Meta's Favourite Child:

The Ads Method That's Actually Working Right Now

If you've been running Meta ads for any length of time, you'll know that the platform has a habit of doing whatever it wants regardless of what you think is best. And honestly? The sooner you make peace with that, the better your results are going to be.

There's a pattern I've been seeing play out consistently across ad accounts recently. It's not complicated, it's not a hack, and it's definitely not the kind of thing you'll find in a Meta Blueprint course (You will find it in my Facebook Ads Playbook though!) . But it's working and I think it's worth talking about.

Launch Wide, Let Meta Decide

Here's how it starts. You launch a new campaign with one ad set, clean, structured, no overcrowding. Inside that ad set, you load up a stack of ads. For budgets under £2k a month, I've found the sweet spot is somewhere between seven and ten creatives. Not random ones either, you want variety with purpose.

That means ads speaking to buyers at different points in their awareness journey: some for people who don't even know they have a problem yet (symptom aware), some for people who know something's wrong but haven't found the fix (problem aware), and some for people who are actively weighing up their options (solution aware). Mix up your formats too, static images, carousels, video, a dynamic product carousel, some UGC, maybe a brand story.

You're essentially casting a wide net and letting Meta figure out what lands.

And figure it out it will.

Meet the Favourite Child

What happens next is where it gets interesting. Meta runs through your ads, gives them each a bit of airtime, and then quietly makes a decision. One ad starts pulling ahead. And from that moment on, the rest might as well not exist.

This is what I call Meta's favourite child. The algorithm found the creative that's driving results, and it is done entertaining the others. It doesn't care what the rest of the pack thinks. It doesn't care what you think. You could pour your heart and soul into briefing a designer or writing copy you're convinced is going to smash it, and it'll be greeted with absolute crickets. The favourite child is getting the sales and Meta is sticking with it.

Here's the thing though: it's working. So why would you fix it?

Your Job Has Changed

Once Meta has picked its winner, your role shifts. You're no longer a campaign manager, you're a support act. And the most important thing you can do at this point (and make sure your client understands this too) is ensure that winning product is in stock. Nothing kills a performing ad faster than sending people to an out-of-stock page.

Beyond that? Leave it alone. Have some creative variations ready to rotate in when reach and frequency start climbing, that's your signal the ad is getting fatigued and a fresh face is needed but until then, keep your hands off it.

I'll say it one more time, just to be clear: do not touch it.

Now Here's Where It Gets Really Good

So you've got your favourite child doing its thing. Great. But you're not done.

What I've been doing recently is building a new ad set alongside the winning one, using audience suggestions (and yes, they really are just suggestions at this point, Meta is going to do what it wants with targeting regardless). Your audience segments will show you how spend is splitting between retargeting and new customers, even if you've tried to guide the algorithm towards fresh audiences. Use that information, but don't stress about controlling it too tightly.

Set your budgets at ad set level, load the new ad set up with fresh creative, and let it run. Your existing ad set keeps the engine ticking over and the sales coming in. The new one goes out and hunts for new customers. You're not blowing everything up and starting again, you're building on what's already working while giving Meta fresh material to find new people with.

The results of this approach in one recent account genuinely stopped me in my tracks. I launched a new ad set with new ads running alongside an existing one that was sitting at a solid 5x ROAS. Within two days, the new ad set hit 11x ROAS.

Now, hand on heart, I'm not expecting it to hold at that number. ROAS like that tends to settle as spend scales and the initial burst of algorithm learning evens out. But the point isn't the exact figure. The point is that this method, in that account, was absolutely gold. It showed that there was a whole new audience out there waiting to be found, and giving Meta the right creative in the right structure was all it needed to go and find them.

Let it be.

Their ai image creation may be a shocker, but Andromeda is working if you let it!

Stop fighting Meta. Stop assuming you know better than the algorithm. Launch smart, let it learn, protect what's working, and feed it fresh creative to find new customers ,without burning down what's already performing.

The favourite child is your friend. Your job is to keep it happy and make sure the shelves are stocked.