The Independent Agency Business Model: What 70+ Founders Reveal About Why Agencies Stall and Scale

One founder I came across had built his dev shop to roughly twenty people on referrals alone. Great work. Happy clients. A pipeline that filled itself for years without a single cold email.

Then it stopped filling itself.

The referrals kept coming, but they came in clusters or not at all. He couldn't forecast them. He couldn't hire against them. He couldn't choose who they came from. The machine that built the agency had quietly become the thing capping it, and he had no idea why.

His story wasn't unusual. It was the most common story in the entire set.

Over the past two years, the Behind the Agency podcast has recorded long, unguarded conversations with more than 70 independent agency founders. Dev shops. Design studios. Marketing firms. We went back through all 70+ interviews looking for one thing: not what these founders sell, but how the agency business model actually works. How it makes money. Where it breaks. And what the few who broke through it did differently.

The patterns were remarkably consistent. And almost none of them are about pricing.

An agency business model is a structural framework where a firm assembles specialized experts to deliver outcomes clients cannot easily build in-house. While frequently reduced to billing mechanics like retainers or hourly rates, the core model dictates how an agency predictably acquires clients, manages utilization, and generates profit over its lifecycle.

How we did this. We analyzed 70+ in-depth interviews with independent agency founders (2024 to 2026), mostly dev and software shops and marketing and creative agencies, roughly $0.5M to $5M in revenue, founder-led. This is qualitative pattern analysis, not a representative census of every agency. The finding is the convergence. When founders who have never met independently name the same revenue ceiling or the same trap, that agreement is the signal. Stats marked 🟢 are patterns we surfaced from our interviews. Stats marked 🔵 are external benchmarks our founders cited. Every quotation traces to a named founder and episode.

What Is the Agency Business Model? (Definition & Core Types)

That second sentence in the definition above is where almost every guide to this topic stops short.

Ask the internet what the agency business model is, and you'll get a list of billing options. Retainer. Project. Hourly. Performance. Hybrid. Useful to know, and worth getting right. But that's the invoice format, not the model. The model is the engine underneath it: how clients find you, how predictable the revenue is, and how the whole thing behaves as you grow.

The canonical pricing structures look like this:

Pricing is the surface. The model underneath is a position on this map, and AI keeps pushing every agency up and to the right, away from the hourly corner.

Pick any one and you still haven't answered the question that actually keeps founders up at night. The work is good, and the agency still feels stuck. That gap is what this research set out to explain.

Agency model vs. consultancy model

One distinction matters before we go further. An agency model is execution-heavy. You're paid to do the work, which means your revenue is capacity-constrained: more output requires more people. A pure consultancy model is strategy-heavy. You're paid for judgment, which decouples revenue from headcount but caps how much you can deliver without becoming the bottleneck yourself.

Most independent agencies live in the messy middle. They sell strategy but monetize execution. Understanding which side of that line your revenue actually comes from is the first step to fixing a model that's stalling. The five traps below are what stalling looks like in practice.

Why Agency Growth Stalls at the $2M–$5M Revenue Ceiling

The single most-cited growth wall in all 70+ interviews was the same one. Referral-driven, founder-led growth runs out of road somewhere between $2M and $5M in revenue, or around twenty people. 🟢

Founders who had never met named the same number from different directions.

At Solwey Consulting, Andrew Drach put the referral ceiling at around $2M, and was direct that scaling toward $10M requires demand generation you can control, not more handshakes. At Warschawski, Sam Tomlinson described agencies hitting a wall around twenty-five clients sourced from referral networks, after which the network is simply tapped out. Tony Wilson, a fractional CFO at Accquip, and Ryan Watson at Upsourced Accounting both put the structural break at $4M to $5M. Paul Wilson of Massive Growth Partners located the same bottleneck at $3M to $5M, the point where the founder becomes the constraint on every deal.

Jason Swenk, who sold his own agency and now coaches hundreds, framed the diagnosis bluntly. Agencies stuck under $1M to $2M usually have a foundation and founder-bottleneck problem, not a tactics problem. As he put it, "There's no such thing as a bad agency client. There's only a bad prospect or a bad process."

