What Actually Works on Meta Ads for E-Commerce in 2026 (And what doesn't)

If your Meta ads have been doing strange things lately, your ROAS keeps slipping, or you have this nagging feeling that the playbook you used to know just does not work the same way anymore, you are not imagining it.

The rules changed. Most online store owners did not get the memo.

In late 2025, Meta rebuilt the entire engine that decides who sees your ads. It is called Andromeda, and it rolled out globally in October. Every single ad account on Meta is now operating under it, whether you have adapted to it or not.

And here is the bit nobody wants to say out loud. Most of the strategies floating around in Facebook groups, agency reels, and old YouTube videos were written for a version of Meta that does not exist anymore.

So in this post, I am pulling back the curtain on what is actually working on Meta right now. Not the shiny version. The practical, sometimes unglamorous things that the brands getting real results are doing in 2026.

We are going to look at what 7 and 8-figure brands like Gymshark, Vero Moda and Princess Polly are doing differently. And more importantly, exactly how you can apply it to your store, even if you have a small budget, a tiny team, and no agency on retainer.

After managing millions in Meta ad spend across e-commerce brands, there's one thing I've learned: the playbook that worked in 2024 is not the playbook winning in 2026.

First, the awkward truth about your old Meta ads playbook

For years, running Meta ads felt like a strategy game you could actually win.

You picked your interests. You stacked your audiences. You excluded the people who did not fit. You tweaked your settings, watched the data come in, and adjusted from there. If performance dropped, you changed your targeting. Simple.

That playbook is gone.

After the Andromeda rebuild, manual interest targeting, lookalikes, and detailed audience segmentation matter far less than they used to. The signals Meta now uses to decide who sees your ad have shifted upstream. It is no longer mostly about who you are targeting. It is mostly about the creative itself.

Industry data is showing that advertisers who have adapted to the new system are reporting 20 to 35 percent higher ROAS than those still running legacy campaign structures. That gap is only going to widen.

If you have spent the last few months staring at Ads Manager wondering whether Meta hates you personally, you are not alone. I have had more conversations than I can count with founders who thought they had suddenly forgotten how to run ads.

They had not.

Meta changed the game.

Your targeting probably is not the problem. Your creative system probably is.

What Gymshark figured out about creative volume

Gymshark is now a brand worth over a billion dollars. They built it largely on Meta.

And here is the stat that blew my mind when I first read it. On a single day in February 2025, Gymshark launched 140 ads. A couple of days later, creative tracking tools showed they were running 340 ads worldwide.

Not three.

Not ten.

Three hundred and forty.

They are not running 340 ads because they have a bigger budget than you.

They are running 340 ads because they know something most brands still have not caught onto.

Winning ads are usually found, not created.

Every ad is a swing. Some miss. Some connect. A handful become your top performers. The more swings you take, the more chances you have to find the angle, hook, or creative that hits.

Inside Meta's new system, this matters more than it ever has. The algorithm needs creative variety to do its job properly. If you are running the same two or three ads for six months, you are starving the system.

What this means for your store

You do not need 340 ads. You do need more than two. Most online store owners I work with are running between three and five ads at a time. If that sounds like you, the single highest-impact change you can make in the next 30 days is to double your active creative count.

And before you panic about budget, this is not a paid spend conversation. This is a creative count conversation. The same ad budget across 10 ad variants will outperform the same budget across three almost every time, because the algorithm has more options to optimise around. That is exactly what we work on inside One Ad Wonder.

What Vero Moda figured out about testing properly

Vero Moda is one of the biggest women's fashion brands in the world. They have something most of us do not. A massive data team.

But what they did with that data is something you absolutely can copy, even on a much smaller scale.

They tested catalogue ad variants in a systematic way and increased sales by 60 percent and traffic by 46 percent. Not by spending more. By testing better.

Here is what makes their approach so worth stealing. They tested one variable at a time. They started with the biggest, most impactful variables first. They did not try to test 15 things at once and end up with data they could not read.

Their order was this. First, they tested types of product imagery: model shots versus packshots versus combinations. Then they tested what product information to include: price versus no price, category labels versus product names. Then, and only then, they tested the smaller details like payment method icons and styling cues.

Their best-performing variant ended up being a model image paired with a packshot. Lifestyle and detail, in the same creative.

The 3-step testing framework you can copy this month

Step 1. Test your hero image

Lifestyle, packshot, or a combination. Pick one as your winner before you change anything else.

Step 2. Test your hook

The first three seconds of a video or the headline on a static. Same image, different hook.

