Written by Donnie Strompf
Most agencies grow the same way. Win an account, hire a body. Win three more, hire a manager for the bodies. By the time you are thirty people, half the company exists to coordinate the other half, and the founder who used to do the work now approves slide decks about the work.
I went a different way, and I want to be honest that it was not a strategy at first. It was a constraint.
I started Good At Marketing in 2017 out of Boynton Beach, after more than a decade doing SEO for other people. I could not afford to hire my way out of problems, so I had to build my way out of them. Nine years later we run about 30 active retainers across home services, e-commerce, B2B, health, finance, marine, and hospitality, and the core team is still deliberately small. My brother Roni runs content. Jay, my technical partner going back to my first agency job, runs development.
That is the whole company. It is not lean because lean is fashionable. It is lean because the system does the coordinating that headcount usually does.
The bottleneck is never the work
Here is the thing nobody tells you when you start an agency. The work is rarely the constraint. The constraint is that the knowledge of how the work gets done lives inside one person’s head.
When a client asks a question, the answer exists, but it exists in whoever built the account. If that person is on a call, at lunch, or on vacation, the whole thing stalls. So you hire someone to cover them. Now you have two people who each hold half a picture, and a third problem, which is keeping them in sync.
That is how agencies end up with a project manager for every three producers. You are not buying capacity. You are buying translation.
Once I understood that, the fix got obvious. Stop hiring translators. Start writing the process down until it no longer needs one.
Build the system before you take the volume
The most expensive mistake I see founders make is winning business they have no infrastructure to deliver. It feels like growth. It is actually a debt you pay back in fires.
We do it in the opposite order. Before we scale anything, we build the repeatable version of it. What does a good month actually look like at the content level? What does a real editorial link look like versus one we would be embarrassed by? What gets checked before anything reaches a client? Write it down, make it repeatable, then let the volume in.
I refused off-the-shelf tools for years and built our own stack instead, and I got some funny looks for it. But at scale the differentiator was never the SaaS subscription everyone else also pays for. It is the workflow underneath it. Everybody can buy the same tool. Nobody can buy your process.
AI is a multiplier, not a shortcut
I will say the quiet part. One skilled operator with the right AI-assisted workflow can out-deliver a twenty-person traditional agency on quality. I believe that, and I have run the experiment on my own P&L rather than on a whiteboard.
But the caveat is the entire ballgame. AI only multiplies what already exists. If your process is a mess, AI gives you a faster mess with better grammar. Hand it to a team with no standard and you get output that is polished, generic, and completely interchangeable with your competitor’s. Clients notice that before you do.
What actually changed for us was not that AI writes. It is that the reasoning got captured in the process instead of living in somebody’s memory. Handoffs stopped requiring a meeting. Context now travels with the work. That is what raised the floor on consistency across thirty accounts, and consistency is the thing clients are actually paying for.
The shift for me personally was going from producing to directing. I spend more time setting the standard now and less time executing it, and quality did not drop, because the standard is written down rather than merely understood.
The part that actually compounds
The scale story people want to hear is about output. The one that matters is about retention.
Our very first client, from 2017, is still with us nine years later. The first account of my career, my uncle’s steel building company, has stayed roughly twenty years. Agencies churn clients constantly. That is the real industry norm, and most people quietly accept it as the cost of doing business.
Retention is the cheapest growth there is. A client who stays is revenue you do not have to sell again, an account your team already understands, and a case study that gets better every year instead of resetting to zero. When you keep clients, you can afford to stay small. When you stay small, you can afford to do the work well, which is why they stay. That loop is the entire business.
Chasing headcount breaks the loop. You hire to service growth, the service gets thinner, clients leave, and you sell harder to replace them. I have watched a lot of agencies run that lap and call it scale.
What I would tell a founder starting today
Pick one process where consistency matters more than creativity. Write it down until someone else could run it without asking you a single question. Then build the AI-assisted version of that one process and measure the difference honestly.
Do not start with the tool. Start with the map. Most people get frustrated early because they drop AI into the wrong stage of a process they never bothered to define, and then decide the technology is overhyped.
And before you take on the next ten clients, ask whether you could deliver them at the same standard with the team you have right now. If the answer is no, you do not have a growth opportunity. You have a fire scheduled for next quarter.
Author Bio:
Donnie Strompf is the founder of Good At Marketing, a Google Partner agency in Boynton Beach, Florida.