Seven founders, no shared notes, the same wall between $2M and $5M. Referrals give you trust, not control.

Here's the mechanism. Referrals give you trust, but they don't give you control. You can't forecast a pipeline you don't own. You can't pick your ideal client from a stream of whoever happens to get sent your way. And you can't hire ahead of demand you can't see coming. The agency that grew on referrals hits the ceiling precisely because referrals worked so well that no one ever built anything else.

Breaking the ceiling isn't about adding a channel. It's about building one source of pipeline you actually control. That's the through-line connecting every trap that follows.

Related reading: Why referral-only growth caps your agency and how to scale a dev agency beyond referrals.

Managing the Feast-or-Famine Cycle in Agency Revenue

The lumpy-revenue problem feels like a personal failing. It isn't. It's a structural property of project-based work, and the founders who treated it as such planned around it. The ones who treated it as a sales-effort problem stayed stuck chasing the next project.

The clearest evidence is how little agencies invest in their own growth. Across the interviews, founders converged on a stark benchmark. Tony Wilson said it plainly: "Most agencies are spending 3% to 8% on sales and marketing... best-in-class spend 15% to 25%." 🔵 JP Holecka of Power Shifter, after years of underfunding it himself, landed on 7% to 12% of gross revenue as the real number. Sam Shepler of StorySnap put the floor at 10% or more of revenue once your margins can support it.

The gap is roughly 3x. Most agencies spend 3 to 8 percent of revenue on growth. The ones that scale spend 15 to 25. Source: Tony Wilson, Wilson Talbot.

Read those numbers together and the famine stops looking like bad luck. It looks like underfunded growth. An agency spending 4% of revenue on getting clients, then acting surprised when the pipeline runs dry, has diagnosed the wrong problem. The fix isn't working the existing channel harder. It's funding a real one.

The second half of the feast-or-famine fix is structural: recurring revenue. Tony Wilson and Ryan Watson independently landed on the same threshold. To scale past $4M to $5M, retainable revenue from existing clients has to clear 50% of the total. 🟢 Below that line, you rebuild your entire revenue base every single year, and no amount of hustle makes that feel stable.

Predictability is a marketing and retention problem, not a billing trick. As Sam Shepler put it, "If you want predictability, solve that with marketing. Not your billing mechanism."

Related reading: why more business development won't fix your feast-or-famine problem.

Why Cold Outbound Marketing Fails for Independent Agencies

Here's a finding that contradicts most growth advice you'll read. Not one of the 70+ founders credited cold email or cold calling as their primary growth engine. More than a dozen explicitly said it no longer works for them at all. 🟢

The language was vivid. Adam Kurzawa, who leads business development for the 250-plus-person agency Infinum, described mass cold email as spam with a CRM attached, not outbound at all. TJ Pitre of Southleft has taken cold outreach off the table entirely, and he doesn't soften the position. As he said, "If you're going into the conversation trying to sell, most people are gonna sniff that out." Hunter Jensen at Barefoot Solutions watched cold outbound stop working after 2020 and replaced it with in-person speaking to CEO groups that now produces several qualified leads per event.

What replaced cold outreach was relationship-led pipeline. Warren Wilansky at Plank sends what he calls fan emails, genuine no-pitch notes to people whose work he admires. One of them turned into a client relationship that lasted a decade. Across the set, the same pattern repeated: value first, relationships before pitches, owned channels over rented ones.

Owning a channel isn't enough on its own. You have to convert it. Logan Lyles of Demand Shift calls the spot where most owned pipeline leaks the conversion cliff. In his webinar work, two gaps do the damage: roughly 65% of registrants never attend, and fewer than 1% of registrants ever book a call.

Most teams pour effort into fixing attendance. The leverage is in fixing conversion. One change, turning a dead-end thank-you page into a qualifying survey that routes straight to a booking, took signup-to-call conversion from 1-2% to roughly 10%. As Lyles described one client, "We went from booking zero calls off webinars to nearly two dozen. We just moved where the CTA showed up."

