Picture two agencies. Both cleared $600,000 in adjusted EBITDA last year. Similar market, similar headcount, roughly comparable client lists.
One sells for $1.8 million. The other sells for $3 million.
Nothing about that outcome came from the second owner being a better negotiator. By the time a buyer names a number, the multiple has already been set, and it was set by things the owner did or did not build over the preceding two or three years.
The multiple is the part of your valuation you control the most and think about the least. Earnings you already track every month. The multiple tends to feel like weather.
What a multiple actually measures
A multiple is a risk score wearing a math costume.
When a buyer pays 3× earnings, they are saying: I believe this cash flow continues, but not confidently enough to pay for more than three years of it up front. When a buyer pays 6×, they are saying something quite different: I believe this cash flow continues without the current owner, without any single client, and without me having to fix anything in the first year.
Every turn of multiple is a statement about confidence. So the work of raising one is removing the reasons a buyer hesitates, one at a time. Three of those reasons dominate.
Lever one: can anyone predict next year
Project revenue is the default in real estate media, and it is the single biggest anchor holding this industry at the bottom of its brackets.
A project-based agency starts every January at zero. Last year's $2 million is history. The clients are real and the relationships are real, but none of it is contracted, so a buyer looking at that is purchasing a track record and hoping it holds.
Retainer revenue is a different asset. Somebody signed something with a date on it, and it survives a bad quarter. That is why retainer-based agencies command the high end of their range and why most agencies in this industry, almost entirely project-based, open every valuation conversation at the floor. In the sell-side audit this summer, recurring revenue was the weakest of the twelve areas scored, which the advisor called "a very common theme within the industry"; the full audit is in What Your Agency Is Actually Worth. At $600,000 of earnings, the floor-to-ceiling band is worth $1.2 million. Predictability is the largest single claim on it.
Predictability shows up in smaller forms too. Clients who pay up front instead of 45 days late. Three years of steady revenue instead of one strong year propped up between two soft ones. The mirror image pulls a multiple down: churn you cannot explain, cash flow that swings hard by season, a profit trend heading the wrong way. Concentration is the one with a number on it. Past 50% from a single client or brokerage, the model takes a flat 15% off the multiple before it does anything else, because that is what we have watched buyers do.
Lever two: does the business need you specifically
The second lever is the one owners find hardest to sit with, because the honest answer is usually yes.
If the founder's relationships bring in the work, the founder's eye is the last check before anything ships, and nothing operational moves without the founder in the room, then what a buyer is looking at is a job with transition risk attached. They price that risk straight into the multiple, because they are reading, accurately, what happens the day you stop showing up.
Over time, when I went to sell my business, there was a lot of interviews. We had to sit down with my staff. You want to know why? Because [the buyer] asked me questions and I didn't know [the] answer because my team had effectively been the one who had created it. For three years, it had been like that.
That is what the top of the range looks like from the inside: a founder who cannot answer the buyer's operational questions because the team built the operation. It took him from year three, when he started the exit, to year five, when he was out, and the sale came at seven and a half years.
What raises this score is unglamorous and entirely buildable. Workflows written down well enough that a new hire can follow them. A sales function that is a person with a quota, not the founder squeezing calls in between shoots. Enough redundancy that one resignation is an inconvenience instead of a crisis. The audit put a number on the last one: the business ran without the owner, but editing and quality control ran through one person, and the advisor's read was that a buyer "might say well what if he leaves?" The fix he proposed was a success fee, paid by the buyer, if that person stays 24 months and revenue holds.
Staff structure is priced the same way. One owner in the group is moving his top shooter from contractor to employee this month for exactly this reason: "as contractors, you know, there's no guarantee that people are going to stick around and that they come with the sale." The coach's answer was that every one of his own employees came with his sale because all of them, including the overseas staff, were on W-2 payroll with timesheets. High turnover or poor morale reverses the effect, because a buyer reads both as evidence that the real system is the founder holding things together by force of will.
This is a longer project than the predictability work, and it is worth more. A business that does not require you daily is also a better business to own even if you never sell it.
Lever three: what kind of company do you appear to be
The third lever is a category question, and it has the largest arithmetic attached.
A media and services company trades in one range. A business running on proprietary technology and data, where clients sit on recurring plans and the operator can see performance across markets in real time, trades in another. Those two ranges barely overlap.
You want to add software dashboards, essentially AI or software-ish kind of things, so you can build recurring revenue inside of it. You can switch the valuation from a service business at 2 to 5× to a tech-enabled model at 8 to 12×.
Two qualifiers belong next to that quote. The realistic version for a media business is the 6× to 8× band the full thesis maps for a data-enabled managed service; 8× to 12× is software-company territory, and a dashboard alone does not make you a software company. And the category only holds when the infrastructure underneath it is real. What earns it is a system clients actually open, data the business actually runs on, and revenue that genuinely recurs. A dashboard built for the diligence room is transparent to anyone who has bought a company before.
The reverse holds too. Obsolete technology and manual process do not just cost margin, they cap the category. An agency doing everything by hand in 2026 gets priced as a labor business, because that is what it is.
Where deals actually land
It is worth being honest about the distance between the top of a range and the middle of it.
A New York creative design agency with $2.15 million in adjusted EBITDA sold recently for $9.85 million, a 4.6× multiple. A California social media agency with $2.3 million in EBITDA sold for $13.5 million, or 5.9×. Both sat in brackets whose theoretical top runs well above where they landed.
That is normal. The top of any range belongs to the cleanest businesses in the strongest strategic positions. Most agencies, without deliberate preparation, sell near the floor of whatever bracket they are in.
You're probably going to get five to seven. That's realistic.
Five to seven is the realistic band for a prepared business at the scale where institutional buyers engage, which starts around $3 million in combined EBITDA, with ten reserved for the very top. I have seen the top of that band once, on a forward basis, and I turned it down.
At our peak, Aerial Canvas was offered like I believe 6.7 million when we're doing 7 million a year for our agency, which was like a 5, 6x multiple, you know, on our forward projected cash flows and profit. Even though [we] weren't super profitable at the 7 million because [we] were investing. And I turned that down because I thought I'd turn Aerial Canvas to a 10, $20 million company on my own.
Two years later, at $4.3 million of revenue and coming through layoffs, the offer on the same company was $1 million. The multiple had moved with the risk, and the risk had moved with the numbers. The same spread shows up at the unprepared end of the market: larger agencies in this space have bought businesses at 3× that could plausibly command 5× or 6× organized into a larger, cleaner entity. Scale and preparation are what keep that difference with the people who built the companies.
Start from where you actually are
None of these three levers moves in a quarter. Predictability takes a year of deliberate retainer building, and independence takes longer than that. Category is slowest of all, because the systems have to work for real clients before anybody will pay you for them. That is the argument for starting now instead of when an offer shows up. By then the multiple is already set, and the only variable left is whether you accept it.
The place to begin is an honest read on your adjusted EBITDA, your bracket, and which of the three levers is costing you the most today. You can value your agency against those inputs in a few minutes. For why scale itself has become one of the levers, the full A27M thesis makes that case at length.
Nobody hands you a multiple. It is the sum of decisions you are already making, priced later by someone else.
