How to structure Meta ads for a subscription brand

If you run a subscription brand, you already have one big advantage on Meta: your customers come back. Whether it’s a consumable, something that needs refilling or a product people order every week, your lifetime value is usually a lot higher than a regular ecommerce store’s. That means you can afford to pay more to acquire someone.

This is the structure we use to make the most of that. It’s the one that helped a meal prep brand we work with grow from under $500K a month to over $1.7 million in about a year and a half.

What holds most subscription brands back on Meta?

Segmentation. Most of the subscription accounts we look at have a bunch of campaigns, each with its own exclusions, interests or lookalikes. Or there are separate ad sets testing different creatives, all going after the same people.

A few years ago that made sense, because you had to tell Meta who to target. Now the algorithm finds the right buyers for you. When you split your budget into lots of small pockets, each one gets less data, and they end up competing against each other.

What’s the best Meta campaign structure for a subscription brand?

It’s really simple. One CBO for scaling, and the only thing it excludes is:

• Active subscribers
• Anyone who’s purchased in the last 90 days

That’s it. The aim is either completely new customers or lapsed ones you can win back, and that exclusion tells Meta exactly which audience you’re focused on. For the meal prep brand, this one campaign now spends around $8,000 per day.

Next to it sits a separate testing campaign, which feeds the CBO its best new ads. More on that below.

If you don’t segment audiences, how do you reach different buyers?

With everything consolidated, the ads themselves do the targeting. That makes the second piece of the system producing strong, diverse content, and a lot of it.

For the meal prep brand, we did a full persona analysis, looking at all the reasons people buy: weight loss, keto, muscle building, that kind of thing. Every batch of ads gets built around those personas. Then we mix up the formats as much as we can:

• Customer review videos
• UGC from creators
• Founder podcast clips
• A whole range of static concepts

Our copywriter writes the scripts and briefs for all of it, and our editor turns the raw material into finished ads. The goal is that no two pieces look or sound the same, because each one is speaking to a specific type of buyer. If you want a way to plan that variety, we break it down by buying stage in our guide to scaling Meta ads on a small budget.

How do you get customer review videos for ads?

Reviews are one of our strongest sources. This is how we collect them:

1. Every new subscriber gets an email, and there’s a little card in their box with a QR code.
2. They scan it, it asks them three questions, and they film their answers.
3. We get a notification every time someone leaves one.
4. We edit the good ones into ads.

It’s real customers talking in their own words, and fresh ones keep coming in.

How do you get more out of UGC?

We send the product out to creators, and the key is that they give us full rights to the footage. That means we can keep using it, and recut it around completely different messages to come up with new concepts.

We also generate AI voiceovers with ElevenLabs and lay them over those clips. The one thing to watch: each version has to stand on its own, with its own message and structure. It can’t be the same ad with another voice on it.

How should a subscription brand test new Meta ads?

Testing is the third piece, and this is where we go deep.

1. Run a separate ABO testing campaign. It runs two or three ad sets at a time. For the meal prep brand, that’s around $2,000 per day in total.
2. Use push delivery on every ad inside each ad set. It spreads the spend fairly across everything being tested.
3. Give it a week, up to a fortnight.
4. Judge each ad outside Ads Manager. When you’ve got consistent repeat purchases, the numbers in there can be skewed quite heavily. Instead, we review each ad in a third-party attribution tool and look for new customer revenue. Was this ad the first click in the customer’s journey, or the last one before they bought? If it’s contributing to either, it’s a real signal.

Push delivery works here because every test gets judged in that attribution tool, not on the spread of spend alone. For more on running testing and scaling side by side, see the ABO + CBO hybrid framework.

How do you decide which ads to scale?

That attribution check is how we decide. Only the best ads from each batch make it across, which means the CBO is full of proven winners that are already warmed up.

When performance is good, increase the budget on the scaling campaign and leave testing where it is. That’s the whole system.

What else decides whether this works?

The structure is only one part of it. Your offer, your landing pages and your retention flows all play a big role too.

And the whole testing process depends on your tracking seeing every sale. Most accounts are missing a big chunk of them, so before you lean on new customer revenue to pick winners, make sure you’ve fixed your tracking before scaling.