Owned pipeline doesn't leak at attendance. It leaks at the second cliff, where a dead-end thank-you page drops booked calls under 1 percent. Source: Logan Lyles, Demand Shift.

The reason this matters for the business model is simple. Cold outbound is a tactic you rent. Relationships, a podcast, a community, a body of published expertise: those are assets you own. Agencies that escaped the referral ceiling did it by building owned pipeline, not by buying lists.

How AI Tools Are Disrupting Traditional Agency Pricing Models

The pricing model most agencies run on is breaking, and AI is the thing breaking it.

The mechanism is straightforward. AI is collapsing the time it takes to do the work agencies bill for. Rishi Khanna, CEO of the 25-year-old agency ISHIR, watched tasks that used to take 40 hours compress toward 10. When the hours fall but the value holds, hourly billing quietly punishes you for getting better. As Khanna put it, "Time-based pricing made sense when time was the bottleneck. AI changed that equation."

The agencies pulling ahead already moved off hours. Michael LaVista at Caxy stopped selling apps and started pricing the business outcomes underneath them. His default deal size went from $100K to $500K, and the company tripled in four years. His framing for why most agencies stay cheap: "As soon as you start talking about what you do, the sale is starting to be over." And on the shift that changed everything: "I spent 10 years offering services, and no one cared. When I started solving $10M problems, things changed."

The payoff shows up in the sales motion too. Travis McAshan at Glide Design replaced long proposals with a one-page, three-option format. In his words, "We cut proposal time by 90% and tripled our close rate. One page, three options."

The three-step pricing shift

Moving off the hourly trap without torching current revenue is a sequence, not a switch.

First, package one repeatable engagement as a fixed-scope, fixed-price offer while keeping your existing hourly work running. Second, attach the price to the outcome the client cares about, not the hours it takes you, and let your delivery efficiency become margin instead of a discount. Third, once the productized offer is converting, make it the default and let the hourly work age out. You change the model underneath your best clients without ever asking them to absorb the disruption.

Related reading: the AI efficiency penalty on agency pricing and how to price when you stop billing hourly.

The Scale vs. Stay Small Dilemma for Agency Founders

Not every founder in the set wanted to scale. A striking number chose, on purpose, not to.

They capped headcount deliberately, often at fifteen to twenty people, and optimized for profit per person rather than people per office. Nick Wilkinson runs Steamclock that way and sells the same way, telling leads they don't need a mobile app when they don't. His rule for the sales conversation: "Sell like doctors, not car salesmen."

Greg Mischio keeps Winbound under twenty on purpose. Jackie Sinex has kept Webii under twenty by design. Karla Santi grows Blend Interactive by one to three hires a year, deliberately, and keeps 75% of revenue coming from retention so a slow quarter never becomes a crisis.

The opposite move is where the danger lives. Founders who scaled on forecasts instead of contracts nearly died doing it. Chris LaFay at Classic City hired full-time staff before he had a sales process and drained his reserves over three years. Chris Morbitzer grew NorthBuilt from two people to roughly ten in anticipation of work that didn't materialize, nearly went under, and cut back to three. The lesson both drew is the same: staff for contracts, not forecasts.

Same starting line. The only difference is whether you hired for contracts or for forecasts. Both paths work. Drifting between them doesn't.

The org chart by growth stage

The practical version of staffing for contracts is knowing who to add and when. Around $1M, the founder is still the rainmaker and the first hires protect delivery quality, not sales. Around $3M, the first real constraint is the founder-as-only-closer, so the highest-leverage hire is a second person who can run sales conversations, paired with a delivery lead who frees the founder from the work. Around $5M, the model needs someone owning recurring revenue and account expansion, because that retainable base is what carries you past the ceiling.

Scaling and staying small are both legitimate. The failure is doing either one by accident.

Strategic Growth Moves of Outlier Agencies That Scale

The founders who broke the ceiling shared a small set of moves. They're worth naming.

