You're Renting Your Customers From Meta: What BuzzFeed's Collapse Teaches DTC Founders

A homewares brand we know spent four years getting very, very good at one thing: Meta ads. Roughly 95% of their revenue came through a single account. The creative was sharp, the structure was clean, the ROAS held in the mid-3s. By every dashboard they looked like a business that had it figured out.

Then their account got restricted for nine days. No clear reason, no warning, just a flag and a queue. Revenue didn't dip in those nine days. It fell off a cliff.

That's the part nobody puts on the dashboard. When one channel is 95% of your revenue, you don't have a strong business with a great channel. You have a channel that happens to have a business attached to it.

The land you build on isn't yours

Here's the thing almost nobody says out loud. In DTC, you don't own your distribution. You rent it.

You're renting it from Meta. From Google. If you go retail, from Target or Costco. You do not reach your own customers directly, not really. You reach them through someone else's pipe, and you pay rent on that pipe every single day in the form of CPMs and CPCs.

I think most founders know this intellectually. What they haven't sat with is the implication: the landlord can change the rules whenever they like, and you don't get a vote.

What happened to the media companies

There's a piece of recent history that should be taught in every DTC onboarding, and it almost never is.

Through the 2010s, a whole generation of media companies built their entire business on Facebook traffic. Facebook was the front door. It sent millions of readers to their sites, those readers turned into ad impressions, and that turned into real revenue. We're not talking about tiny operations. BuzzFeed at its peak was one of these. So were a lot of the big millennial content engines you remember.

Then one day Facebook decided it would rather keep that content on its own platform. So it changed the algorithm. One switch.

It didn't kill those companies for being bad at what they did. A lot of them were excellent at what they did. It killed them because the thing they were excellent at lived entirely on rented land, and the landlord changed the terms overnight.

To be clear, Meta wasn't being evil. It's their platform, their rules, their right. That's exactly the point. You built on someone else's land, and that's a choice with a cost baked into it, even when everything's going well.

The math founders never run

Now, the standard advice in business is focus. Go an inch wide and a mile deep. Find the thing that works and pour everything into it. And honestly, for pure efficiency, that advice is correct. If a brand had gone all-in on Meta from 2016 to 2021 and ignored everything else, it would almost certainly have made more money than a brand that split its attention.

So why diversify at all? Because the focus advice quietly ignores one number.

Picture it as an annual extinction risk. Say there's roughly a 5% chance each year that your one channel breaks in a way you can't fix fast: an account ban that sticks, a policy change that guts your category, a CPM environment that makes your unit economics stop working. 5% sounds small. It feels ignorable on any given Tuesday.

But run it forward. A 5% annual chance of a knockout blow compounds to roughly a one-in-four chance of getting taken out at least once over five years. You wouldn't board a plane with those odds. Yet plenty of brands run their entire company on exactly that bet and never put a number on it, because the channel is working right now and the second channel would be "inefficient".

Here's my take. That second channel isn't an efficiency play. It's insurance. And you don't judge insurance by whether the house burned down this year. You judge it by what it costs you to still be standing if it does.

To put it in money: if your single-channel brand turns over ~$200k/month, even a "wasteful" second channel running at half the efficiency of your main one is cheap compared to the months of revenue you lose the day your main channel flickers. The premium is small. The thing it protects is the whole company.

How to add a second channel without wrecking your numbers

This is where most founders freeze, and I understand why. They've heard "diversify" and they picture pulling 30% of their budget out of the channel that's paying the bills and dumping it into one that isn't proven yet. That feels like setting money on fire, so they do nothing.

You don't have to do it that way. Here's how I'd sequence it.

Start small enough that it can't hurt you. Carve out a fixed slice, maybe 10-15% of monthly spend, and ring-fence it. Tell yourself up front this money is buying resilience, not this month's ROAS. If you measure a brand-new channel against a four-year-old one on day one, you'll always kill it. That's the trap.

Pick the channel that's closest to demand you already have. For most Meta-led brands, that's Google. Specifically the search terms where people are already typing your brand name or your category. That's the cheapest, highest-intent traffic in the building, and a lot of Meta-led brands are quietly letting competitors bid on it. Starting there means your "inefficient" second channel often isn't very inefficient at all.

Judge the blend, not the line item. This is the bit people get wrong. When you turn on a second channel, your isolated ROAS on that channel will look worse than your main one, almost by definition. The question isn't "is Google's ROAS as high as Meta's?" The question is "did my blended MER hold while my single-point-of-failure risk dropped?" If blended efficiency stays roughly flat and you're no longer one ban away from zero, that's a win, even if one line in the dashboard looks uglier.

Layer it in over quarters, not weeks. You're not trying to flip from 95/5 to 50/50 by Friday. You're trying to get to a place where, if any one channel went dark tomorrow, you'd be wounded but not dead. That's a multi-quarter project. Add a little, let it stabilise, read the blend, add a little more.

You should assume it's coming

The mental shift underneath all of this is to operate slightly paranoid. Not anxious. Paranoid in the useful sense: assume that what's working today won't work forever, and make small moves now on that assumption.

Because the brands that get hurt aren't the ones who see it coming. They're the ones who were so heads-down riding the wave that they never built a second wave to step onto. The wave always crashes eventually. Sometimes it's a ban. Sometimes it's your category getting expensive. Sometimes it's an algorithm update you'll never get an explanation for.

The founders who survive it aren't smarter. They just refused to bet the whole company on land they don't own.

Where to from here

If you genuinely don't know what share of your revenue rides on a single channel, or what actually happens to your blend the day that channel stutters, that's worth finding out before the platform finds it out for you.

That's the kind of thing a Signal/Noise Audit is built to surface. We map where your revenue actually comes from, how concentrated the risk is, and where a second channel could go in without dragging your blended efficiency down. No pitch, just a clear read on how exposed you are right now.

So here's the question I'd sit with this week: if Meta restricted your account tomorrow morning and it took two weeks to sort out, what would your revenue look like on day ten? If you don't like the answer, you already know what to build next.