Five People, $100M: How AI Is Shrinking the DTC Org Chart (and What to Keep In-House)

A few years back I watched a brand I was close to hire a six-person performance team to "get serious about scale". Within four months they were spending ~$80k/month on Meta and most of that team was busy keeping itself busy. The founder was in standups about standups. The account wasn't any better than when one sharp person ran it.

That's the mistake I keep seeing, and it's the exact opposite of where the smart operators are heading.

There's a thesis doing the rounds at the moment that's worth taking seriously. The best builders are saying they could scale a brand toward nine figures with around five employees. One operator put it bluntly: with the right channel decisions, you could run a brand to roughly $100M in revenue with five people. Not five people per function. Five people total.

I don't think that's hype. I think it's the direction the whole industry is moving, and it changes how you should be drawing your org chart right now.

Why the org chart is shrinking

Here's the thing - the cost of doing most things is collapsing.

Building a landing page, drafting a creative brief, cutting a first version of an ad, pulling a dashboard together, writing the email flow. All of that used to need a person, or three. Now a lot of it is near-free if you've got someone with judgment pointing the tools in the right direction.

The same operators making the five-person argument said something I keep coming back to: the actual creation of assets is getting easy, but you still need judgment. And the value of that judgment doesn't fall. It goes up. Maybe 10x. When everyone can produce a decent ad in an afternoon, the person who knows which ad to make, to whom, and why, is worth more than ever.

So the new org chart isn't "the same roles but fewer of them". It's a different shape entirely. A small core of high-judgment people, plus tools and partners doing the execution that used to soak up headcount.

For a 6-8 figure DTC brand, I'd think about it in three layers.

Layer one: what a founder team keeps in-house

This is the bit most people get wrong. They outsource the strategy and keep the busywork. It should be the other way around.

What I'd keep in-house, even if you're a two-person founder team:

• The brand and the offer. What you sell, who it's for, how it's priced, the story. Nobody can own this for you. This is the actual company.
• Customer truth. Reading your reviews, watching your DMs, talking to people who bought and people who churned. The texture of why people buy is your single most valuable input, and it should never leave the building.
• The numbers that decide things. Your contribution margin, your CAC-to-LTV, the handful of decisions that actually move the trajectory. You don't need a finance team for this. You need to not delegate it.
• One ops hire. Someone who keeps the machine running - inventory, logistics, the systems, the shared services. The unglamorous spine of the business.

That's basically it. Founders on growth, one person on ops. Everything else is a question of partner or tool.

I believe the trap is hiring to feel legitimate. A bigger team feels like progress. But every person you add is a person to manage, a Slack thread to follow, a decision that now needs a meeting. To put this in perspective: a lean two-person team that ships a new angle the same week they think of it will usually beat a ten-person team that needs three meetings to approve it. Speed is the whole edge, and headcount quietly eats speed.

Layer two: what you hand to a partner

Some things shrink to a tool. Some things need a person with reps, just not a person on your payroll.

Paid acquisition is the clearest one. It's the layer where a small mistake is expensive, where the platform changes under you constantly, and where the difference between "fine" and "good" is the difference between a brand that scales and one that stalls. You want someone who has run a few hundred accounts pattern-matching against yours, not someone learning on your budget.

This is where I'd argue an agency earns its place in the lean org chart, and it's specifically what we do at Pigeon. Not as another layer of overhead, but as the execution and judgment for the acquisition engine so the founders don't have to build and manage that whole function in-house. One relationship instead of four hires, a creative testing system instead of a creative team, kill and scale discipline instead of a media buyer you have to performance-manage.

The test I'd apply: is this a function where reps and pattern-matching across many brands genuinely beat doing it yourself with AI? If yes, partner. If it's just labour, that's layer three.

Layer three: what you hand to AI

Then there's the execution that used to be headcount and is now mostly tooling.

First drafts of nearly everything. Variations on a winning ad. Research synthesis. Dashboards. The repetitive production work that used to justify a junior hire or two. None of this is where your edge lives, so none of it should cost you a salary.

A worked example of the shape I mean. Picture a supplements brand doing ~$3M a year. The old chart: founder, a marketing manager, a media buyer, a designer, a content person, a part-time analyst, a couple of VAs. Seven or eight people, and the founder still feels stretched.

The lean version: two founders on growth and brand, one ops hire on the spine, an agency partner running acquisition, and AI tooling doing the production that the marketing manager, designer, and analyst used to split between them. Four humans instead of eight, moving faster, with the founders closer to the customer rather than buried in management.

That's not a cost-cutting story. It's a speed-and-focus story that happens to cost less.

What doesn't shrink

I want to be honest about the limits here, because the "five people" line gets quoted like a magic trick and it isn't one.

A few things don't compress.

The making of physical things still costs what it costs. Manufacturing, inventory, MOQs, the cash tied up in stock. AI doesn't ship your pallets.

And the whole lean model rests on one thing: the quality of the small team. The operators making this argument were clear that it only works if the few people you've got are genuinely excellent and can run independently. A lean org with mediocre people isn't lean, it's just understaffed. The model amplifies whoever's in it - so a weak team gets you a weak company faster.

The other quiet risk is hiding execution inside your own head. When you cut the team right down, the founders end up holding a lot. That's fine if you've handed the heavy execution layers to partners and tools. It's a problem if you've just deleted the roles and kept all the work.

Where to from here

If I were drawing my org chart for 2026, I'd start from a blunt question: which of these roles is judgment, and which is execution I'm paying a salary to do by hand?

Keep the judgment. Hand the execution to a partner where reps matter, and to tooling where they don't. Then protect the speed that small teams have, because that's the only real advantage you're buying.

If you're not sure which layer your acquisition actually sits in - whether it's genuinely a judgment problem you should keep close, or an execution layer you're overstaffing - that's exactly the kind of thing a Signal/Noise Audit is built to show you. We look at your account, your creative history, and your unit economics, and tell you where the headcount and spend are actually earning their keep, and where they're just keeping busy.

What would your org chart look like if you had to draw it from scratch tomorrow, judgment first?