The first is specialization, and the math is dramatic. Corey Quinn, the former CMO of Scorpion, watched the company grow from $20M to $150M in six years by going vertical, with the sales team expanding from six people to roughly a hundred. His one-line case for it: "If you market to everybody, you market to no one."

Specialization shows up in retention, too. The most durable agencies in the set cluster around vertical depth: Dusted has kept a single client for twenty years, Fusionbox and ISHIR run ten-plus-year relationships, and Oomph's average client tenure tops five. Depth, not account-management tricks, is what makes clients stay.

The second move is owning a discovery channel before your competitors notice it exists. Shaun Davidson, VP at Sembit and co-founder of Zero Channel, tracked something most agencies haven't measured yet. Over nine months, 28% of Sembit's revenue came from prospects who found the agency through AI and LLM recommendations. In his words, "Twenty-eight percent of our revenue came from a channel nobody on the team was managing." Treat that as a leading-edge signal rather than a cohort average, but the direction is clear: buyers are starting to shortlist agencies through AI before they ever open a search tab.

How structure caps your margins

There's a financial consequence to skipping these moves, and it doesn't require quoting a single profit benchmark to see. An agency with no niche competes on price, which compresses what it can charge. It rebuilds pipeline from scratch every year, which raises the cost of acquiring each client. And it bills for hours, which caps the upside on its best work.

Those three behavioral traps put a ceiling on net margin that no amount of cost-cutting can lift. The structural choices set the margin headroom long before the bookkeeping does.

Capital efficiency differs by niche

One nuance matters here. These dynamics aren't identical across agency types. Dev and software shops carry higher fixed salary overhead, so utilization and retainable revenue matter enormously to whether the model works. Creative and marketing firms often run leaner on payroll but live or die on differentiation, because without it they're one of a thousand interchangeable shops. The trap is universal. The exact pressure point depends on where your costs sit.

Related reading: the micro-niche multiplier and agency positioning as a strategy decision.

Fixing Structural Agency Flaws: The Haus Advisors Approach

Read all 70+ interviews back to back and one conclusion is hard to avoid. The agencies that stall have excellent delivery and a business model they've never actually examined. The work is great. The structure underneath it is accidental.

That gap is the entire reason Haus Advisors exists. We kept seeing founders with flawless delivery and a pipeline that felt like gambling, and the fix was never a new tactic. It was sharper positioning that gave them a pipeline they controlled.

This is the work we do in the Bottleneck Sprint, a focused engagement that finds the one structural constraint holding an agency at its ceiling and reverse-engineers the positioning hiding inside your best client relationships. For founders ready to build the whole growth system, Breakthrough is the embedded version, replacing referral dependence with predictable, owned pipeline over six months. Both exist because the traps in this research are fixable, but only when you treat the model as a thing you design rather than a thing that happens to you.

Research Methodology & Dataset: 70+ Agency Founder Interviews

This research is qualitative. We analyzed 70+ long-form interviews with independent agency founders from the Behind the Agency podcast, recorded between 2024 and 2026. The sample skews toward dev and software shops and marketing and creative agencies, founder-led, in the $0.5M to $5M range. It is self-selected, and it is not a statistical census of all agencies. The value is in the convergence: patterns that recurred independently across founders who never spoke to each other.

We mark findings two ways. 🟢 means a pattern we surfaced from our own interviews. 🔵 means an external benchmark our founders cited. Every quotation in this piece was verified against the source interview.

This is the qualitative companion to the quantitative agency research published by firms like Promethean Research. Their numbers answer what: margins, revenue per employee, growth rates. This answers why: the behavior behind those numbers, in operators' own words. Both are true at once, and you need both to fix a stalling agency.

Cite this research:The Independent Agency Business Model: What 70+ Founders Reveal About Why Agencies Stall and Scale. Haus Advisors, 2026. The charts in this piece are free to embed with attribution and a link back to this page.

The Independent Agency Business Model (Haus Advisors, 2026): https://hausadvisors.com/blog/agency-business-model