Author of Scale At Speed https://xitter.cf/t.co/5EKuSpVMD7 Founder, https://xitter.cf/t.co/k2vxOxGwgS the world's leading agency accelerator Try out https://xitter.cf/t.co/9GP3Lkgh1q
London
Joined July 2007
- Tweets10.7K
- Following1.8K
- Followers2K
- Likes831
Pinned Tweet
*When stuff that's not in the transcript suddenly becomes findable!*
I am so proud to announce GoldenRetriever.ai is being launched in public beta today on ProductHunt – please upvote it for me!
It's our new software that does a truly revolutionary thing: searches for moments in videos and docs and audio, like when dad and grandma were both laughing, or when that slide about EBIT appeared on a video call.
Please check it out! producthunt.com/products/gol…
On day one of my agency's acquisition, the buyer handed us a welcome gift: a client worth around £2m to us. My whole team refused to work on it.
The account clashed with what the agency actually stood for, and my people would not touch it. So I went back to our brand-new owners, on the first day of a multi-year relationship, and said no thank you. The relationship never really recovered. I lost millions, and I would make the same call again, because the alternative was watching the team's belief in the place die instead.
What that taught me about values has nothing to do with the poster in reception. Corporate values (curiosity, passion, integrity, laminated) are worthless. Personal values, the ones formed by the time you were eleven, are the most stable thing in an agency: the economy shifts, AI arrives, clients churn, and your people's actual values do not move. When they align, you have a core you can build a proposition and an exit thesis around. When they clash with a partner's or an acquirer's, the schism is not a risk. It is a certainty, and it will surface in the worst month of the deal.
I once walked away from a partnership worth ten million to me for exactly that reason, and it was cheap at the price.
So before you sign anything: surface the real values, yours and theirs, and check the overlap. It is a structured exercise, it takes ninety minutes, and I have run it scores of times, occasionally with fireworks. It is part of how I work as a board advisor. felixvelarde.com, or DM me.
Right now, someone in your agency is pasting client data into a personal ChatGPT account. It costs them twenty dollars a month, and it could cost you the client.
This is not a hypothetical. Your staff are already using AI daily, because it makes their work faster and you have not given them a sanctioned way to do it. So client campaign data, customer lists and confidential briefs are moving through personal accounts, outside your controls and in breach of data protection law and most of the contracts you have signed. If a client's lawyers ever go looking, the words wilful and negligent will both be available to them.
The fix is not a ban. Bans drive the behaviour further underground while your competitors compound their advantage. The fix is to get in front of it: corporate accounts for every single member of staff, governance around client data, and explicit encouragement to use the tools ambitiously inside those rules. Yes, at hundreds of dollars per head per month it looks expensive, until you compare it with losing a client for cause.
Done properly, this stops being compliance and becomes capacity. People hand the trivial work to the machines and spend the recovered hours on judgement, taste and client relationships, the things clients actually pay agencies for.
Getting agencies onto the front foot with this is board work I do every week. DM me if you honestly do not know what your exposure is today.
A large share of American agencies are dead men walking, and the tell is white-labelled software.
The US agency market runs on it. By my estimate the substantial majority of American SEO agencies resell the same handful of platforms: white-label the dashboard, add a markup, charge three hundred nail salons and auto shops a grand a month each. For years that was a perfectly good business. It was also, though nobody said it out loud, a commodity wearing an agency's logo.
AI has now called it. A client can optimise their own site with a couple of prompts. They can build a perfectly serviceable website in ten minutes. The undifferentiated agency has nothing left to sell, which is why so many are quietly trying to exit at once, and why buyers will not touch them. An acquirer looking at tens of thousands of agencies running identical platforms sees tens of thousands of businesses with no reason to exist independently, worth nothing beyond their client list.
The agencies that survive this are the ones that own something: a proposition nobody else can claim, a niche they are genuinely best in the world at, and increasingly their own tools, built around their own way of working. Generic platforms produce generic agencies. Generic is not attractive to clients, and it is worth nothing to a buyer.
Repositioning out of the commodity zone is doable, but it is strategy work, not a rebrand. It is what I do with agency boards. DM me if this one stung.
Your agency exists for one reason: to deliver bags of cash to your clients. Everything else is decoration.
