Behind the Agency

"It's not like an artificial intelligence. It's just a bucket of files that you run an AI agent over."

TL;DR: Key Takeaways

• Chris Bolton spent 15 years at Murmur, a Portland agency that grew to 26 people and nearly $4M in one year. Then it shrank to about 15 and had layoffs.

• He left to go solo. He planned to hire an assistant once he got busy. He never did, because he kept finding ways for AI to do that work.

• His big idea: an AI assistant isn't really an intelligence. It's your own files with an agent run over them.

• So he keeps one folder per client, with notes, tasks, contacts and email in it, and launches an agent inside that folder.

• The risk he sees: AI companies want to hold your folders and data for you. Then you pay them forever.

• He runs agents on a schedule too. One sends him a morning SEO "weather report" for every client. Another tracks what he did in each project, so billing is easy.

• Test for today: where does your agency's client history live right now? If the answer is "in six different tools," that's the place to start.

Every week there's a new AI tool for one more job. One writes emails. One takes notes. One tracks keywords. One summarizes calls.

So you sign up. Then you sign up again. And every tool wants its own login, and every one keeps a slice of your client data.

Chris Bolton saw this up close. When he went solo after 15 years at an agency, he hit what he calls "tool and login overload." Every new client added five more logins. Every login pinged his phone for a code.

His answer wasn't another tool. It was a folder.

Start with the free Bottleneck Score →

Chris Bolton is the founder of Diviner, a Portland marketing agency focused on SEO and AI visibility. Before that he spent 15 years at Murmur, the agency he ran with his brother, where he led strategy and later the web team.

He also builds his own software. Crystl is a terminal for running many AI agents at once, and Diviner One is the marketing suite he uses with clients.

Visit diviner.agency →

1. Fifteen years, and a $250K warning sign

Chris and his brother turned a freelance design practice into Murmur, with four people sharing leadership. It grew fast. At one point it had 26 people and did close to $4 million in a year.

But the money came in big projects, and the projects weren't making money. Chris remembers the moment it hit him. Murmur signed a $250,000 website project, and he told the team, "Guys, we've never profited on a website project." Losing money on a job that size is a much bigger loss than on a $20,000 one.

He saved that project by getting the client to sign a $70,000 change order when they paused it. Then he took over the web department and spent about a year and a half trying to fix it. It was still too damaged. So he hired his replacement and left.

2. Off the rollercoaster

Starting over at 49 was scary. He had to work with smaller clients and figure out what to even offer. It took about a year and a half before he could say, "Okay, I've got clients now. It's working. I can pay my bills."

But he knew one thing for sure. He'd ridden the project rollercoaster for 15 years. Lots of work, lots of money, then nothing. So he built the new business on retainers.

"That's why I'm relaxed talking to you right now," he said, "because I can predict how much money's gonna come in next month."

He also named the habit that kept Murmur on the ride. "We were a little bit addicted, honestly, to the big new business." New wins got celebrated. Existing clients got less energy.

Watch this moment (6:39) →

3. The assistant he never hired

Chris thought about starting a software company, or doing consulting. In the end he followed the money into ongoing services, and kept building his own tools on the side.

As he got busier, he figured the first hire would be an assistant. It never happened. "I seem to just keep on finding more efficiencies with AI that just I don't need it."

His first wins were simple. Stop logging into everything, because the AI can reach most tools through their APIs. And stop copying and pasting. He's lost whole days at agencies to moving content from one place to another. Now he just asks the agent to do it.

4. It's just a bucket of files

This is the heart of the episode. Chris explained something about AI that changed how he works. Every message you send starts a fresh AI. It reads everything it's given, answers, and is gone. As he put it, "it is born and it dies."

So the only thing you control is what it reads on the way. That's why he keeps a folder for every client, with notes, task lists, contacts and email. When he launches an agent in that folder, it already knows the client. He can ask what's going on with a client this week, or tell it to send an email with his exact words.

"It's not like an artificial intelligence. It's just a bucket of files that you run an AI agent over."

And here's the warning that came with it. AI companies want to hold those folders for you. "You'll pay them forever, and if they keep raising the price you just get screwed."

I think this is the most useful thing an agency owner can hear about AI right now. The tool isn't the asset. Your context is. Your client history, your voice, the way you work. That only stays an asset if it lives somewhere you own.

Watch this moment (18:28) →

"You don't need to use like 10 different services. You just need one AI to look at your data and help you out when you need it."

5. Own it before someone else does

Chris thinks most companies, even the ones doing AI well, are still running on disconnected systems. Project notes live here, sales notes live there. Nobody has a perfect setup yet.

That's what makes the big vendors' pitch tempting: give us your data and we'll handle it. His view is the opposite. "I think that the real solution is to own it yourself."

He told a story about a client in agriculture who has years of crop photos, each one with notes. Now an AI can read that whole library. When a farmer calls with a problem, he can say, "It's probably this. Send me some pictures."

6. Agents that run on their own

Chris has agents that run on a schedule without him. One checks Search Console, analytics and keywords for every client site and emails him an "SEO weather report" each morning. "This website is sunny. This website is rainy." Then it says why.

Another tracks how much time he's spent in each project. That fixes a problem every agency owner knows. The client asks, "What have you even done for us lately?" Now he can just look and answer.

He has others that check his software for bugs and watch his sites' health. Setting them up is easy, he said. He tells the agent it's a good idea, and it's done.

• "It's not like an artificial intelligence. It's just a bucket of files that you run an AI agent over."

• "I seem to just keep on finding more efficiencies with AI that just I don't need it."

• "Guys, we've never profited on a website project."

• "That's why I'm relaxed talking to you right now, is because I can predict how much money's gonna come in next month."

• "I think that the real solution is to own it yourself."

• Ep. 76: 4 Bids, 5x Apart. He Built It Himself in 3 Months.

• Ep. 21: How a Video Agency Scaled on Retainers Without Losing Its Soul

• Ep. 60: The Collective Model (and Why This Founder Stopped Hiring Full-Time)

Visit → https://diviner.agency

Crystl → https://crystl.dev

Email → [email protected]

Subscribe to the show on YouTube

Subscribe to my weekly newsletter

Pick the path that fits where you are right now. No pitch deck. No pressure. Just clarity.

Start with the free Bottleneck Score →

See how the Program works →

"What if you go from the shame to new hotness launch point, and then it just slowly gets better over time?"

TL;DR: Key Takeaways

• A website starts losing value the day it launches. The business changes and the site doesn't, so three years later the owner is embarrassed and buys a new one. Wayne calls it a reverse hockey stick.

• Wayne sells the opposite line. The site launches, then gets a little better every month, for one flat subscription.

• Scope creep is the product. Clients put any request in an inbox and it gets done, with no quote and no haggling. That relief is what keeps them subscribed.

• He priced it by working back from lifetime value. A $10K site every 3 years is about $30K over 9 years. He launched at $280 a month, which is $30,240 over the same 9 years.

• He raised the price until some prospects said no. That told him where the ceiling was, and he still added about 20 subscribers this year.

• Switching cost does the retention. Hiring a web person was a lot of work, and nobody wants to do it twice.

• Test for today: look at the last project you shipped. Is it worth more to the client now than on launch day, or less? If it's less, that gap is what a subscription could sell.

You finish a project and the client is thrilled. Then it's over. The invoice is paid and the phone goes quiet.

So you go back to hunting. Some months you're buried. Some months there's nothing, and summer and the holidays make it worse. You can't do marketing while you're heads down, and by the time the project ends you're starting from zero again.

Wayne Pelletier lived that cycle after he went out on his own in 2020. He found it, in his words, "incredibly stressful." So he rebuilt the offer around one question: what if the thing he delivered kept getting better after launch, and the client kept paying for that?

He sells websites. But the idea travels to any agency that ships a project and walks away.

Start with the free Bottleneck Score →

Wayne Pelletier is the founder of Resonant Pixel. He's been making websites since 1996 and spent years as a creative director on brands like The Home Depot, Verizon and Spanx. After a stint running a large internal agency at a public company, he went out on his own in January 2020.

Resonant Pixel builds and runs Squarespace websites on a monthly subscription. His clients are small business owners who've had a rough time with web designers before.

Visit resonantpixel.co →

1. Feast and famine was the real problem

In his first year Wayne took all comers. Branding, ads, social creative, web design. Whatever his network asked for, he did. And it worked: six months in, he was tracking to the income goal he'd set with his family.

But project work has a rhythm he couldn't stand. "When the project's done, you're starting from ground 0 all over again." So he went looking for a model that smoothed out the peaks and valleys.

He'd been reading about productized services for years. He named three books that shaped it: The E-Myth, Built to Sell, and The Automatic Customer. The last one is about moving a business from project work to smaller, steady monthly billing.

2. The reverse hockey stick, and why scope creep is welcome

This is the heart of the episode. Wayne's pricing is fixed, and so is the deliverable. But he doesn't fence scope. "You actually welcome it in. You make it part of the deal."

To explain why, he drew a picture. A website launches and the owner goes from shame to celebration. Then nobody updates it. The business changes, the site doesn't, and it slowly decays for about three years. Then comes a new period of shame, a new website, and the cycle starts over.

His fix is to keep the launch spike and flip the line after it. Clients can have as many pages and forms as they want. Every request goes into a Trello inbox, moves to in progress, and lands in done. "The done column just keeps getting longer." There's no billing and no negotiating, so there's nothing stopping the client from asking.

And here's the part I didn't expect. Clients who ask for more publish more. They're active online, the site works harder for them, and they're happier with the engagement. The scope creep most agencies fight is what makes Wayne's clients successful.

Watch this moment (7:53) →

3. A niche starts in the mirror

David asked where this model works and where it doesn't. Wayne's answer started with the owner, not the market. Look at the work you're proudest of and the clients who gave you the most energy. Then pick.

