When an agency owner asks me whether their numbers are any good, I ask a question back before I open any benchmark table: if you stopped marketing today, how many quarters would pass before you’d notice? The tables below will tell you whether your margin is normal, but that question tells you whether your agency is durable, and durability is what most owners are really asking about when they go looking for industry statistics.
So here are the 2026 numbers, every one tied to a named source you can check, and one frame to read them through: in this industry, normal is not the same as healthy. The average agency is under-margined, referral-fed, and honest enough to admit it, which means matching the average is a strange thing to aim for.
In 2026 the average marketing agency runs a 13% after-tax net margin against a 20% industry target, keeps an agency-of-record client about seven years, and still names referrals as its best lead source, so the benchmark that predicts survival for a small agency isn’t margin, it’s how much of the pipeline rides on referrals.
Quick Take
- The margin story favors staying small: agencies with fewer than 10 people averaged 19% after-tax net margins in 2025 while agencies over 50 people averaged 8% (Promethean Research), so adding headcount is a cost decision long before it’s a profit one.
- The risk story is concentration: 75% of firms call referrals their best lead source and 93% admit their growth engine isn’t strong enough to hit their goals (RSW/US), which is the same problem measured twice.
- My working ratio for agencies under $1M: put roughly 80% of your marketing effort into demand work that produces conversations this quarter and 20% into brand, and shift the mix only as content starts lowering your cost per lead.
What this covers: how many agencies you’re competing with, what the average agency makes, what a healthy margin looks like, where new clients come from, how long they stay, where AI adoption stands, and the three benchmarks I check for agencies under $1M.
How many marketing agencies are there in 2026?
There are 114,014 advertising agency businesses in the United States as of 2026, up 4.4% from 2025, and the count has grown 6.7% per year on average since 2021, according to IBISWorld. On the digital side, Promethean Research counts more than 50,000 digital agencies in the US and Canada and over 179,000 worldwide, and finds the average digital agency is tiny: fewer than 10 full-time employees, with 88% of the industry running under 50 people.
Read those two numbers together and the competitive picture gets uncomfortable. Supply is growing near 7% a year while the buyers’ budgets aren’t, and almost everyone entering looks like everyone already there. If every prospect seems to be comparing you against six lookalike shops, that’s because they are, and the sameness is the expensive part. In a market this crowded, being small is survivable, and being interchangeable isn’t.
How much does the average marketing agency make?
The average digital agency in Promethean Research’s latest survey brought in $4.43 million in revenue and earned a 13% after-tax net margin in 2025, which works out to roughly $575K in profit. Treat that revenue figure carefully, because the survey’s average respondent had 31 employees while most of the industry runs under 10, so the “average agency” in the data is far bigger than the average agency on the street.
Growth tells the sharper story. Agencies averaged 7.5% revenue growth in 2025, large agencies grew 2.7x faster than small ones, and the agencies that narrowed their services grew 13% on average (Promethean’s 2026 State of Digital Services). That last number is the one I’d sit with. Narrowing what you sell reads as giving up revenue, but the data keeps landing on the opposite result, and it matches what I tell owners about niching: it doesn’t mean firing every client outside the niche, it means work outside the niche becomes a deliberate choice instead of a default yes.
What’s a healthy profit margin for a marketing agency?
A healthy agency targets 20% profit on adjusted gross income, per the Agency Management Institute’s 55-25-20 rule: 55% of AGI goes to salaries and benefits, 25% to overhead, and 20% is profit. Parakeeto, an agency finance consultancy, sets the bar higher, calling 25% net the standard benchmark while noting most agencies land at 10-15%, and it backs into that number through delivery margin: 50% or better agency-wide, 60-70% of AGI on each project.
Actuals run well under target. The average after-tax net margin was 13% in 2025 and 14% in 2024, against a long-run average of about 15% since 2015 (Promethean Research). And the spread by size is the part most benchmark posts bury: studios under 10 people averaged 19%, agencies at 10-24 people dropped to 12%, 25-49 came in at 9%, and 50-plus averaged 8%.
Scale, in this industry, works like a tax on margin, and the tax rarely arrives as a decision. Nobody sits down and chooses a 9% margin. A project manager gets hired to absorb one crunch, an office lease follows the third hire, a management layer follows the tenth, a benefits package grows to keep pace with bigger shops, and the cost structure assembles itself out of individually sensible moves until the margin is gone and no single line item took it. The owners who keep the 19% aren’t tougher negotiators, they’re the ones who treated each of those additions as a choice with a price tag. If you want the per-head version of this, revenue per FTE is the metric that catches the drift earliest.
Where do marketing agencies get new clients?
Most new agency business still arrives through referrals, and that’s the finding that explains the rest of the survey. In RSW/US’s 2025 report, 75% of marketing services firms named referrals their best lead source, and 93% said their growth engine isn’t strong enough to reach their goals. Those read like two findings, but they’re one finding measured twice, because a growth engine built on referrals is really goodwill other people spend on your behalf, and goodwill doesn’t take instructions.
Referrals feel safe for real reasons. They close faster and arrive warm, and every one is evidence you do good work, since someone staked their reputation on you to make the introduction. The trap is structural: the only way to get more referrals is to keep doing good work, so you pour everything into delivery and stop marketing for yourself (this is delivery drag doing exactly what it does), and then at some point the referrals slow down and you can’t predict when, because you were never the one generating them. You’re betting that the right person in your network happens to know the right person in theirs, every quarter, forever.