In 2007 my CRM agency won Virgin Holidays. A couple of hundred grand to run an email programme. At a certain point we could walk into the room and say: for every pound you put into the top of this machine, twenty-six pounds of revenue comes out of the bottom. The budget went to two million, almost overnight. Nobody 10x's your budget because they like the work. They do it because you have made the return undeniable.
I think most agencies have forgotten this is the job. You specialise through search or creative or CRM or a vertical, fine, but the underlying transaction is always the same: money in, more money out, provably.
It matters more right now than at any point in my thirty years, because AI has put a blunt question under every agency relationship: what do you do that the client cannot now do themselves in ten minutes? Agencies answering that with efficiency (same work, fewer people, lower price) are racing to the bottom of a market that no longer needs them. The defensible answer is value: use the technology to deliver measurably more revenue to clients, prove it, and charge accordingly.
That reframing, from cost to cash machine, changes your pricing, your proposition and eventually your exit multiple. It is the first conversation I have with any agency board I join. felixvelarde.com, or DM me.
The polished management meeting is no longer where agency deals are won. Buyers have learned to look underneath.
The traditional process let a founder control the story: buyer meets the two founders, maybe a board member or two, talks to a couple of hand-picked clients during due diligence, and never sees below the top slice. Deals done that way kept failing, because the thing being bought (the team, the culture, the machine that does the work) was never actually inspected.
So serious buyers changed their approach. When we acquire, we offer the target a full strategy day as a gift: a proper workshop with their best people, building a three-year plan. They get real value from the day whether or not a deal happens. We get to see the quality of the team with the polish off. Secret-shopper enquiries are routine now too, and I would call them entirely legitimate. And your clients will take the diligence calls; big clients are used to them and speak candidly.
Which means the only reliable strategy left is the obvious one: be genuinely good under the hood. Depth of talent beyond the founders. A team that could run the place if the top three vanished. Clients who volunteer that the agency would survive without you.
None of that can be staged in the month before a process starts. It gets built over a couple of years, deliberately. That is the work I do as a board advisor. DM me.
The most expensive sentence in agency M&A: we're thinking of selling next year.
By then the choices have already collapsed. Anything financial in your valuation gets averaged over three years, so this year's improvements barely register. The discount factors (succession, concentration, proposition, systems) each take months to fix and another year to show up in the numbers. A founder who starts thinking about exit twelve months out is not preparing a sale. They are choosing between a markdown and a delay.
The founders who do it well treat it as a time-money equation, decided coldly. Sell now and accept the discount, or spend a year fixing everything, a year letting the numbers prove it, and sell for multiples more. Neither answer is wrong. Not knowing you were making the choice is what costs the money.
My advice to any founder past the first million, even one with no intention of ever selling: build the agency as if it were going to market in three years. Put in the succession team, the management rhythm, the fractional CFO, the systems. If a buyer turns up, you are ready. If one never does, you own a calmer, more profitable agency that runs without you, which was rather the point of starting it.
Buildable, with a system, in about a year. That is the board work I do with a small number of agencies at a time. DM me if you want the honest assessment of where yours stands.
One client can quietly cost you two-thirds of your agency's value. The arithmetic takes thirty seconds.
Say you run a $3m agency producing $600k of profit, and your biggest client is $1m of the revenue. You go to market expecting, at four times, something like $2.4m. The buyer looks at the same numbers and does a different sum. The big client is a risk, so they price you as the $2m agency you would be without it: $400k of profit. And because smaller agencies carry lower multiples, that $400k gets a multiple of two, not four. Your $2.4m just became $800k. A two-thirds haircut, from one line on your client list.
Founders experience this as being low-balled by everyone at once. It is nothing of the kind. It is the same risk calculation, run by every competent buyer, producing the same answer.
And the buyer's caution is justified, because the concentration is worse than the spreadsheet suggests. When a whale leaves, profit does not fall proportionally. You carry the dedicated team for months while you unwind it, so $600k of profit goes to roughly break even, along with most of the morale in the building.
The repair is a deliberate, unglamorous programme: rebalance the portfolio, grow the mid-tier, put an account plan under every client that matters. It takes a year or so, and it is worth more than almost anything else you could do with that year. This is board-level work, and it is what I do. felixvelarde.com, or DM me.
Multiples come from buyers. They were never designed to get you a good price.