He niched into a type of person, not an industry. His people are small business owners who've been burned by web designers. Many are stuck on WordPress sites that were never set up well. "They become allergic to logging in and touching it."

So he moves them to Squarespace, where the site just runs. He called WordPress "a great tool, but it's a sledgehammer," which is too much when you only need a few thumbtacks. His bigger point was about what buyers actually want. "I don't think anybody cares about the tech stack. They just want the thing to work."

"Hiring you is a massive ton of work. Nobody wants to go through that again, and that itself is its own barrier to exit."

4. The math: working back from $30K

Here's how he set the price. People buy a new website every 3 to 5 years. Call it 3. They start thinking about a new one around month 30 and start budgeting for it around month 20. So Wayne wants to be in front of them at month 18 "and remove the need for another website."

Then he did the lifetime math. His project sites ran about 10 grand each. Over 9 years that's three websites, about $30,000. So he aimed for a $30K lifetime value and backed into a monthly price. He launched at $280 a month, low because he had no subscribers yet. Hold that for 9 years and it comes to $30,240. The client never writes a big check.

He pointed to Nike and Amazon to explain the thinking. Nike will spend $800 to win a customer it knows will spend $3,000 on shoes over a lifetime. Amazon knew what each signup was worth and spent up to that line. The number you can afford to spend comes from knowing what a customer is worth over time.

I want to flag this because it's the move most project shops never make. They price the deliverable. Wayne priced the relationship, and the deliverable fell out of it.

Watch this moment (14:46) →

5. Raising the price until he heard no, and why clients stay

Wayne has raised his price since launch. Earlier this year he moved it "a little too far" and started hearing no. He called that awesome. It told him where the ceiling was, and he's since landed on a number that's working. Net of a couple of losses, he's added about 20 subscribers this year.

Prospects always ask if they're paying for the website. His answer: yes, over time. "You're gonna stay subscribed for a year, you just don't know that yet."

Churn has been rare. The few who left mostly shut their business down. And Wayne thinks the fear of churn, like the fear of niching, isn't very rational. The client had a hard time finding you. They did a ton of work to get you what you needed. Leaving means doing it all again, or admitting to their team they hired the wrong person.

6. Getting himself out of the maintenance

David asked if the model lets him take a vacation. Mostly, yes, though Wayne admits he often feels attached to the business. Two roles make it possible. One person traffics requests and handles account service, and Wayne was blunt about why: "that person needs to not be me, because I'm a mess." A designer with good judgment does the bulk of the site maintenance.

His job is to set the goals and the system, then hold people to it. Like any director, he isn't doing all the deliverables.

7. AI, repair work, and done with you

Clients now show up with long requirements docs written by their AI assistant. Wayne's first move is a call, because those docs miss context and the real fix is usually simpler.

He's also changed the offer. There are now two levels: a base "credibility starter" and a "credibility expansion" level. The second can include e-commerce, a franchise site, or now a small AI app that adds one specific job to the website.

Next on his list is repair work: "Bring me your AI design. I will fix it for you." He said the hard part comes later, when you ask the tool to change what it built. "It's like digging a hole. It just gets deeper and deeper."

And he's honest that AI could erode done-for-you work in his space. So he's planning a move to done with you, helping people use their own tools with systems he already has. One to many instead of one to one. "The process doesn't change, it's just who's doing it and how fast it gets done."

• "Instead of putting barriers and penalizing scope creep, you actually welcome it in. You make it part of the deal."

• "What if you go from the shame to new hotness launch point, and then it just slowly gets better over time?"

• "I need to be in front of them maybe at 18 months and remove the need for another website."

• "Hiring you is a massive ton of work. Nobody wants to go through that again, and that itself is its own barrier to exit."

• [Ep. 21: How a Video Agency Scaled on Retainers Without Losing Its Soul](/podcast/retainers-beat-rollercoasters)

• [Ep. 9: The "5 Ones" Framework That Hit 7 Figures in 11 Months](/podcast/scale-your-agency-without-scaling-headcount-the-5-ones-productization-ip)

• [Ep. 45: Stop Saying Yes to Everything (Before It Kills Your Agency)](/podcast/specialization-pricing-discipline-building-long-term-client-relationships)

Visit → https://resonantpixel.co

YouTube → search "Wayne Pelletier" (his channel is under his name)

LinkedIn → https://www.linkedin.com/in/waynepelletier

Subscribe to the show on YouTube

Subscribe to my weekly newsletter

Pick the path that fits where you are right now. No pitch deck. No pressure. Just clarity.

Start with the free Bottleneck Score →

See how the Program works →

"We had four bids with identical inputs, and what came back was a four to five times spread. That told me the numbers were just opinions."

• One brief went to four agencies. The quotes came back between roughly £600K and several million, a four-to-five-times spread on identical inputs. Nobody could show the data behind their number.

• Every agency said yes to the AI question. None could evidence it. One gave a genuinely strong demonstration, and then sent a contract with fixed scope and a change-request fee on every deviation.

• One bid put 40% of a multi-million-pound budget on management. That was a no before the meeting.

• The bid he privately wanted to win was priced honestly, as a range rather than a number. It lost on a risk nobody had answered: a transient workforce.

• Nobody won. The founder built the whole module to production grade in Cursor in three months, then hired people from one of the losing agencies as augmentation he manages himself.

• Skills are no longer the differentiator. Fit is. Skills can be augmented; understanding the buyer's world cannot.

• Test for today: take your last proposal and ask what evidence sits behind the number. If the answer is experience and instinct, you are selling an opinion.

You put a lot into a proposal. You price it carefully. You lose, and you never really find out why.

And the reason you never find out is that the person who decided is not going to tell you. They have their own reasons, their own board, their own risk they are managing. So you guess. Maybe it was price. Maybe they went cheaper. Maybe it was always going to be the incumbent.

Tatum Bisley is that person. He sits on the other side of the table as a fractional CPO, which means he writes the brief, reads what comes back, and tells the client which way to go. He ran exactly this process for a UK construction company that wanted to turn an internal tool into a product they could sell. It ended with no agency hired and the product shipped anyway: the founder built the module to production grade in three months.

Four agencies. One document. And an answer none of them saw coming.

Take the free Bottleneck Score →

Tatum Bisley is a fractional CPO and the founder of Hoursly. As a fractional CPO he works alongside founders, CEOs and CTOs, usually on nought-to-one products. He takes one-to-three-day seats rather than full-time roles, which puts him on the buying side of agency pitches more often than most people ever are. He describes the job as a bridge between the client and the delivery team.

Hoursly, which he is building now, is a delivery ledger for people who sell delivery — agencies, consultants, fractional operators — that records the effort behind the work, whether that effort was a person, an AI agent, or both. It is in closed beta.

Visit hoursly.app →

Tatum does not normally run a competitive process. He has two or three agencies he trusts, relationships built over years, and he uses them. But this was construction, and construction gets three quotes for everything. It is just what the industry does.

So he did the work properly. He architected what the SaaS product would be, listed the requirements, and made sure every bidder could get into the in-house systems so their estimate would resemble reality. Same document, same access, four agencies, one of them UK and the rest overseas.

That matters, because it removes the usual excuse. Nobody was working from a worse brief than anybody else.

The quotes came back somewhere between £600K for a year of effort and several million for two.

Sit with that. Identical inputs, a four-to-five-times spread. And when Tatum went looking for the reasoning, there wasn't any he could inspect. "No one could actually show me the data that turned that estimate into a number rather than just a feeling."

This is the line in the episode I keep coming back to. Not because agencies are dishonest, but because most pricing genuinely is a feeling — an experienced person's gut, dressed in a spreadsheet. That works fine until a buyer asks to see behind it.

The brief asked how each agency used agentic coding and AI. Every one of them said yes. Not one could demonstrate it in their response. The best honest answers were in-house templates and code reuse, which is real, but it is not what was asked.

Tatum was not looking for a buzzword. He wanted one of three things back: lower cost, faster delivery, or higher quality. Break the triangle somewhere. Show me where.

He is fair about the timing, and this matters. This was about a year ago. There were no agent factories then, no Slack ticket that becomes a bug fix and a production deploy inside an hour. He would expect more today than he did then. What he wanted at the time was simply alignment between what an agency claimed and how it proposed to work.

One vendor did demonstrate it, and did it well. Token usage, AI doing a large share of the QA, the whole picture. Genuinely impressive.

Then the contract arrived. Fixed scope. Heavier upfront requirements. Every deviation routed through a change-request process with its own labour cost attached.

"The commercial model wasn't aligned to the way of working."

That is a sentence worth reading twice if you sell agile delivery on a waterfall paper. The delivery team was telling the truth about how they work. The contract was written by people solving for a different risk. The buyer sees both and concludes the demo was theatre — even when it wasn't.

Here is the part you rarely get to hear.

One came back with a cost breakdown that put roughly 40% of a multi-million-pound proposal on the management team. Instant no. There was barely any detail beneath it, just stage gates and weekly meetings offered as the answer to change control.

One was a Nepalese agency he had used before and liked, but only as an extension of himself, where he fed them the work in pieces. Given the whole thing up front, they could not evidence that they could deliver it, and the client had no frame of reference for that part of the world.

One was Eastern European, and this is the one he wanted. They priced on labour cost plus a thin markup, and they gave a minimum and a maximum instead of a single number. Honest, and harder to sign off. "I would have loved to have gone with them." They lost on a transient workforce nobody had given the client confidence about.

And the last one had the lowest unit costs, clear technical capability, and — crucially — had actually worked in construction. That empathy is what took them to the final round. Then the contract lost it.

No bid. The client accepted none of them.