The 62% of firms now packaging services into productized offers (same RSW/US report) suggests the industry knows the sales motion has to change. But a productized offer with no demand behind it is a nicer brochure, and demand is the part the 93% admitted they don’t have. The way out is building acquisition you control next to the referrals, not instead of them.
How long do clients stay with a marketing agency?
Agency-of-record relationships now last about seven years on average, more than double the 3.2-year average reported in 2016, per the 2025 ANA and 4As tenure study. Independent agencies hold relationships longer than holding-company shops (7.3 years versus 5.8), while media agencies churn fastest at 3.7 years (MediaPost’s read of the same study).
One honest caveat before you compare yourself to it: that study population skews toward national advertisers and their AOR rosters, not eight-person shops. What transfers to a small agency is the direction, because relationships compound, and that compounding is exactly why referral-fed growth feels so safe from the inside. Your clients stay, they say nice things, the work keeps arriving, and the pipeline question gets easier to defer another quarter. Long tenure is a real asset, but it answers “will they stay,” and the number that kills small agencies is “who’s next.”
How many agencies are using AI now?
Most agencies are already using AI somewhere in the business. Promethean found 34% of agencies had implemented AI across the business with another 28% actively rolling it out, and RSW/US puts AI use in business development at 77%, with 80% planning to expand it.
At those adoption rates, “we use AI” has already stopped being a differentiator, the way “we’re data-driven” stopped being one a decade ago. When three-quarters of your competitors run the same tools, the tools cancel out, and the compare-me question goes right back to the thing AI doesn’t answer: what problem you own and why a buyer should believe you over the six lookalikes. The efficiency gains are real, but efficiency at producing undifferentiated work just makes the sameness cheaper.
Which benchmarks matter when your agency is under $1M?
An agency under $1M should watch different numbers than the industry tables measure. I coach referral-built agencies under $1M and founders in the $1M-$2M band, and margin is almost never the first number we look at, because at that size the binding constraint is nearly always the pipeline, not the P&L. Three checks, in order:
Referral share of new revenue. Take the last 12 months of new business and mark how much arrived through a referral. Above half, and the 93% from the RSW/US survey is you: the growth engine isn’t strong enough because it isn’t yours. This is the number I’d move before touching the margin conversation.
Working channels. I hold clients to a two-channel rule: at least two marketing channels producing conversations, because relying on one means a shadowban or an ad-account suspension takes your marketing to zero overnight. Referrals count as a channel only once you operate them, with a named partner list and a scheduled ask, instead of waiting for goodwill to strike.
The demand-to-brand mix. An agency under $1M should put roughly 80% of its marketing effort into demand work that produces conversations this quarter and 20% into brand, and shift that mix only as content starts lowering the cost of a lead. Brand compounds and it matters, but it pays off on a timeline a thin pipeline doesn’t have, and most small agencies I meet have the ratio inverted: plenty of posting, no mechanism that turns attention into a booked call.
And when you change something, judge it by exposure, not by time. A new message that only a few hundred of the right people have seen hasn’t been tested yet, and the same message in front of a few thousand of them will tell you plenty. Set the test as a number of the right eyeballs against calls booked, and read the result off data instead of the calendar. If you want the standing version of these three checks, they sit inside a simple KPI dashboard rather than a monthly scramble.
The Monday version of all this: pull up the last 12 months of new clients, mark each one referral or not, and divide. If the referral share is over half, that single number is your benchmark for the next two quarters, and every hour you’d have spent comparing your margin to Promethean’s tables is better spent building the second channel. If you’re doing that math from inside a referral-built agency and don’t love the answer, this is the exact problem we work on in the Dynamic Agency Community.
FAQ
What is the 55:25:20 rule for marketing agencies?
It’s the Agency Management Institute’s allocation target for adjusted gross income: 55% to loaded salaries and benefits, 25% to overhead, and 20% left as profit. Drew McLellan, who runs AMI, calls it the most important metric for running an agency profitably and recommends checking it monthly. Miss the ratio and the fix is usually in the 55, since payroll is where agency costs drift first.
What’s a good profit margin for a small marketing agency?
Under 10 people, the data says you should expect to beat the industry average, not match it. Studios that size averaged 19% after-tax net margins in 2025 per Promethean Research, against a 13% all-agency average, and the 20-25% targets from AMI and Parakeeto are realistic at small scale because the overhead layers that eat larger agencies’ margins don’t exist yet.
Are referrals a bad way to grow an agency?
No, and I’d never tell an owner to turn them off. A referral closes faster than any cold lead and proves the work is good. The problem is exclusivity: when referrals are the only source, you’ve outsourced your pipeline to other people’s memory, and it fails without notice. Keep the referrals, then build one channel you control next to them, and treat the referral network itself as something you operate with a partner list and a scheduled ask.
Do bigger agencies make more profit than small ones?
They make more dollars on worse margins. Promethean’s size bands run 19% net for under-10 studios, 12% at 10-24 people, 9% at 25-49, and 8% at 50-plus, while revenue and growth rate climb with size (large agencies grew 2.7x faster in 2025). Scaling an agency trades margin for absolute profit, which is a fine trade if you chose it and an expensive one if it just happened.