Most Agencies Won’t Survive AI: Igor Ivitskiy and Paul Ryazanov on What Comes Next

A lot of agencies will just die. Igor Ivitskiy says it plainly, sitting across from Paul Ryazanov in Manchester. Not most in some vague sense, but the majority in every single niche. What follows is his case for why, and what the ones left standing will look like.

Winning clients was never the hard part

Paul frames the issue early. Agencies pour enormous effort into winning business and then fail to deliver at the level they promised. That gap is where the industry is breaking.

Igor’s diagnosis is more specific. The average agency never builds a feedback loop from client results. They set up advertising or SEO services, produce some outcome, and then simply move on. Those results never get reused for the next client or the next industry. What the agency actually runs on is a playbook written five years ago, applied to every new account as though nothing has changed.

This is where he sees AI mattering, and it isn’t the use case most people reach for. The value isn’t in generating campaigns faster. It’s in finding the patterns inside results you’ve already produced, then folding those patterns back into your processes. Handled that way, every client makes the agency measurably better at serving the next one. Handled the current way, every client is a fresh start.

Two ways to narrow, and one way to lose

Asked what a successful agency looks like in 2027, Igor doesn’t hesitate. Narrowing focus. There are two ways to do it and both work.

You can narrow by vertical. An agency exclusively for solar panel installers, or for a specific kind of e-commerce store. Or you can narrow by tool, becoming the people who genuinely know YouTube ads or local service ads better than anyone.

What doesn’t work is the default. Most agencies, as Igor puts it, try to be all-in. Any niche, any tool, everybody. That approach collects money from a broad market in the short term, and it’s a poor long-term strategy.

Paul’s counterexample is worth sitting with, because it complicates the advice. His company runs forty people across automotive, pharmaceutical, and other verticals. Broad on the surface. But internally, the specialist who handles a pharmaceutical client keeps every pharmaceutical client that follows. One person owns all automotive. The company looks generalist from outside while every individual inside it is a specialist with real depth in a niche. Focus doesn’t have to mean shrinking. It can mean structuring.

Why the billable hour is finished

Paul is blunt about where this leads for pricing. In development work, charging by the hour with AI in the picture means the business model is simply gone.

His reasoning is about what clients actually buy. They don’t care how much they pay in the abstract, and they certainly don’t care how many hours went into their account. They care about results. How long something took is an internal problem, not a client-facing one.

So his company is restructuring around outcomes. Work out the balance of budget and expense, understand what the client needs to achieve, and price against the result. He’s candid that it’s still in progress rather than solved.

Igor’s read on the shakeout is that the survivors will thrive and create considerably more value than they do now. He’s particularly optimistic about one-founder agencies. A single person, a stack of AI agents and tools, a narrow niche, and a tight feedback loop. That combination can now compete in ways it couldn’t before.

Finding a partner rather than an employee

The conversation’s most personal stretch starts with Paul admitting something. He’s good at delegating work. He has project managers. What he can’t find is someone who will take real responsibility, not just for clients and projects but for the unglamorous parts, the finances and operations and billing.

How do you find those people?

Igor’s answer is that only one thing has ever worked for him: a partner agreement. The title comes with a share of the company. That’s the mechanism that gets someone working for results rather than for time.

He doesn’t pretend it was easy. The first time he gave half his company to a business partner, it was genuinely hard. He’d built the thing from scratch, grown it, gotten it running properly, and then handed over half.

Five years on, he calls it the best decision of his life. Better to hold half of a pie that size than the whole of a smaller cake.

Where he finds these people is the part most people won’t expect. Business clubs. He belongs to Ukrainian business clubs, several in the US, the Perry Marshall community, groups across Singapore and elsewhere. Every time he travels he meets people, spends time with them, shares ideas. He describes it as natural filtration. After a hundred minutes with someone you know whether the chemistry is there, and if it is you suggest working on something together.

One of his partners came from the same field. The other came from a completely different background and brought a structure of operations nobody in their industry was using. That outsider perspective turned into an advantage.

Paul’s response is the honest one most agency owners would give: he should probably make the time to actually go do this.

What separates the survivors

The through-line is that the agencies which survive won’t be the ones that work hardest at winning clients. They’ll be the ones that learn from the work they’ve already done, know exactly who they serve, charge for outcomes instead of hours, and find partners with genuine ownership rather than employees with job titles.