When Should You Actually Diversify Off Meta? Both Sides of the $20M Debate

"We're at A$22m now, surely it's time we got serious about other channels."

A founder said that to me on a call a few months back, almost as a foregone conclusion. Revenue had crossed some number in his head, so diversifying off Meta felt like the obvious grown-up move. And I get why. But the number that triggered it - the revenue - is the wrong one. It tells you almost nothing about whether you've actually run out of room on Meta.

This is one of the loudest arguments in DTC right now, and it's worth laying out honestly because smart people land on opposite sides of it.

The case for staying on Meta longer

One camp, and Taylor Holiday is the clearest voice here, argues that most brands leave Meta too early. They get a year or two older, decide they "should" be doing influencer or CTV or whatever the timeline is excited about, and they spread thin before they've squeezed what Meta can still give them.

I think this side is mostly right, and here's the uncomfortable bit: when you diversify, your measurement gets exponentially harder, not a bit harder. Meta is easy to measure. Your business is more or less either working or it isn't. Post-purchase survey data is more trustworthy when one channel dominates. Click-based attribution actually holds together.

The moment you light up five channels, every one of them needs its own measurement approach, and a lot of them - CTV, linear, audio, out-of-home - can't be measured on a click at all. So the risk of pouring money into a channel and not really knowing if it worked goes way up. If your Meta account is still reaching new people efficiently, leaving it to go "diversify" is often just trading a channel you can measure for several you can't.

The case for leaving sooner

The other camp says go earlier, suck at the new channels on purpose, and start paying tuition before you're forced to.

The logic is sharper than it first sounds. Eventually almost every brand reaches a scale where it spends money everywhere - that's just what big brands do. So the question isn't if you'll diversify, it's whether you do it while Meta is still carrying you, or after it's tapped out and your acquisition cost is climbing.

And here's the kicker that won me over to taking this seriously: waiting doesn't make you good at the new channels. Spending two extra years only on Meta doesn't teach you a single thing about YouTube or TikTok or CTV. You're going to be bad at them whenever you start. So if you're going to be bad anyway, you'd rather be bad while Meta is still profitable and covering the learning curve, not bad while it's inflating your CAC and bleeding contribution profit.

There's a second reason that gets missed. People like us, who live online, badly underestimate how many of our customers don't. Plenty of buyers - often the wealthier ones - are actively being told to put their phones down, and they're doing it. Offline impressions reach people Meta structurally can't. The alpha tends to sit in the channels that are harder to access, not the wide-open auction where you're bidding against the entire planet.

Why "what revenue?" is the wrong question

So you've got one side saying don't leave too early and one saying don't leave too late. Both are right. Which means the real question was never timing in the abstract - it's "how do I know which situation I'm actually in?"

And the answer almost everyone reaches for is revenue. A$10m, A$20m, A$50m - pick a threshold, cross it, diversify. I think that's the single biggest error in this whole conversation.

Here's my take: saturation isn't a revenue number. It's how many customers you've acquired relative to the size of the category you sell into.

You can be a A$30m or A$40m brand, still mostly on Meta, that has quietly acquired a huge share of the people in your category who'll ever realistically buy. At that point the next customer on Meta is genuinely hard to find, and you should be spending more off it. You can also be a A$30m brand that's barely scratched its category, in which case Meta has loads of new people left to show you and diversifying early would just spread you thin for no reason. Same revenue, opposite decision.

The cleanest way I've seen this framed: two brands can sit at wildly different revenue and need almost identical media mixes, because they acquire customers at a similar rate into a similar-sized market. Imagine a cookware brand with an A$400 average order value and a wallet brand at A$50. The cookware brand could be eight times the revenue. But if they're signing up roughly the same number of customers each Black Friday, into categories of roughly comparable size, their saturation is similar - and so is their need to diversify. Revenue hid that completely. Customers-into-category exposed it.

A penetration model you can build in a spreadsheet

You don't need a data team for this. You need an honest estimate and a few columns. Here's how I'd run it.

1. Estimate your serviceable category size. Not the fantasy total market - the realistic pool of people who would actually buy a product like yours at your price, in the geographies you sell to. If you're a A$120 premium dog supplement in Australia and New Zealand, that's a specific, finite number of dog owners who'll pay premium, not "everyone with a dog." Be conservative. This is the denominator.

2. Count cumulative unique customers acquired. Pull lifetime unique buyers out of Shopify. That's your numerator. Use customers, not orders - a returning buyer isn't a new person reached.

3. Divide to get rough penetration. Customers over category size. If you've sold to ~40,000 people in a realistic pool of ~200,000, you're at roughly 20% penetration. That single number tells you more about your Meta runway than your revenue line ever will.

4. Layer in the account signals, because penetration is the strategy and the signals are the proof. Penetration tells you the shape of the problem. These four tell you it's actually happening right now in the account:

• Rolling reach on Meta. What share of accounts you're reaching are new impressions? If only one in five is new, you're recycling the same people.
• Meta efficiency year-on-year. Is the account getting less efficient even as you spend more? Rising CAC into a shrinking pool of new people is the classic saturation fingerprint.
• Unique new-visitor rate to site. Of your sessions, how many are first-time? A falling first-time rate means your top of funnel is drying up.
• Cost per new visitor. Trending up means each genuinely new person is getting more expensive to reach.

5. Read the two together. High penetration plus all four signals flashing red is a real saturation problem - diversify, and pay the tuition. Low penetration with the same signals is usually not a channel problem at all. It's a creative problem. You haven't exhausted Meta's new people; you've exhausted the angles you're showing them.

That last line is the one I'd underline. The brand that's at 8% penetration but seeing reach collapse doesn't have a Meta problem. It has a sameness problem. Fresh angles, fresh creative, less purchase-optimised campaigns, and Meta will keep finding new people. Diversifying there would treat a fixable creative issue as if it were a structural ceiling, and you'd waste a year and a pile of cash learning channels you didn't need yet.

So when do you actually go?

Get more juice out of Meta first. Almost always. Before you touch a new channel, ask whether the real issue is that you've run out of people, or run out of ideas. If reach is down but penetration is low, the answer is new creative and new angles inside Meta, not a new channel.

If penetration is genuinely high and the account signals confirm it, then go - and go in sequence, not all at once. Start with the channels where you already have ammunition: if your vertical video works on Meta, YouTube and TikTok are a short hop with self-serve buying and easy-ish measurement. The harder, more bespoke stuff - CTV, audio, the things that need fresh creative and can't be measured on a click - come later, one at a time, so you can actually tell what each one did. Launch eight channels in a quarter and you'll never untangle which one moved the business.

And do not, under any circumstances, vibe-diversify. If you can't point to the penetration number and the account signals that say you've run out of new people, you haven't earned the move yet. "We feel like we're big enough" is not a reason. The data either says you've hit the edge of your category on Meta, or it doesn't.

Which brings me back to that founder on the call. The honest answer to "we're at A$22m, isn't it time?" is: I don't know yet, and neither do you, because A$22m isn't the number that decides it. Go build the penetration model. Count the customers, estimate the category, pull the four signals. Then we'll both actually know whether you've saturated Meta or just need better ads.

So before you spread your budget across five new channels this year, here's the only question worth answering first: have you genuinely run out of new people in your category, or have you just run out of new ways to talk to them?