The founder, who is not a developer but had built the original internal tool in Power Apps, sat down with Cursor and built the entire module at production grade in three months. Then they hired a full-stack developer and a junior. Tatum's work was done and he moved on.

Months later the founder told him something better than any of it. His whole team now comes from one of the agencies that lost — as augmentation he briefs, manages and controls directly, paying the agency rather than employing the people. And he likes it that way.

So the agency did not lose the relationship. It lost the shape of the relationship it wanted, and kept a worse-margin version of it. Meanwhile that agency went from around 700 people to about half that, and moved to full agentic delivery.

Asked what a strong response would have looked like, Tatum's answer is not about proposals at all.

"We're now in a world where skills are less important than fit." Skills can be augmented with AI now, and there is no reason to hide that. So the question becomes how you fit the organisation — whether you can turn business talk into technical delivery, whether you understand the security posture they are worried about, whether you know their user.

His example of that last one is the best argument for niching I have heard in a while. On a construction site the average age is 60s and 70s. Big hands. Often a missing finger or two. If your UI shows that you know that, you have proved more than a capability deck ever will.

One more thing, because it is the cheapest fix in the episode. Faster, cheaper, better is meaningless on its own. Cheaper than what — your last project, or a pint of milk? Better than your in-house team or the last agency you fired? Faster than which go-live date?

Ask first. Speak 20%, listen 80%. Then your number has something to sit against.

• "We had four bids with identical inputs, and what came back was a four to five times spread. That told me the numbers were just opinions."

• "No one could actually show me the data that turned that estimate into a number rather than just a feeling."

• "The commercial model wasn't aligned to the way of working."

• "We're now in a world where skills are less important than fit."

• "Just because you can doesn't mean you should."

• Ep. 68: "We're in the '94 and '95 Era of the Internet. Well, for AI.": 26 Years Building Orases with Nick Damoulakis

• Ep. 47: Why Most Software Projects Fail Before a Single Line of Code

• Ep. 61: Win Better Clients by Going Narrow

Hoursly →

Tatumdale (his fractional CPO practice) →

LinkedIn →

Subscribe to the show on YouTube →

Subscribe to my weekly newsletter →

Pick the path that fits where you are right now. No pitch deck. No pressure. Just clarity.

Start with the free Bottleneck Score →

See how the Program works →

"We're bringing probably 90% first-time clients, higher average order values, higher conversion rates than the overall traffic to our client sites."

• Affiliate traffic is not coupon traffic. For Revit's clients it is first-time buyers in the low 90s percent, with 25 to 50 percent higher order values and conversion rates than the site's overall traffic.

• One in three consumers now use AI tools for product research at least monthly, and the fastest-growing source of citations in AI answers is creators: affiliates, influencers, and independent content. The people writing about a brand are becoming its visibility.

• "We do the same thing, but better" gives a buyer no reason to choose you. "We do it differently, and here is why" does.

• Recruit creators outbound. Most of the affiliates on Revit's programs were sought out on purpose, briefed on a call, and often sent product before they wrote a word.

• Sort channels by shelf life. Blogs and YouTube get found again and again. Instagram and TikTok are for a moment.

• Position around one industry, then expand to its neighbours. Golf first, then pickleball, hiking, fishing.

• Test for today: pull your last 90 days of traffic by channel and ask which sources bring first-time buyers. If you cannot answer, that is the gap Reid built a business in.

You know affiliate exists. You have probably run a program, or been pitched one. And somewhere along the way you filed it under coupon codes, last click, and margin leak. A channel to tolerate, not to build on.

Meanwhile your buyers have changed how they shop. They ask an AI which brands are worth a look before they ever see your site. The answer that AI gives comes from what other people have written about you. Not your case studies. Not your homepage.

Reid Colson saw both halves of that shift from an unusual seat. He spent 25 years in data and analytics and made it to chief data officer. On the side he ran a golf simulator site as an affiliate himself. Then one night on a drive home he asked his daughter, on a whim, whether they should start an agency together. She said yes. Three years later Revit Digital runs affiliate programs for golf and outdoor brands, and Reid has the numbers most founders never see — for most of his clients, affiliate traffic is over 90 percent first-time customers. This episode is about what those numbers say, and why the channel you dismissed may be the one that gets you found.

Take the free Bottleneck Score →

Reid Colson is the founder and CEO of Revit Digital, an affiliate marketing agency for golf and outdoor brands that he runs with his daughter, a brand strategist. Before Revit he spent 25 years in data and analytics, finishing as a chief data officer, and five years in technology consulting. He came to affiliate as a practitioner first: a golf simulator site he built and monetised himself. That combination, data discipline plus a creator's view of the channel, is why this conversation is so specific.

Reid and his daughter had talked for years about doing something together. It stayed abstract until an evening when she was working at a brand agency in Los Angeles and he was driving home, comparing notes. He said, with no plan behind it, "maybe you and I should start an agency together." She said, "Yeah, let's do it."

"Nothing cements that commitment for you like making a commitment to your children," he said. From there the shape of the business came from the assets they already had: her brand strategy training, his data background, and his side project as a golf affiliate.

Reid's clearest signal that affiliate management had a gap came from his inbox. As a golf affiliate he received broadcast messages asking whether he would like to promote somebody's makeup. "There's some management in this space that didn't really have their eye on the ball," he said. He compared it to getting direct mail from a company that already serves you. A data person reads that as a system nobody is watching.

So the founding question was not "how do we do affiliate" but "how could we do it differently." David picks this up: agencies that enter a category saying "we do the same thing, but better" give a buyer no way to choose. Reid credits his daughter's brand work for the standard they hold themselves to. "Differentiating your brand is actually being able to authentically live your promises, not just put something on your website that anybody else could do."

Reid is candid that pipeline is the part of the business he is still learning. His top channel is relationships: introductions from people who can bring credibility before the first call. "I don't think this is really a secret to anybody," he said, but they produce the best discovery conversations and the highest close rate.

Two newer channels are starting to compound. Account executives at Impact, the affiliate platform his clients run on, have begun sending referrals. And content on Revit's own content sites, not the agency site, recently brought in a client who sought them out. Reid is a data person, so he adds the caveat himself: "these aren't scale. This is onesie, twosie kind of stuff."

David's read is that this is exactly how sub-million-dollar agencies grow, and that even the large ones stay relationship-driven. Cold outbound works when your reputation arrives first. Events work because trust gets built in a room. Every channel is really a way to build authority, and authority is what makes an agency fee feel safe next to a software subscription.

This is the section to send to anyone who still thinks of affiliate as last click. Reid starts with the industry number: roughly 9.5 to 10 percent of US e-commerce traffic has an affiliate touch somewhere in the funnel. Then he narrows to his clients.

First, who arrives. For most Revit clients, affiliate traffic is in the low 90s percent first-time customers. "The affiliates are bringing new discovery." Second, what they do. Order values run 25 to 50 percent or more above the site's average, and conversion rates run 25 to 50 percent higher too. Some of that is coupon behaviour. Most of it, Reid argues, is an educated buyer who has already read the comparison.

Watch this moment →

"For most of our clients, first-time customers are in the low 90s. The affiliates are bringing new discovery."

Then Reid connects the channel to the thing every brand is anxious about. At a recent event an eMarketer presentation put two numbers together. One in three consumers now say they use AI tools for product research at least monthly. And the fastest-growing category of sources that AI answers cite is creators: affiliates, influencers, and independent content.

"Having this network that is writing about you, that isn't you writing about yourself, is going to be very important in this world of AI visibility," he said. The affiliate program he sells as a revenue channel has quietly become a discovery channel inside the AI answers his clients' customers are reading.

Watch this moment →

Why do those buyers spend more? Reid offered it as opinion, not data. Affiliate content gets consumed at two points. At the top of the funnel someone searches "best golf clubs for a senior golfer" and narrows the field. Further down they read comparisons of brand X against brand Y on the features they now care about. By the time they click, they feel informed and confident. "They won't have regret afterwards."

The operating model is more deliberate than most brands expect. Revit sorts affiliates into categories, coupon, cashback, sub-affiliate networks, newsletters, blogs and sites, and pays each on terms that match the work involved. Recruitment is mostly outbound. "A majority of the folks that are on the programs that we manage are ones that we've intentionally sought out."

Each creator gets a briefing call. Revit's marketing director spends a large share of her week on those calls, learning what a creator makes well and matching it to a client's content strategy. Sometimes the creator sets the terms: no review without product in hand. Sometimes Revit places the content on the client's site as well as the creator's, and layers user-generated content into the client's paid media on top.

On channels, Reid sorts by shelf life. Websites, blogs, and YouTube keep getting discovered. Instagram and TikTok are short-term promotion. His research effort goes mostly to publishers, bloggers, YouTube, and Instagram.

David closes on the constraint every niche agency hits: there are only so many golf brands before you compete with yourself. Reid's answer is a small firm's answer. "The nice thing about being really small right now is that we've got a long runway." He will not chase everything. Revit will stay golf and golf adjacent: sports, outdoors, active lifestyle. Pickleball, hiking, camping, tennis, fishing. The creators overlap and so do the people who read them.

• "Nothing cements that commitment for you like making a commitment to your children."

• "Differentiating your brand is actually being able to authentically live your promises, not just put something on your website that anybody else could do."

• "Having this network that is writing about you, that isn't you writing about yourself, is going to be very important in this world of AI visibility."

• "I don't want to try to be everything for everybody."

• Ep. 63: 28% of Revenue Came from AI Recommendations

• Ep. 61: Win Better Clients by Going Narrow

• Ep. 73: The 3-Legged Stool: Why Agency Partnerships Fail Without All Three Relationships

Visit → revitdigital.co

LinkedIn → Reid Colson

Subscribe to the show on YouTube →

Subscribe to my weekly newsletter →

Pick the path that fits where you are right now. No pitch deck. No pressure. Just clarity.

