You ship or you die
Agencies don’t get the luxury of “we’ll ship next quarter.” A retainer ends. An invoice goes out. A client signs off. The cadence trains a kind of decisiveness that’s genuinely rare in product organisations that have never had to bill for time.
When money is on the line every month, you learn to stop waiting for perfect conditions and start making calls with what you have. That muscle turns out to be one of the most transferable things from agency life, and one of the hardest to explain to people who’ve never experienced the pressure of it.
Scope is a tool, not a constraint
There’s a reason the Iron Triangle is such a durable idea in project management: scope, budget, and time are always in tension. You cannot move one without touching the others. In an agency, you live inside that tension constantly. It stops being abstract pretty fast.
But the real challenge was never the triangle itself. It was getting out of a client’s head.
You write the brief. You build the prototype. You walk them through the mock-up. You iterate. You get sign-off at every stage. And then, and this happened more than once, the project ships and suddenly it’s: “Why does this work this way? Why is it designed like that?”
The honest answer is: because you approved it three months ago. But that’s not a useful answer.
What I learned is that a lot of clients don’t fully process what they’re agreeing to until they’re living inside the finished thing. The prototype is abstract. The real product is real. And that gap, between what someone imagines and what they experience, is where scope conversations actually live.
So scope discussions become the work. You learn to read what a client actually needs versus what they asked for. You make the trade-offs explicit before you start. You name the triangle out loud: if you want to add this feature, we drop that one, or the timeline moves, or the budget goes up. Which is it? That muscle transfers directly to product roadmapping, maybe even more than I expected.
The team is the product
You can’t out-process a bad team, and you can’t under-invest in a good one. That’s the clearest thing I know from those seven years.
When I hired, I always looked more at soft skills than technical ones. The most important things were a growth mindset, openness to feedback, and the ability to learn from mistakes, genuinely learn, not just nod and repeat the same thing. Technical skills you can teach. Mindset is much harder to shift.
And here’s the thing: in the AI era, that’s becoming even more true, in a different way. Growing technical skills is easier than ever. But what becomes differentiating is something else entirely. The ability to structure problems, to supervise AI systems well, to understand systems deeply enough to simplify them. A junior with excellent reasoning and genuine AI fluency may outperform a traditional senior who mainly accumulated framework knowledge. Meanwhile, elite systems thinkers become even more leveraged than before. The middle of the distribution compresses. Average developers become more capable. Top-tier thinking becomes more differentiated.
The question was never really junior versus senior. It was always: who actually thinks?
What moving into product taught me about the agency years
When I made the move, I thought the biggest shift would be about the product. It turned out to be about the people. Building a team, building a culture, that was the work. I hadn’t expected that to be more true on the product side, but it was.
Some things transferred cleanly. Creating a mission. Thinking about competitors and the market. Running a team. A lot of the operating system was familiar.
The sharpest difference was focus. In an agency, you have many clients and you adapt constantly. In a product company, you have one thing, and you go deep. That depth was something I hadn’t fully anticipated, and it turned out to be one of the things I valued most about the switch.
There was also something I had to unlearn. In agency work, you develop a certain resourcefulness, a confidence that you can always find a way to deliver something, even under difficult conditions. That’s genuinely useful. But in product, “something” isn’t enough. You have to deliver something actually good, because there’s no client relationship to smooth over a mediocre outcome. The market doesn’t grade on a curve.
And then there was the frustration I hadn’t known how to name at the agency: you’d spend months, sometimes over a year, building something real, and then the client would shelve it. Internal priorities shifted. Marketing never happened. The founder decided it didn’t fit the vision anymore. The product just didn’t get used.
That was hard. You put so much into building something and it disappears without trace.
Working in product, you can see the impact of your work more directly. That, more than anything, is what I didn’t know I was missing until I had it.