Every founder knows the going rate. The formulas have been stable for thirty years: a certain EBITDA earns a certain multiple, and on that basis your agency is worth eight times whatever. The trouble is that the formula is not a promise. It is the buyer's opening anchor, and everything that happens next moves in one direction: downwards, as they discount for every risk they find.
No succession team, discount. One client too big, discount. A proposition nobody can distinguish from the agency next door, discount. Gross margins that wobble, a pipeline that is really the founder's address book, a leadership team that has never made a decision alone: discount, discount, discount. Due diligence exists to find these things, and it will.
In the US market this bites harder than anywhere. A very large generation of agency owners is trying to sell to a small pool of buyers at once. Buyers do not need to negotiate hard; they simply move to the next agency on the list. You have to be clean to sell at all, let alone at the top of the range.
The full repair job takes about a year, and the improved numbers take another year to show up in your accounts. So the window to start is roughly two years before you want out. I sit on agency boards to run exactly that programme. DM me and I will tell you honestly how long yours would take.
Your multiple is not a reward for twenty years of hard work. It is a measure of the buyer's confidence, and nothing else.
Founders tend to treat the multiple as a grade: for the reputation, the awards, the sacrifices, the quality of the work. Buyers do not grade any of that. A buyer is asking one question: how confident am I in this agency's cash flow over the next three to five years, after the founder has gone? The multiple is the number that confidence produces.
Sit with the arithmetic for a moment. A buyer who pays four times EBITDA is on a five-year mission just to break even if your agency does not grow. That is the bet you are asking them to make. Every risk they can see (the client that is 30% of revenue, the founder who does all the selling, the team that has never run the place alone) shortens the odds against them, and they price it in. Sellers call this being low-balled. It is not. It is risk being removed from the calculation, line by line.
The good news is that every one of those discount factors is fixable, and fixing them is worth more than a year of new business wins. You just cannot fix them during due diligence. You fix them in the two or three years before, which is board-level work, and it is the work I do. DM me if you want to know what a buyer would discount in your agency today.
Felix Velarde retweeted
full disclosure I run a b2b lead gen agency, and the "founder cannot work out why it feels like a grind" line is the whole game. 20% growth that just replaces the client you lose isnt growth, its treadmill maintenance dressed up as a target. the piece I'd add is most agencies never separate "new revenue" from "revenue that offsets churn" in their own reporting, so they cant see which one theyre actually buying with their pipeline spend until its too late to fix
A 20% growth target sounds ambitious. It is barely treading water.
Run the numbers on your own agency. Take off inflation. Then take off wage inflation, which in this market runs ahead of your ability to raise prices. Then ask when you last put your rates up by anything close to RPI. In my experience the honest answer, for almost every agency, is not this year, not last year, and not the year before.
Now lose one small client. Not a whale, just an ordinary account that resigns for ordinary reasons. There goes the rest of your 20%. The agency that hit its stretch target ends the year exactly where it started, and the founder cannot work out why it feels like a grind.
This is why I named the programme I designed 2Y3X: two years, three times. Not because tripling is easy, but because anything much less ambitious is not actually growth once the arithmetic has finished with it. Agencies plateau because they optimise themselves, very competently, for staying the size they are. Getting past that takes a transformation programme, a team responsible for running it, and a founder willing to get out of the way.
Plateaued at the same revenue for three years is not stability. To a buyer it reads as decline that has not started yet. DM me if your last three years look flat.
Your senior leadership team was hired to keep things the same. Asking them to triple the agency is a conflict of interest.
Think about how they arrived. You climbed a bit, then brought in a project manager to protect the projects. Climbed again, added an account director to protect the clients. Eventually a COO, to protect the systems. Each hire put a pin in the mountain below you. Safe, sensible, and entirely designed around preserving what already exists. Preservation is their job, and the good ones are excellent at it.
Growth is a different job. When I scale an agency I build what I call a Growth Lab team: four or five superstars drawn from every level of the business, not just the top. They write the three-year vision, then work backwards to what has to happen this year. They run the experiments, break things safely, and only roll changes out once they have proven them and the leadership team has agreed.
The pitch to the SLT is honest: you keep running the business you are brilliant at running, while this team does the risky work that will not touch you until it works. The result is a company that grows without every idea coming from the founder, which is precisely what buyers pay a premium for. The superstars become your succession team, grown from below rather than anointed from above.
I install this as a board advisor. DM me if your growth still depends on you.