Start with the free Bottleneck Score →

See how the Program works →

"I started to replace myself in every aspect of the business."

• The agency scaled when the founder stopped being the product. Jason ran Hennessey Digital as a lifestyle business for three years, then hired a COO, CFO, and CTO, set OKRs, trained two salespeople, and went from $3M to $8M in one year.

• Premium was a decision, not an outcome. His second agency opened with a $15,000 a month floor and turned away the $6,000 lawyers his first agency was built on.

• Offer the tier you don't expect them to take. Three proposals ($20K, $50K, north of $100K) turned one client into a $110,000 a month account.

• Growth mode has a bill. At $12M in revenue his margin was 7 percent and he moved personal money to cover payroll. He calls that by design.

• The exit test is a vacation. If the business keeps growing while you're gone, a buyer can see what they're paying for. He sold at about $22M in revenue with just under $5M of EBITDA, kept the CEO seat, and now runs a $110M group.

• The niche was the first unlock. It solved pipeline before he ever had to market, and the personal brand now drives most of the leads.

• Test for today: list every function you still personally hold. Each one is a line item a buyer discounts.

You're the best salesperson at your agency. You're also the person who approves payroll, reviews the big proposals, and keeps the two or three anchor relationships warm. The business grows when you push and stalls when you travel. You know that's a problem, and you know a buyer would see it too.

Jason Hennessey ran an agency exactly like that for three years. Then he changed one thing, and revenue went from $3M to $8M in a year. He didn't change the offer. He didn't change the market. He replaced himself, function by function, starting with the things he was bad at.

This episode is the clearest account we've had on the show of what "make the business not need you" actually involves, and what it's worth. He sold at about $22M with just under $5M of EBITDA, kept the CEO seat, and has since bought six agencies.

Take the free Bottleneck Score →

Jason Hennessey is the founder and CEO of Hennessey Digital, an SEO and digital marketing agency for personal injury law firms, and the author of Law Firm SEO and Honest SEO. He spent eight years doing SEO for his own projects and for online poker before a 2008 talk to a mastermind of DUI lawyers handed him his first four clients. He sold his first agency to his partner, moved to Los Angeles, and started Hennessey Digital with a rule: $15,000 a month or no. He sold a majority stake at roughly $22M in revenue, stayed on as CEO, and now runs a group at about $110M.

Jason never planned to run an agency. He had been doing SEO for eight years, some of it in online poker, "very competitive niches," when a friend brought him in to speak to a mastermind of about fifty DUI lawyers who didn't compete with each other. He opened by admitting he knew nothing about ranking law firms, then showed them how he ranked number one for "online poker." He left with seven business cards and four clients. "You now own an agency, and you're working with lawyers."

The lawyers referred him to colleagues in other markets, then told him to buy a booth at their conference. Relationships compounded, case studies got better, and a personal brand formed. "Now the personal brand drives most of our leads and revenue for our agency."

David's read: Jason demonstrated expertise in a niche he had never worked in, and the lawyers hired the clarity. Jason took it further. The more crowded a category, the harder it is for a buyer to choose, and "we're the folks who just work with lawyers" makes the choice for them. That is now literal. Ask a large language model for an agency that works with lawyers and Hennessey Digital comes up, because of years of books, podcasts, and content silos on one subject. He titled his book Law Firm SEO for the same reason: he knew he couldn't outrank Amazon, so he let Amazon rank the book for him.

His first agency, Everspark Interactive, sold $3,000 to $6,000 a month packages and never had to market itself. When he and his partner split over appetite for risk, he sold to the partner, moved to LA, and started over with a rule. "No more $3,000 packages. If I'm going to do this, I want to be $15,000 per month." He found one client to pay it, made that the best case study he could, and used it to win the next personal injury firm in the next market.

The trade was deliberate. "Oh, Jason Hennessey, he's really good, but he's really expensive." He was fine with that, and he said no to lawyers with $6,000 budgets on purpose. "Our agency could have been way bigger if we just said yes to everybody. But we said no a lot more than we said yes."

Watch this moment →

"No more $3,000 packages. If I'm going to do this, I want to be $15,000 per month."

David asked whether going premium meant giving up the repeatable, McDonald's-style delivery Jason had described earlier. It didn't. The build is the same whether the firm is in Los Angeles or Chicago: architecture, internal links, schema. What the bigger budget bought was reach, more content, better links, digital PR into national outlets, and every new client inherited the lessons from the last four.

Packaging evolved by accident. Ben Crump's team asked what Jason would do if budget were no object. He wrote three proposals: the standard $20,000 a month, a $50,000 version with far more content and links, and one north of $100,000 that he "just kind of threw out there." They took the third. Four hundred pages of content, before AI, written and edited by humans. "Oh, shit, what did I get myself into?" The strategy worked, and Ben Crump is still one of the most cited law firms in AI answers today. Jason's clients, he says, have big egos and big budgets. If you can deliver signed cases below their target cost, "you've got a blank checkbook."

This is the section for anyone who has wondered what "set the business up for acquisition" actually means. For three years Hennessey Digital was a lifestyle business: car through the company, money taken at will, and, he admits, more personal income than he made later. Then he decided to build an asset instead.

He brought in a COO, a CFO, and a CTO. He built layered management and set OKRs. He trained two salespeople so he wasn't the only revenue source. He replaced himself in every function that drained him, starting with accounting. Revenue went from $3M to $8M in one year and the firm hit number 290 on the Inc. 500. Then $10M, $12M, $20M.

Watch this moment →

It was not all profit. There was a stretch at $12M in revenue where margins were 7 percent and he moved money from his personal account to cover three payroll cycles. He calls that by design: growth mode, reinvested. The test that mattered was simpler. "If I took a vacation, maybe there's a couple fires I've got to put out, but the business is still growing while I'm away."

Jason was 48 and not interested in a beach. He wanted a financial backer who would let him take chips off the table, push some back in, and stay CEO of something bigger. He sold at roughly $22M in revenue with just under $5M of EBITDA. The group has since bought six agencies and sits at about $110M in revenue and $30M of EBITDA, with a larger private equity partner, and maybe a public listing, ahead.

David closed on the thread that ran through the whole conversation. The niche solved pipeline before Jason ever had to market. Solving pipeline is what let him step out of the day to day. And stepping out is what made the business worth buying.

• "No more $3,000 packages. If I'm going to do this, I want to be $15,000 per month."

• "Our agency could have been way bigger if we just said yes to everybody. But we said no a lot more than we said yes."

• "I started to replace myself in every aspect of the business."

• "If I took a vacation, maybe there's a couple fires I've got to put out, but the business is still growing while I'm away."

• Ep. 66: The Partner Buyout, the $4 Billing Error, and "Do More With More"

• Ep. 61: Win Better Clients by Going Narrow

• Ep. 69: The Agency That Grew Revenue Without Adding Clients

Visit → hennessey.com

Book → Law Firm SEO

LinkedIn → Jason Hennessey

Subscribe to the show on YouTube →

Subscribe to my weekly newsletter →

Pick the path that fits where you are right now. No pitch deck. No pressure. Just clarity.

Start with the free Bottleneck Score →

See how the Program works →

"If you just treat yourself as a vendor from day one, trying to switch later to more of a business advisor role, it feels weird. So you wanna do that from day one." — Adam Weil

• White label partnerships don't fail from bad code — they fail from missing relationships

• The 3-legged stool: you need a relationship with leadership, sales, AND the PM at every partner agency; if one leg goes, the whole thing falls over

• Trust and timing are the two variables that control whether a referral happens — staying top of mind is the actual work

• Positioning yourself as a strategic partner has to happen on day one; you can't layer it on later without it feeling forced

• Free value (something genuinely useful, no ask attached) is the most underrated partnership-building tool in the playbook

• Cold email once hit 20% reply rates; now it's too saturated — the shift is from demand gen to brand awareness

• Audit your partner roster: agencies that aren't interested in actively growing don't deserve your full account management investment

You've got a referral agreement in place with a few other agencies. You both signed it, shook hands, and said you'd send work each other's way.

That was eight months ago. Nothing's moved.

The problem isn't bad faith. A referral agreement doesn't create a relationship, it just documents the hope of one. And without a relationship, there's no trust. Without trust at the right moment, there's no timing. And without timing, the project that was perfect for you went somewhere else.

Adam Weil has spent seven years on the other side of this dynamic, as the white label development partner that agencies either learn to rely on or forget about entirely. In this episode, he maps out exactly what separates the partnerships that generate consistent work from the ones that never get off the ground.

Take the free Bottleneck Score →

Adam Weil is a partner at White Rabbit Group, a white label development shop that works exclusively with agencies. He came to the role through an unusual path — running his own web design agency, burning through a dozen development teams across three continents, and eventually becoming a client of White Rabbit before joining as a partner in 2019. That experience of being on the agency side shapes how White Rabbit approaches every aspect of its partner relationships.

Visit whiterabbitgroup.com →

1. From door-to-door sales to a $3M dev shop

Adam's path to running White Rabbit Group started with a failed college clothing company and a late-night obsession with Illustrator. Teaching himself design software in his dorm room — staying up until sunrise — gave him the first hint that he might want to work in tech. But his next move was unexpected: door-to-door alarm system sales, pure commission, no base.

That experience shaped something more durable than a skill set. "You learn to deal with rejection, to persevere, and just go to the next one," Adam said. When he eventually pivoted into web design, the willingness to do outreach — and to be bold about it — came naturally. He started freelancing, built a small shop, and hired employees. What he didn't have was a reliable development team.