Step 3. Test the offer or angle

Same image, same hook, different framing of what you are actually selling.

Do not change three things at once. You will not know what moved the needle.

This is the part most online store owners get wrong. They get a new content shoot, a new offer, and a new hook, and they launch everything at once. Then when something works, they have no idea why. And when something does not work, they have no idea why.

One of the biggest mistakes I see founders make? They treat every bad day in Ads Manager like a five-alarm fire. One bad Tuesday does not mean your strategy is broken.

The brands that scale are the ones that test one variable, read the data, and then move.

What Princess Polly figured out about identity

Princess Polly is one of the rare Australian fashion brands that has cracked the US market in a serious way. And their Meta strategy is something every fashion and lifestyle founder should pay attention to.

Here is the thing about Princess Polly. They do not actually sell clothes in their ads. They sell an aesthetic.

Nobody wakes up wanting a beige linen shirt.

They wake up wanting to feel confident. Stylish. Effortless. Like the version of themselves they have in their head.

Their ads carry headlines like "your back to school style schedule has been released." They have positioned themselves as the authority on a specific look. They named an entire aesthetic.

Campus core.

When customers Google campus core, when they search for it on TikTok, when they think about that look in their head, Princess Polly is the brand that comes up. They own the category.

Most small brands are still selling products. Big brands are selling a version of who their customer wants to become.

What this means for your store

What aesthetic, lifestyle, or identity does your brand own? What word or phrase do you want your customers to associate with you? If you do not know the answer to that, your ads are likely doing the harder job. They are trying to sell to people who do not yet know why they would choose you specifically.

This is not the sexy part of marketing. It is usually the part that separates the brands that scale from the brands that stall.

I cover this in depth inside the mastermind, but you can start with one exercise tonight. Write down three words your dream customer would use to describe your brand. If you cannot get to three, that is your starting point.

What all three brands know about Advantage+ that most online store owners do not

Every brand we just talked about uses Meta's Advantage+ Shopping Campaigns. Every. Single. One.

Advantage+ now accounts for around 62 percent of all e-commerce conversion spending on Meta globally. It surpassed a $20 billion annualised revenue run rate in 2024 and grew about 70 percent year over year.

If you are not using it yet, or if you are using it but it is not working for you, this is one of the biggest levers left to pull.

And here is the part most guides do not tell you. Advantage+ does not work because of clever audience hacks. It works because the brands using it well are feeding it the right inputs. Strong creative volume. Clean pixel data. Sensible budget allocation. A properly set existing customer cap so they are not just retargeting their own customer list and inflating ROAS.

I have a separate in-depth guide on exactly how Advantage+ Shopping Campaigns work and whether they are right for your store: Meta Advantage+ Shopping Campaigns explained for e-commerce.

How to apply all of this to your store (without a 50-person team)

You do not need a billion-dollar budget. You do not need an agency. You do need a system.

After scaling and selling my own 7-figure fashion brand, then spending the last few years working inside other people's online stores, here is the simple framework I keep coming back to. It works for a $10K month store and it works for a $100K month store. The principles do not change.

1. Build your creative library, not just your campaigns

Stop thinking in ads. Start thinking in assets. Every month, your job is to create enough new creative variants to keep the algorithm fed. Not 340. But not three either. Aim for 8 to 12 new pieces a month, across static, video, and UGC. Different hooks. Different angles. Different formats.

2. Pick one variable to test at a time

Steal the Vero Moda framework. Hero image first. Hook second. Angle and offer third. Read the data after each round. Do not move on until you know which one won.

3. Use Advantage+ properly, not as a magic button

Feed it. Cap your existing customers. Let it run for at least 7 to 14 days before you judge anything. The brands burning budget on Advantage+ are the ones expecting it to perform in 48 hours with three creatives. That is not how it works in 2026.

4. Own an identity, not just a product

Get clear on what you actually sell. The aesthetic. The lifestyle. The transformation. Your ads should look and feel like the world your customer wants to live in, not like a catalogue page.

5. Read the numbers, do not react to them

A bad week of ads does not equal a bad strategy. Pause before you pivot. Look at your ROAS in context with your average order value, your margin, and your traffic. Most of the time, the smarter solution is sitting right in front of you in Ads Manager. You just need to step back and read what the data is actually telling you.

Either way, you do not need a billion-dollar brand to run your ads like one. You just need the right framework.

Now go and audit your active ad count. If you are running fewer than five, that is your homework for the week.

Jodie x

Frequently asked questions