Never Vibe-Diversify: 5 Data Points That Say It's Time to Spend Beyond Meta

Getting the "should we spend beyond Meta" decision wrong is expensive in two directions, and almost nobody prices in both.

Diversify too early, into TikTok and YouTube and CTV all at once, and you don't just risk wasted spend. You blow up your ability to measure anything. Meta is easy to read. The moment you have five channels live, each needs its own measurement approach, and if you launched them together and revenue moved, you genuinely can't tell which one did it. You've bought confusion at a higher daily budget.

Diversify too late, and the cost is quieter but just as real. If your account has genuinely run out of new people to reach, every extra dollar into Meta is paying to show the same ad to the same humans again. You feel it as rising costs and softening growth, and you keep blaming the creative, when the truth is the channel is tapped and you're refusing to go anywhere new.

The way you avoid both mistakes is to never make this call on vibes. "We're a year older, we should probably be doing influencer by now" is not a reason. "We're at A$2m, isn't it time for TikTok" is not a reason. The only thing that should move you off Meta is data telling you that you have a new-reach problem.

So here are the five data points I'd actually look at. Treat it like a scorecard. The more of these flashing red at once, the more obvious the answer becomes.

1. Rolling reach on Meta

This is the first thing I check, and it's the most telling.

Rolling reach is, roughly, how much of the audience you're touching is genuinely new versus the same people over and over. A blunt version of the same idea: of all the accounts you reached in a window, what share was a first-time impression?

When that number gets ugly, you'll see something like one in five reached being new, which means four out of five impressions are landing on people who've already seen you. That's a channel starting to talk to itself.

The reason this matters so much is causal, not cosmetic. You can't acquire new customers from people you never reach. If Meta is mostly re-serving your existing pool, no amount of bid tweaking fixes it, because the problem isn't efficiency, it's that you're fishing in a pond you've already fished out.

2. Net-new visit rate

The second signal moves one step down the funnel, from impressions to actual site visits.

Net-new visit rate is the share of the traffic a channel sends you that's hitting your site for the first time, the new-versus-returning split in your analytics. A healthy acquisition channel should be dragging in a strong wave of first-time visitors. That's its job.

When it's working, the contrast is stark. I've seen accounts where a fresh top-of-funnel channel is running an 80%-plus new-visitor rate while the tired channel underneath it is down around 30%. That gap is the whole story: one source is finding new people, the other is recirculating the ones you already had.

So if your main channel's new-visit rate has quietly slid from healthy to low, that's signal number two. You're paying to bring back visitors you'd probably have gotten anyway.

3. Cost per new visit

The third data point is my favourite, because it turns the first two into a number with a dollar sign on it.

Cost per new visit is exactly what it sounds like: how much you're paying to get one genuinely new person onto your site. Not a click, not a returning browser. A new human.

This one cuts through a lot of noise. Topline ROAS can look fine while this number is creeping up underneath it, because you can hold blended returns together for a while by leaning harder on people who already know you. But the cost to reach someone new is the honest price of growth, and when it climbs, it's telling you the channel is getting more expensive at the one job that actually grows the business.

What I love about it is how it travels across channels. Once you know your cost per new visit on Meta, you've got a clean yardstick to judge whether YouTube or CTV is actually cheaper at finding new people, instead of arguing about it on feel. If a new channel pulls in first-time visitors at a lower cost per new visit than Meta, that's not a vibe, that's a reason.

4. Channel efficiency trend

The fourth signal is about direction, not a single snapshot.

Here I'm looking at whether the channel is getting better or worse at its job over time. Is efficiency on Meta flat-to-improving as you spend more, or is it sliding year-on-year even though you're doing everything right? One quarter means little. A clear downward trend across a few of them means something.

The trap to avoid is reading efficiency in isolation. A channel can look "fine" on a given day and still be in a slow decline. So I want the trend line, and I want it next to the growth picture, because the two together tell you what's really happening.

Here's the combination that should make you act. Revenue is roughly flat or slowing, and efficiency is drifting down, and reach and new-visit rate are both falling. When those line up, you don't have a bidding problem or a landing-page problem. You have a reach problem, and the fix is to go find new people somewhere else.

And here's the combination that should make you sit still. Revenue is still growing, efficiency is holding or even improving as you spend more, and your new-visit rate is climbing. That's every light green. You do not need to diversify, no matter how old the brand is or what number you just crossed. Pushing into new channels there would be solving a problem you don't have.

5. Revenue trend, read against the rest

The fifth data point is the one everyone already watches, but most people read it wrong on its own.

Revenue growth is the headline, but in this decision it's only useful in context. Flat revenue with healthy reach and a strong new-visit rate is a conversion or offer story, not a channel-mix one. Flat revenue paired with falling reach, a sinking new-visit rate, and a rising cost per new visit is the channel-mix story, loud and clear.

That's why this is a scorecard, not a single metric. Any one of these alone can mislead you. Stacked together, they make the call almost obvious. If you can tick most of the boxes, losing reach, new visits down, cost per new visit up, efficiency sliding, growth stalling, then the answer is yes, go diversify, and go in knowing exactly which new-reach gap you're trying to close.

The uncomfortable part: most brands fail this test

Now the honest bit, and it's the thing I most want you to hear.

Most brands that come to us asking for TikTok or YouTube don't actually fail this scorecard. They've got a fixable Meta problem dressed up as a channel problem.

When I see a low new-visit rate inside Meta, the cause is very often the creative, not the channel being tapped. A classic version: an account pouring nearly all its budget into sale and discount creative, then wondering why its new-visitor rate sits around 30%. Of course it does. Discount-led ads mostly re-convert people already circling you. They don't go find strangers. That's not a reason to bolt on three new platforms. It's a reason to fix what you're feeding Meta.

So before you diversify, the first move is almost always to get more juice out of the channel you already have. Less of the same purchase-optimised, discount-heavy stuff. More genuinely new creative, more varied angles, more evergreen work built to reach people who've never heard of you. That counts as diversifying your mix too, and it's far easier to measure than standing up CTV from scratch.

Then, if you've done that and the five signals are still flashing red, expand, one or two channels at a time, sequenced and measured properly, not eight at once.

So before you go shopping for a new platform: have you actually run these five numbers, or are you about to spend real money on a hunch that it's "probably time"?