2. How White Rabbit accidentally became an agency-first business

Before Adam ever joined White Rabbit Group, he was a client — and a reluctant one. He'd tried outsourcing development to South America, Eastern Europe, India, and local contractors. "I went through a dozen different development teams," he said. "They'd start off well, then I'd get handed off to someone else. Communication would be spotty. No QA. The CMS was a mess."

White Rabbit reached out via cold email and it took Adam a full year to give them a chance. Once he did, he never hired another development team. The model clicked: a stateside technical project manager for timezone communication, paired with co-founder Abilash managing the India team on-site. "Those two together, with high-quality output and some cost savings, allows an agency to upsell. I thought the whole model was brilliant."

When White Rabbit's founder Greg acquired Adam's shop in 2019, Adam joined as a partner and immediately pushed the company to double down on its agency-facing positioning. He knew exactly what it felt like to be on the other side — and he saw how much opportunity the model had for agencies that couldn't build or afford a full dev team.

3. The shift from project managers to account managers

For nearly a decade, White Rabbit ran entirely on technical project managers. It worked well enough — until Adam started asking why the agency relationships that were good weren't becoming great.

"When it comes to expanding that account, growing it, being a trusted advisor for that agency — not just a technical vendor — you need a different role," he said. They discussed hiring account managers as far back as 2022 and shot it down. The margins were good. The PMs were performing. And critically: "You don't know what you're missing until you have someone in that role experimenting."

What changed was internal. Their director of services, Marshall Bellinger, raised his hand and said he wanted to lead it. He had the right instincts — sales-minded, entrepreneurial — and they let him run with it. It's still early: "We've been around almost 10 years and only had account management for about a year and a half. We're still spread way too thin relative to the number of agencies we have."

4. The trust-and-timing principle

Adam describes agency partnerships with a deceptively simple framework: it's all about trust and timing. You can have both parties ready and willing — but if the timing isn't right when the project comes in, someone else gets the call.

"I'll reach out to someone I haven't talked to in quite a while, and a few days later they say, 'That's funny timing — we actually got a project.' That's not a coincidence. It's because I was top of mind and relevant in their mind." The implication is uncomfortable: there were probably projects in between that didn't come their way, simply because Adam wasn't in front of them.

The fix isn't complicated, but it requires a system. Stay in touch. Have something meaningful to say when you do. "Not just 'hey, how's it going?' — actually have something relevant to bring them." He recommends every agency have two to four free pieces of value they give away each year — not as marketing collateral, but as a genuine excuse to show up.

"It's not that there were no projects in the time since I last talked to them. That's not a coincidence. It's because I was top of mind."

5. Free value as outreach infrastructure

White Rabbit is currently building an AI-powered estimator — agencies can upload a brief or RFP, and it produces a detailed estimate with timeline, recommended tech stack, risks, and hour ranges. All historical project data, fed in. And they plan to give it away for free.

"That alone gives you something meaningful and relevant to say," Adam explained. "If you don't have anything else, at least have some free value you can just send them." The concept he's toying with is almost radical in its honesty: an email that says upfront it's automated, explains exactly why they're a fit based on research, passes along the free tool, and ends with zero ask. No follow-up. Just top-of-mind.

"Almost like you're handing out flyers," he said — a callback to his door-knocking days. The principle is the same: volume and goodwill, without pressure.

6. Cold email: what worked, and why it doesn't anymore

During the COVID years, Adam cracked cold email. He'd find agencies on LinkedIn — filtering for case studies but few developers — write genuinely personalized outreach, and send carefully. "I wasn't blasting a million people. I had a 20% reply rate for a solid year. Looking back, that's incredible."

That era is over. "We don't even really do cold email anymore. It's just not effective." The culprit, he thinks, is saturation — when everything went digital in 2020, everyone flooded the channel simultaneously, and inboxes built up resistance fast. AI has made it worse: personalization is now table stakes, but AI-generated personalization is easy to smell, so the bar for what reads as genuine keeps rising.

The shift he's making is conceptual: stop treating cold outreach as demand generation. Treat it as brand awareness. "If you're also posting content regularly or running paid ads, when you do send that cold email, they think, 'I've heard of you guys before.' They're more likely to open it." The outreach works harder when it's not doing all the work.

7. The 3-legged stool: the full partnership process

This is where the conversation gets operational. Adam walks through the exact sequence White Rabbit uses to turn a new agency into a long-term partner — and the framework that keeps it from stalling.

The process starts at the estimate stage. As soon as a new agency comes in with a project, Adam brings in an account manager immediately — not to close, but to introduce. "If the project closes, they're going to continue working with that account manager. I'd rather have them know each other instead of having to hand it off."

From there, the kickoff meeting is used to plant two seeds: make a strong first impression, and get the retro on the calendar. "Before the project is over, you make sure you get that retro on the calendar — because once it's over, it's hard to get it." At the retro, you're asking about their goals, their pipeline, whether this type of project is a one-off or a direction they want to grow in.

If there's interest, you propose a strategy call — within 30 days. And in the window between the retro and that strategy call, you're building what Adam calls the 3-legged stool.

"You have your leadership leg — a CEO, COO, or head of services. Your sales leg — a business development leader who needs to be able to sell what you do. And your PM leg — the project manager you work with day-to-day. You want a relationship with all three."

If any one of those legs goes missing, the stool falls. The sales team might love you but leadership keeps killing the budgets. Or the PM trusts you completely but the salesperson doesn't know how to present your services. "That's when I started to map it out," Adam said. "It was just pattern recognition. Why has this agency only brought us 4 projects this year instead of 10? Because their sales team doesn't know how to talk about us."

After the strategy call, you set up quarterly business reviews. The QBRs plus ongoing one-on-ones with each of the three stakeholders form the maintenance layer. And critically — you qualify who deserves that level of investment. "If they tell you they're not really interested in actively growing that side of the business, put them in a quarterly email check-in and keep it light. You don't want to spend all this time investing into an agency that doesn't want to grow.

• "When you're outsourcing that sort of stuff, you're not just outsourcing code. It's your reputation."

• "You don't know what you're missing until you have someone in that role experimenting and seeing what could come of really good account management."

• "If you just treat yourself as a vendor from day one, trying to switch later to more of a business advisor role, it feels weird. So you wanna do that from day one."

• "I've seen some agencies earn half a million dollars a year just from the projects they brought us — and we're doing 90% of the work or more."

- Ep. 27: The 4-Direction Partner Framework That Turns 3 Deep Relationships Into a Referral Engine

- Ep. 58: How a 4-Person Shopify Agency Builds a Partner Engine That Actually Pays Off

- Ep. 49: Weekly Client Calls, Referral Networks, and 15 Years of Retention

Visit → whiterabbitgroup.com
LinkedIn → Adam Weil on LinkedIn

Subscribe to the show on YouTube →

Subscribe to my weekly newsletter →

Pick the path that fits where you are right now. No pitch deck. No pressure. Just clarity.

Start with the free Bottleneck Score →

See how the Program works →

"Niche your advertising, not your agency."

• LinkedIn's traditional ad formats can't carry a complex strategic argument -- so most agencies default to generic messages that make them look interchangeable

• Thought leader ads change the game by letting you communicate the "wrong way vs. right way" contrast that shows your unique approach

• The most effective message structure: symptom the prospect recognizes, root cause they didn't see coming, consequence if ignored, and your solution as the only logical path

• After the ad, send traffic to a "niche offer page" -- not a homepage -- where you walk through the full message at a deeper level and make the engagement tangible (scope, price, next step)

• Best fit for this approach: agencies that have both a strategic component and an executional deliverable, serving a market category where demand already exists

• Generalist agencies can still use this by niching their advertising to one service or vertical, running specialized thought leader campaigns to each audience without exposing that segmentation to anyone else

• The goal is to narrow down 100 possible solutions in the prospect's mind to one -- the solution you own

You've tried LinkedIn ads. Maybe more than once. You ran the campaign, watched the spend go up, and waited for the pipeline to respond. It didn't.

So you concluded LinkedIn ads don't work for agencies. Or that they only work at enterprise scale. Or that your targeting just wasn't dialed in enough. But what if the problem wasn't the platform -- it was the format?

Most LinkedIn ad formats cap what you can say. You get a headline, a line of copy, and an image. That's enough to say "we get you more leads," but it's not nearly enough to show why your approach is genuinely different. You end up looking like everyone else in your category, competing on vague promises instead of real expertise.

That's the problem Dylan Ciaccio identified -- and it's why he built his work around a newer format that finally gives agencies enough room to make a real argument.

Take the free Bottleneck Score →

Dylan Ciaccio is a solopreneur and founder of Focus the Lens, where he helps agencies build and run LinkedIn thought leader ad campaigns. He specializes in the message creation side -- developing the wrong-way/right-way contrast that makes these campaigns land -- and has worked with clients across deal sizes from four to seven figures.

Visit LinkedIn -- Dylan Ciaccio →

Dylan has run this format across an unusually wide spread of deal sizes, from clients with a $1,000 lifetime value up to one seven-figure engagement. That range is the reason it is worth listening to. An ad format that only works at enterprise budgets is a different claim from one that works at both ends.

1. Why traditional LinkedIn ads fail agencies

Dylan opened by identifying the two things most agencies get wrong with LinkedIn ads: format and message. Traditional ad formats -- static image ads with a single generic line -- physically can't carry a nuanced argument. The result is messaging like "get more leads with LinkedIn ads," which signals generalist, interchangeable, and forgettable.

The format limitation isn't just a creative problem. It's also why LinkedIn has always been more expensive than Meta or Google. The targeting is genuinely superior for B2B, so LinkedIn charges accordingly -- but when the format prevents you from showing why you're worth the premium, the ROI math never works.

"A very generic message gets you seen as a generalist, interchangeable," Dylan said. "And so why buy from that person compared to anyone else in that category?"

