stuff: the evening I came home from London, sat down on the sofa and couldn’t stop crying and the warning sign I only recognised afterwards. I’d started cutting my dog’s morning walk short because I felt like I was late for my desk at half seven. Nobody was telling me I was late.
We also get into something I see a lot more clearly now I’m on the outside: the LinkedIn version of agency life, where everyone appears to be smashing it. I now know some of those posts well enough to say: I was there, and that is not what happened.
The full conversation is below, with the key takeaways underneath.
Key takeaways from this episode
The early growth secret was boring: senior hires and client retention. Aira’s first few hires were deliberately mid-level or above, people who already knew how to work with clients. That foundation meant the agency didn’t lose a single client for the first 18-24 months. If you’re churning clients every six months, you have to replace them before you can even think about growth.
In the AI era, the client relationship is the moat. Strong relationships have always mattered, but now you can’t give a client a reason to question you as an expert or a strategist. A tool can process context; it can’t sit down with a client and get under the skin of their business. If Paddy were starting again, he’d put even more emphasis on that human value than he did ten years ago.
Your job is to effect change, not deliver reports. A line from an external trainer early in Paddy’s career at Distilled that stuck: agencies that have been genuinely helping businesses by effecting change, rather than delivering reports, are in a far stronger position to ride out what’s coming.
Normal problems, compounded, are what break you. Nothing that went wrong at Aira was unusual. Win a client, lose a client, hire someone, lose someone. It was the relentlessness over two to three years that did the damage and Paddy didn’t see it because each individual issue looked like something you just deal with.
The warning signs are small and behavioural. Cutting the dog walk short. Skipping the gym, skipping breakfast, getting more agitated with your partner than usual. Paddy thought he was self-aware, and he spotted none of it until he was out the other side. Watch for the small behaviours that aren’t like you.
LinkedIn is not a representative sample. Most agencies in 2026 are either struggling or plodding along level year on year, and there’s nothing wrong with that. But if you believe the feed, everyone is winning except you, and that comparison is corrosive precisely because it’s built on posts that leave the hard parts out.
Forget company purpose; know your personal one. Paddy is openly sceptical of agency vision and mission exercises, but firm that individuals need to know why they’re doing this. It’s fine if the answer is money, because it’s really about what the money enables. His own purpose right now is deliberately small: to get up in the morning and not hate his job.
What got you here won’t get you there, and you don’t have to do all of it. Around 25 to 30 people, departments and tribes emerge and the founder’s job changes. It’s legitimate to say “I didn’t sign up for this bit” and hire around it. Paddy and his co-founder also learned that two founders can want different things from the same business, and that’s workable too.
There may be a headcount sweet spot, and it’s smaller than the industry’s default ambition. Paddy was happiest somewhere between 30 and 40 people, and observes that agencies hit their most serious problems pushing past 40 to 50. Robert tells the story of an owner who capped his agency at 34 on purpose: small enough to know everyone, big enough to matter, and no happier at 60.
To stop the compounding: talk, and shrink your involvement to what only you can do. Paddy wasn’t even telling his co-founder how hard he was finding it. Talking to someone, therapist or otherwise, comes first. Then look honestly at everything you’re doing: post-Covid he got closer to clients his team already had covered, because being under pressure made him feel he should. It didn’t help them, and it drained him.