2. What thought leader ads actually make possible

The shift came with LinkedIn's thought leader ad format, which allows a post from an individual (rather than a company page) to be promoted directly to a targeted audience. The format is long enough to carry a real argument -- which changes everything for agencies with a genuine point of view.

Dylan distinguished between two types of agency positioning problems. Strategic agencies struggle to show why their approach is a "need to have" rather than a nice to have. Executional agencies struggle to show what's different about their work in a crowded category. Both problems have the same solution: a message that names the wrong approach, shows the pain it causes, and positions the agency's method as the only logical alternative.

The franchise marketing agency example made this concrete. Dylan's first client served franchise brands and had identified what they called "fragmented franchise marketing" -- the problem of each franchisee running their own ads with no coordination. The agency had built proprietary software to unify everything. But they were marketing themselves as "the best franchise marketing agency," listing benefits that any competitor could claim. The thought leader ad reframed everything around the contrast: fragmented versus unified. Symptom, root cause, consequence, solution.

3. The message structure that drives the ads

Dylan walked through the post structure he uses with clients. It's layered, almost like peeling an onion. The ad opens with the symptom -- the pain point the prospect already feels and recognizes. Then it names the root cause they haven't diagnosed yet. Then it shows what happens if they don't address that root cause. Then it introduces the right approach.

"It's valuable, it's empathetic, and it helps them see your approach as the only solution," Dylan said.

This structure does something most advertising doesn't: it rules out the alternatives. Most marketing addresses a pain point and says "we fix that." But the prospect's mind immediately generates a list of other ways to fix it. Dylan's method walks them through that list and eliminates the options one by one, until your solution is the only one left standing.

"You thought you wanted X, but what you really need is Y. Some people might call it damning the demand."

4. Where to send people after the ad

David asked a practical question: what's the destination? Homepage, article, booking page?

Dylan's answer has evolved. His recommendation for most agencies is what he calls a "niche offer page" -- a long-form landing page written in an article style, but still designed to convert. It continues the message from the ad at greater depth, goes further into the root cause and solution, and then makes the engagement tangible: what working together looks like, how it's scoped, what it costs, and what the next step is.

The reason for the last part matters. Buyers can be fully bought into your perspective and still have no idea what hiring you actually involves. Without that clarity, booking a call feels like a big, scary leap. The niche offer page closes that gap -- so by the time someone books, they're not wondering what they're getting into.

5. Who this works best for

Dylan was direct about fit. This approach works best in market categories where demand already exists -- where a meaningful percentage of the target market is actively thinking about the problem, not just the few percent who might happen to be in the market at any moment.

High-demand categories let you differentiate within the category rather than create the category from scratch. More strategic agencies that have to build awareness of a non-obvious problem face a longer timeline and more reconditioning work before the ads pay off.

The sweet spot, Dylan said, is an agency with both a strategic component and an executional deliverable -- something tangible enough to trigger existing demand, but distinctive enough to stand apart on approach. "Very strategic content service" was his example: not a commodity content mill, but an agency where message development and distribution strategy are central to the work.

Deal size, he added, is less of a factor than people assume. He's seen the approach work from $1,000 LTV clients to one client with a seven-figure engagement.

6. Generalist agencies and the segmentation play

David raised a natural objection: doesn't this only work for specialists? Dylan agreed in principle -- it's hard to make a sharp "wrong way / right way" argument when you do everything for everyone.

But he offered an important nuance. Unlike organic content (where your whole audience sees the same message), paid campaigns can be targeted to a specific segment. A generalist agency with three main verticals can run three separate thought leader campaigns, each with a specialized message for that audience. The other audiences never see it. The perception of specialization is created -- and perception, Dylan argued, is what actually matters in a buying decision.

"Advertising is about perception and not reality," he said. "Here you can make sure that a specific audience sees a certain message and the others don't."

He summarized the strategic principle this way: "Niche your advertising, not your agency." If niching your entire agency feels too risky, niche the campaigns instead. You get the conversion benefits of being seen as a specialist without locking the whole business into a single lane.

• "A very generic message gets you seen as a generalist, interchangeable -- and so why buy from that person compared to anyone else in that category?"

  "The format is not very valuable, and it doesn't give you the ability to create a message that you need to communicate your unique position."

  "You thought you wanted X, but what you really need is Y. Some people might call it damning the demand."

  "Niche your advertising, not your agency."

• Ep 71: LinkedIn Paid Ads for Agency Growth (Brian Koffler)

• Ep 48: Best Practices Isn't a Differentiator

• Ep 10: If You Market to Everybody, You Market to No One

Visit → Dylan Ciaccio on LinkedIn

Subscribe to the show on YouTube →

Subscribe to my weekly newsletter →

Pick the path that fits where you are right now. No pitch deck. No pressure. Just clarity.

Start with the free Bottleneck Score →

See how the Program works →

"The messaging and the ICP will determine the budget and the success more than the ads or the posts." — Brian Koffler

• Paid ads work for agencies, but positioning comes first. If you can't clearly articulate what makes you different, no amount of ad spend will fix that.

• Target a tight list. Diggs Marketing targets roughly 14,000 people on LinkedIn. At 20-30 clients, that list will last a long time. Smaller and more defined beats bigger and diluted.

• Spend 70-80% of your budget on top-of-funnel awareness. The goal is repetition and recognition, not immediate conversion.

• Bottom-of-funnel ads (book-a-call CTAs) can be toggled on and off based on capacity. Top-of-funnel never turns off.

• Test content organically before putting paid dollars behind it. If it doesn't resonate without spend, it won't resonate with spend.

• Commit to 6 months minimum. The ad platform needs time to learn, and so do you. A 30-day trial is a way to burn money, not a way to test a strategy.

• Thought leadership ads work best when they don't read like ads. Authentic, first-person, slightly imperfect content consistently outperforms polished promotional content.

You've built a solid agency on referrals. The work is good, clients stick around, and word-of-mouth has carried you this far. But you've hit capacity a few times, the pipeline has gotten thin between engagements, and you know that referrals alone aren't a growth strategy. They're a lifeline.

So you look at paid ads. You run a few LinkedIn campaigns. Maybe you spend $3,000 over 90 days, get a handful of leads that weren't a fit, and shut it down with the conclusion that "paid ads don't work for agencies."

Here's the thing: they might work fine. It's just that they failed for reasons that had nothing to do with the ad platform. Audience too broad. Content too promotional. Positioning too vague. Not enough runway for the algorithm to learn anything useful.

Brian Koffler has run demand gen for B2B SaaS companies for years. When Diggs Marketing was ready to grow beyond referrals, he applied the same playbook to his own agency. Two years later, it's still running. The results are compounding. And the mechanics are something any agency founder can follow.

Take the free Bottleneck Score →

Brian Koffler is the founder and head of Diggs Marketing, a boutique demand generation agency focused exclusively on B2B SaaS. He runs paid LinkedIn campaigns for software companies, and in 2024 began applying that same methodology to grow Diggs itself. He joins the show to break down exactly how he built the playbook, what it costs, and what agency founders get wrong when they try paid ads for the first time.

Visit diggsmarketing.com →

1. Why the transition from client work to self-service demand gen felt natural

For Brian, running ads for Diggs wasn't a leap. It was the obvious next step when referrals had taken the agency as far as they could. "My brain went to, 'How do we get business from people that don't know me personally?'" he said. The answer was the same toolkit his clients use every day.

The main difference between doing demand gen for an agency versus a B2B SaaS company isn't strategy. It's budget. Venture-backed software companies are under pressure to hit growth targets for fundraising or exit. Agencies can grow steadily, which changes the risk profile. But if you're selling to B2B leaders, the buyer psychology is nearly identical.

2. Positioning as the non-negotiable prerequisite

Before a single ad runs, Brian insists on clarity. Not a tagline. Not "we do digital marketing." Something specific enough that the right person reads it and thinks "that's exactly what I need."

He described the failure mode common to smaller agencies: generic language that could apply to any firm on the market. "You need to be remembered for something that is unique. It doesn't mean you have to do something completely unheard of. You just need a unique twist, a unique package, or a unique point of view."

The example that landed: an agency that works exclusively with hospitals, with a specific methodology for getting X result. Stack enough differentiating elements, and the content almost writes itself. Every post has a natural point of view, because you know exactly who you're writing to and what they care about. Positioning also has to hold all the way through. It needs to match the sales call, the onboarding, and the client experience. If any part of the chain breaks, the whole thing fails.

3. Building the target audience: start with your best clients

Diggs targets about 14,000 people on LinkedIn. That number came from analyzing the agency's best clients and pulling out the characteristics they shared: job title, company size, and how long they'd been in their role.

That last variable was a real signal. "People who are 6 months into their role or less are often looking for an agency like Diggs or help." New-in-seat operators are in evaluation mode. They're building or rebuilding. That's a trigger, and it made the list sharper.

Brian pushed back on the instinct to want a bigger audience. "I would personally rather have the same 14,000 people who are perfect fits see our stuff 5 times than have people who are not perfect fits see our stuff." For an agency with 20-30 clients, that list will last for years. The trap is confusing reach with relevance.

"Paid is a great amplification for things that are working. It is not a way to test if things are working." — Brian Koffler

4. The 70/20 funnel split and when to turn things off

Brian's budget allocation philosophy is simple: 70-80% toward top-of-funnel awareness, 20-30% toward warm retargeting and bottom-of-funnel conversion.

The top layer has no conversion ask. No demo request. No sales call CTA. Just content in front of the right people, repeatedly, so that when the time is right they already know who Diggs is. This layer never turns off, even when the agency is at capacity.

The bottom layer, ads with a call to action, can be modulated. When Diggs is full, Brian scales it down or turns it off entirely. He doesn't want five qualified calls in two weeks when the team can't properly service them. But the awareness machine keeps running. "Those are people who are super warm who have now been seeing our stuff and just not having the chance to convert."

The result: even without bottom-of-funnel ads running, roughly 90% of inbound calls still attribute to LinkedIn when Brian asks how they found him. The top-of-funnel work drives organic conversion on its own schedule.

5. Content that works: organic first, paid second

Brian writes consistently on LinkedIn and his team has started contributing too. About a third of his posts are educational, a third are adjacent to his buyers' world, and a third are sarcastic observations about being a VP of marketing. All of it resonates with the same person.

The rule for what gets promoted: if it performs organically, it's a candidate for paid amplification. "Paid is a great amplification for things that are working. It is not a way to test if things are working."

For thought leadership ads specifically, the quality signal is whether it reads like a human wrote it. Posts that open with "I" or "me" tend to outperform. Cell phone photos outperform polished studio shots. The more it looks like an ad, the less it performs. The format comes from a person's profile, not a brand page, and readers can feel the difference.

6. Budget, timeline, and patience

If your ICP is tightly defined and the list is small, Brian's rough minimum is $2,000 per month on one platform with one content play. If the ICP is broader, expect $4,000 to $6,000 to generate enough signal to learn from.

The mindset matters as much as the number. "If you're going to run paid ads for a month and see what happens, that's not a good mindset." Brian committed to a 6-month budget before he started Diggs' own campaigns. Not because he expected results in month 6, but because he knew he needed time to learn, adjust, and let the platform learn alongside him.

Within the first two months, the goal isn't pipeline. It's leading indicators. Which posts got engagement? Which audience segments responded? Did anyone reach out and tell you how they found you? Those data points become the hypotheses that guide the next two months. "You will have some indicators that make us have 4 more educated guesses over the next 2 months of what to do. And then we'll learn 2 other things, and it compounds."

7. The long game: pre-warmed buyers and compounding awareness

Brian described clients who had been engaging with his LinkedIn content for over a year before ever reaching out. They'd like posts, read threads, see who else comments. Then they take a new job and within two weeks they're booking a call because they already know how Diggs thinks.

That's the actual promise of this approach. Not click-to-conversion. Pre-built trust at scale. "People come knowing who you are and your point of view. I literally have people on sales calls all the time say, 'Your post about X spoke to me.'" The sales conversation starts at step 2 or 3 instead of step 1.

It's not a referral, but it's closer to a referral than a cold ad impression. And the paid motion is what gets people into that sphere who wouldn't have found Diggs organically. Once they engage, the algorithm serves them more. The relationship deepens without requiring a single direct outreach.

• "The more it comes off as organic and authentic, the better it will perform."

  "I will never turn off the top-of-funnel awareness piece ever, because that is what fuels future growth."

  "If you expect paid ads to just work after a month or two, that's just not how it works."

  "You want to produce content that speaks to an individual person, and have that person in mind."

• Ep 63: 28% of Revenue Came from AI Recommendations

• Ep 15: Your Best Thinking Is Invisible If You Never Hit Publish

• Ep 31: Why "Always Be Closing" Is Killing Your Agency Pipeline

Visit → diggsmarketing.com

Email → [email protected]

LinkedIn → Brian Koffler on LinkedIn

Subscribe to the show on YouTube →

Subscribe to my weekly newsletter →

Pick the path that fits where you are right now. No pitch deck. No pressure. Just clarity.

Start with the free Bottleneck Score →

See how the Program works →

"We focus 99.9% of our energies on doing the work for our clients. We spend literally zero time on biz dev -- and our revenue has grown quite a bit this year through upselling alone."

• Dev shop owners get stuck on the "agency treadmill" because they stay in the developer role too long. The shift to sales is what breaks the ceiling.

• The best technical founders can become exceptional salespeople -- selling services is really just explaining a credible future state to someone who needs confidence before they sign.

• Giving away architecture and strategy advice for free during the sales process builds more trust than any paid discovery engagement. By the time the numbers come up, resistance disappears.

• The metrics that matter are forward-looking: forecasted resource utilization, projected project margins, and how many days of runway you have before you need another sale.

• Resource utilization is the single most actionable operational number for a dev shop. Know your break-even percentage and track it weekly going forward -- not backward.

• The "bench buffer" tactic: offer slower-growth clients a discounted rate with no fixed timeline, letting you fill gaps with pre-sold work rather than scrambling for new projects.

• AI tools like Claude Code can add 30% productivity on good days -- and 0% on bad ones. The execution still matters. What changes is volume: a 10-person shop can now take on 4-5 projects instead of 2-3.

You started as the best developer in the room. That's how you won clients early on, and how you kept them. But somewhere along the way, being the best developer stopped being the thing your business actually needed from you.

The treadmill kicks in. Big project lands, cash is good. Project wraps, pipeline is thin. Repeat. You're doing great work and staying exactly in the same place -- because the constraint was never the quality of your code. It was the top of the funnel, and you were too buried in delivery to see it.

The other thing nobody warns you about: all those tools you stitched together -- Harvest, HubSpot, QuickBooks, spreadsheets -- they work fine in isolation. The damage happens in the gaps between them. Wrong invoice timing, misaligned expectations, a client who thought you'd be further along by now. Each one a small fire, each one stealing time you don't have.

If you're running a 10-30 person dev shop and that's landing a little close to home, this episode is worth your time.

Take the free Bottleneck Score →

Joe Conway is the CEO of Treya, an operations platform built specifically for dev shops. Before Treya, he grew a software development agency to $24 million in revenue and sold it -- then spent two years not building anything before AI pulled him back in. He's now applying everything he learned the hard way to a product designed to give small agencies the financial visibility that only bigger shops can usually afford.

Visit heytreya.com →

1. The moment worth building something new

After selling his agency and taking two years to decompress -- skydiving, MMA, all of it -- Joe wasn't planning to start another company. Then AI tools started arriving, and he approached them the way he approached everything: build a small project, see what happens.

He started with a pricing calculator for software projects, something he'd always done in spreadsheets. Using Claude to handle the parts he didn't want to code himself, he rediscovered the fun in programming. Then his mind went back to the real operational headaches of running a dev shop: the gaps between Harvest and HubSpot and QuickBooks, the invoicing errors, the financial forecasting that was always a beat too late to be useful.

He started reaching out to other dev shop owners on LinkedIn as he built. "The best barometer of whether you're building something interesting," Joe said, "is if people have suggestions for it." They did. He kept going.

2. The identity shift: from best developer to best salesperson

Joe admits the transition took longer than it should have. For the first few years, he kept his head down on the technical work -- and the business reflected it. Feast-or-famine project cycles, no real control over the pipeline, a ceiling that came down to how many hours were in a year.

The turning point was realizing he wasn't adding unique value as a developer anymore. His team had people who were better than him. The place he could make the most impact was sales -- which turned out to be the hardest professional shift he'd made.

What helped him make the leap was reframing sales as an engineering problem. "How do you turn your sales pipeline into a number you can measure and fiddle with?" Once he could think about it that way, the analysis paralysis started to lift.

3. The counterintuitive sales lesson from an 18-month deal

The single biggest sales education of Joe's career came from a deal that took a year and a half to close. A prospect told him they didn't need him -- they had IBM. He kept in touch anyway. Nine months later, they called back.

What followed was an extended process of diagnosing their real problem (the existing platform was essentially unsalvageable), interviewing five different business units, and building an architecture plan and roadmap -- all before a dollar changed hands. Joe became an internal champion, walking the halls selling the future vision to executives.

By the time numbers came up, nobody pushed back. It was 2x the largest deal they'd closed.

The lesson he took from it changed how his team approached every sales process after that. Paid discovery -- the $15K to $30K "we'll figure out what it costs to build what you want" model -- wasn't getting traction. Giving the strategic work away for free, and letting the trust compound, worked.

"We used to come at it from the approach of let's sell discovery. A lot of people were hesitant to spend this money when they had a limited budget on just telling you how much it costs to build the thing they want built in the first place."

4. What Treya actually solves

The problem Treya addresses isn't dramatic. It's the slow bleed of small things falling through the cracks: a client who didn't expect an invoice, a team member sitting on the bench for two weeks before anyone noticed, a project margin that looked fine in theory and leaked in practice.

Larger agencies hire people to manage this. A 200-person shop has a CFO, a finance team, dedicated ops staff. A 10-person shop has everyone wearing five hats, and administrative overhead is a direct cost with no billing counterpart.

Treya's goal is to give smaller dev shops the same financial visibility without the headcount. Joe is clear about the fit: if you're a 10-30 person shop that's still growing, this is the tool. If you're already running Oracle or Dynamics, you've outgrown it.

5. The numbers that actually matter

When asked what single metric he'd surface for agency owners who aren't tracking it, Joe gave three that form a coherent picture: cash in the bank, accounts receivable, and booked business. Compare that combined figure to your burn rate, and you know how many days you have before you need another sale.

From there, resource utilization is the most actionable operational number. If 8 out of 10 people are on billable work, that's 80% utilization. Every dev shop has a break-even percentage -- maybe it's 67%, maybe it's 70% -- and the goal is to know that number and track it weekly, looking forward.

The key distinction Joe kept coming back to: backward metrics are for accountants and the IRS. Forward metrics are the only ones you can do something about. "You can't resell an hour. If someone on your team isn't working this week, that hour's gone forever."

6. The bench buffer tactic

One tactical idea Joe shared for smoothing over pipeline gaps: offer clients who can't keep pace with your rate growth a discounted retainer with no fixed timeline. They sign an agreement, you commit to getting their work done within the year, and you slot it in whenever you have bench capacity.

It keeps those relationships alive without sacrificing margins on active projects. And it gives you a buffer to pull from during slow stretches rather than scrambling for new business at the worst possible time.

• "The best CEOs just have 5 really good stories they tell every day."

  "The future's the only thing that matters in this business. Going backwards is just -- I don't care."

  "I don't care about your net profit margin last year. I care about your forecasted profit margin on this project, because you can do something about that."

  "A services company -- you can just one day turn the light on and get a project and you're profitable. Within the first week of starting my company, we closed a deal and we were automatically profitable."

• Ep 33 — The Financial Hierarchy Most Agency Founders Skip

• Ep 52 — Founder-Led Sales Is a Growth Ceiling (Here's How to Escape It)

• Ep 47 — Why Most Software Projects Fail Before a Single Line of Code

Visit → heytreya.com

LinkedIn → Joe Conway on LinkedIn

Subscribe to the show on YouTube →

Subscribe to my weekly newsletter →

Pick the path that fits where you are right now. No pitch deck. No pressure. Just clarity.

Start with the free Bottleneck Score →

See how the Program works →

"We focus 99.9% of our energies on doing the work for our clients. We spend literally zero time on biz dev -- and our revenue has grown quite a bit this year through upselling alone."

• Kreativa Group grew revenue significantly in 2026 by upselling existing clients, not by landing new ones -- they signed just one new client all year

• Selling outcomes (projected revenue, ROAS, CAC) instead of services (Google ads management) lets you charge a premium and reduces churn because the client sees you as a strategic partner, not a vendor

• "AI forward" is their positioning: AI handles middle-layer efficiency (sitemapping, content frameworks, meeting summaries), while humans own judgment, taste, and final output

• A tool called Noted AI ingests Zoom transcripts, Slack messages, emails, and reports to surface client health, churn risk, and task lists -- saving an estimated 8+ hours per week

• Thoughtful constrained growth beats aggressive scaling: at ~25-30 clients and approaching $1M gross, they run a tighter, more profitable operation than comparable agencies managing 200+ local clients at $1K/month each

• Not niching by vertical keeps the team engaged and prevents internal boredom -- context switching costs are real, but so is the competitive advantage of working across consumer electronics, property tech, AI SaaS, and luxury watches

• Try this: audit your last 90 days of client conversations and count how many upsell requests came inbound from clients you were already doing great work for -- that number is your growth floor

You've been told to niche down. Pick a vertical. Own a category. And there's real logic to that -- economies of scale, deep expertise, faster delivery.

But maybe you tried it, or you're considering it, and something feels off. You're already doing great work across a handful of different industries. Your clients stay. They refer others. They keep asking for more. And the idea of turning away a $100K client because they're in the "wrong" vertical feels absurd.

Or maybe you're looking at your agency and wondering why revenue feels flat when you haven't actually lost clients. You've been heads-down, doing the work -- but you haven't built a sales machine, and the thought of one exhausts you.

Tommy Chang built Kreativa Group exactly in that space. No cold outreach system. No vertical niche. No relentless biz dev motion. Just obsessive client focus, outcome-oriented positioning, and a deliberate choice to grow thoughtfully rather than at all costs.

The results speak for themselves: three client losses in three years, one of the longest-tenured clients still on the roster from month two, and revenue growth this year driven almost entirely by clients asking for more.

Take the free Bottleneck Score →

Tommy Chang is the co-founder of Kreativa Group, a generalist digital agency he runs alongside creative partner Karina. Together they bring 20+ years each from the marketing and creative industries into a boutique operation of eight people serving mid-market companies and VC-backed startups. What makes Tommy's perspective worth hearing is the deliberate tension he lives in: a co-founder who wants to grow fast, held in productive balance by a partner who champions thoughtful restraint -- and the proof that the slower path is actually working.

Visit kreativagroup.com →

1. How a real estate niche became a generalist agency in six weeks

Kreativa Group launched with a clear premise: real estate agents needed websites and nobody was serving them well. They signed their first client in that niche and felt good about the direction.

Then Tommy and Karina started talking publicly about what they were building on LinkedIn. The DMs flooded in. A former boss of Tommy's, now a CEO at a non-real estate company, reached out about paid media. Prior colleagues and contacts surfaced with completely different needs. Within four to six weeks, the original niche had dissolved and Kreativa became a generalist agency.

What emerged over time wasn't chaos -- it was a natural sorting process. They kept doing work they were good at, cut services that didn't fit (logo packages and brand kits are now an off-menu exception, not a promoted offering), and gradually identified the verticals where they'd built the deepest work: consumer electronics, property tech, luxury watches, and AI SaaS.

2. The "AI forward" operating model

Rather than describing themselves as an AI agency or an anti-AI agency, Kreativa landed on a term Karina coined: AI forward. The phrase captures something specific -- they use AI to improve efficiency and output quality, not to eliminate the human layer.

In practice, that means AI shows up in the middle of workflows, not at the edges. They use Relume to generate sitemaps and wireframe structures before handing off to Figma for final design. They use AI to create content outlines and frameworks, which a human content strategist then edits heavily -- sometimes to the point where the final output barely resembles what Claude produced. They tested AI design generation tools like Cloth Design and found the output repetitive and generic: "It produces the same stuff because it has access to the same free libraries. That day may come, but it's not now."

The human judgment layer isn't a formality. It's the core of what they're selling.

3. Noted AI and the 8-hour-per-week recovery

One tool stood out in Tommy's rundown of their stack: Noted AI, which also happens to be a Kreativa client. The platform ingests all client touchpoints -- Zoom transcripts, Slack threads, emails, Google Drive files, reporting -- and synthesizes them into actionable client overviews.

Before a client call, Tommy no longer needs to spend time catching up on where things stand. Noted surfaces what was discussed last time, what tasks are outstanding, who's at risk of churning, and what the upsell opportunities are. Post-call, it converts the transcript directly into a task list that can be copied into Monday for follow-up.

"That has saved us literally a good eight hours a week, if not more." That's a full workday per person per week -- recovered by removing the coordination overhead that accumulates invisibly across an agency.

"You're not just taking $100K and throwing it into Google ads hoping and praying it does something. There's thinking, planning, and strategy behind all of it."

4. Outcomes over outputs -- and why it lets you charge more

Tommy's pitch to clients doesn't lead with services. It leads with a media plan that projects traffic, conversion rate, order count, revenue, ROAS, and CAC. The deliverable is the outcome model; the execution is downstream from that.

This framing comes from his time as a head of marketing before the agency, where building and defending media plans was his job. He brought that same discipline into Kreativa's client relationships -- positioning himself and his team as something closer to a fractional head of marketing than a vendor executing tasks.

The business impact is concrete: clients who see you as a strategic partner churn less, escalate conflict less, and are willing to pay a premium. "There are 10,000 agencies in the US that do Google ads. And probably 100,000 freelancers. When you focus on outcomes instead of the service, you can charge what you charge -- not $100 a month."

5. Why horizontal beats vertical (for them)

The conventional agency growth advice is to niche by industry. Tommy is aware of the argument and has a considered response to it.

First, the economics at their ICP level don't favor a local-services model. A lawn care PPC agency managing 200 clients at $1K/month can generate similar gross revenue to Kreativa's 25-30 clients averaging $50K/month in Google spend -- but the operational complexity and staff requirements are incomparable. "I can't fathom dealing with 200 clients."

Second, there's a non-compete dynamic at their client size. When you're working with companies that know each other, working for direct competitors creates relationship risk. In the luxury watch space, three clients know each other well enough that they referred Tommy between them -- and gave explicit permission for him to work with all three. That's an exception that required trust, not a scalable vertical play.

Third, and Tommy is candid about this: variety keeps the team engaged. Context switching has real costs, but so does doing the same thing every day. "No one can complain about being bored."

6. Thoughtful constrained growth

Kreativa is approaching $1M gross revenue. They could push harder -- Tommy acknowledges that more biz dev energy would probably get them to $2-3M. But that's not the current choice.

Karina is the primary driver of what she calls thoughtful growth. The premise is that the agency exists to do great work for clients, and scaling beyond what the team can manage well undermines the reason it was built in the first place. That commitment to quality is also their main retention engine: in three years, they've lost three clients total. One didn't pay. One was let go after becoming abusive. One left for a cheaper freelancer.

Their largest client signed in month two and is still with them -- and the account has grown. This year's revenue growth came almost entirely from existing clients expanding their scope. "A lot of times we're not even the ones doing the upsell. They're asking."

Tommy holds a counterpoint internally: he wants to grow, and thinks there's more to capture. But he's also seen what the alternative looks like. Right now, the bet is on depth over breadth, retention over acquisition, and sustainable pace over maximum scale.

• "We use AI to supplement what we do -- not in such a way where I let AI do all the work and sit here and collect the check."

• "Our deliverables are like, here's the media plan. But the outcome is: here's how much traffic we're going to drive, the conversion rate, your projected order count, your revenue, ROAS, CAC."

• "We've grown our revenue by quite a bit this year, and we've landed just one new client. It was all through upselling existing clients."

• "You want to do great work for a smaller subset of clients -- not have a massive company with 200 clients making the same amount of money at the end of the day."

• Ep 56: Stop Selling Services and Start Solving Million-Dollar Problems

• Ep 46: How an 18-Year CMS Agency Keeps Clients for 5+ Years (Without a Sales Team)

• Ep 61: Win Better Clients by Going Narrow

Visit → kreativagroup.com

LinkedIn → Tommy Chang on LinkedIn

Email → tommy@kreativagroup.com

Subscribe to the show on YouTube

Subscribe to my weekly newsletter

Pick the path that fits where you are right now. No pitch deck. No pressure. Just clarity.

Start with the free Bottleneck Score →

See how the Program works →