Meta AI Shopping Launched in March — How to Track Who's Winning

Meta AI Shopping Launched in March — How to Track Who's Winning

Meta's AI now recommends products to 3 billion users. You can't apply. Here's how to figure out who's getting picked and why.

Most Operators Don't Know This Exists Yet

Meta launched an AI Shopping tool in March 2026. It recommends products directly inside Facebook and Instagram feeds — carousel cards with images, brand name, price, and a buy button.

No ad spend required. No campaign setup. Meta's AI picks which products show up based on its own signals.

That's a big shift. For the first time, Meta is recommending products to users without anyone running an ad for that specific placement. The algorithm decides who gets shown.

This isn't Meta's first try at commerce. Facebook Shops launched in 2020 and went nowhere because it was a passive storefront. AI Shopping is different — it pushes products into feeds proactively, the same way TikTok Shop recommends products based on watch behavior.

Google did the same thing with Shopping results. Amazon does it with "frequently bought together" and "customers also viewed." The platform that controls the recommendation surface controls who gets seen. Meta controls the feed for 3 billion people.

Most operators haven't caught on. They're still focused on ad campaigns and ROAS targets. But a parallel discovery channel just opened — one that runs on brand signals, not ad budgets.

The brands building the right signals now will own this channel before competitors notice it exists.

You Can't Apply — Meta Picks Based on Signals

There's no sign-up form. No application. No partner program.

Meta's recommendation engine pulls from behavioral data, ad performance history, and brand-level signals it already has. If you've been running Meta ads for months, Meta already knows your creative velocity, your audience engagement, and your product catalog depth.

The brands most likely to get recommended share three traits:

Consistent ad spend. Brands spending EUR 1,000+/day across EU markets feed Meta more data than anyone else. More data means stronger signals. Stronger signals mean higher likelihood of getting picked.

High creative velocity. Brands launching 10+ new creatives per month give the algorithm fresh data points. Stale campaigns with the same three ads running since January don't generate the behavioral signals Meta needs.

Clear brand positioning. Meta's AI reads your landing pages, ad copy, and product descriptions. If your positioning is scattered — selling to everyone, saying nothing specific — the algorithm can't categorize you. Brands with sharp messaging around a specific audience get matched to the right users.

You can't control whether Meta picks you. But you can study which competitors are already building these signals — and match or beat them.

Step 1: Find Who's Spending the Most

The first signal to track is ad spend. Brands investing heavily in Meta ads generate the most behavioral data — clicks, conversions, engagement patterns, audience overlap. Meta's AI Shopping engine relies on this data to understand which products work for which users.

You need to know who's outspending you in your category. Not a rough guess — actual numbers.

I open Brandsearch Discovery, filter to my niche, and sort by Total Adspend (EU). The EU Adspend data shows real Meta spend figures — not estimates from third-party scrapers.

A competitor spending EUR 2,400/day across France, Germany, and the Netherlands generates 10x the signal volume of a brand spending EUR 200/day in one market. That's the brand Meta's AI knows best.

I check the top 5 spenders in my niche. If three of them are scaling spend month-over-month, that category is about to get crowded in Meta AI Shopping. If I'm not in the top 5, I know exactly the gap I need to close.

Here's a real example. I looked at the supplements niche last week. The top spender was running EUR 3,800/day across four EU markets with 45 active ads. The second-place brand was at EUR 1,200/day with 18 active ads. That gap tells me the top brand is generating roughly 3x the signal volume. If Meta's AI is picking favorites in supplements, brand #1 has a massive head start.

Step 2: Track Who's Actually Scaling (Not Just Spending)

Ad spend alone doesn't tell the full story. A brand could be spending heavily on a campaign that's failing. You need to see who's scaling — not just who's burning budget.

In Discovery, I filter to Phase: Winning. This shows ads that passed the testing phase and are actively scaling. A winning ad means the brand found a creative that works and is pouring budget behind it.

A brand with 15+ winning ads across Meta is doing something right. Their creative strategy is feeding Meta exactly the kind of engagement data that powers recommendations.

I look for three things in the winning ads:

Format. Are the winners mostly video or static? If 80% of the winning ads in your niche are video, that's the format Meta's algorithm responds to.

Running days. Ads running 30+ days in the winning phase are printing money. The brands behind them have stable unit economics and consistent signal generation.

Creative diversity. A brand with 10 winning ads that all look different is testing aggressively and finding multiple angles. A brand with 10 winning ads that look the same found one angle and is milking it. The first brand builds a stronger signal profile.

I also check the Brand Analysis Overview tab for each of these competitors. A brand with 15 winning ads AND traffic up 25% month-over-month is a confirmed threat. Their ads are working, their traffic is growing, and Meta's algorithm is getting fed from both sides.

A brand with 15 winning ads but flat traffic? They're spending money but not building a moat. Their signals are weaker than the numbers suggest.

I save the top 3-5 scaling competitors to a Swipe File folder — "Meta AI Shopping Competitors Q2" — so I can check back weekly.

Step 3: Reverse-Engineer Their Positioning

Meta's AI doesn't just look at ads. It reads landing pages, product descriptions, and brand messaging to understand what a brand is about — and who it's for.

Your positioning isn't just a marketing exercise. It's a signal that determines whether Meta's AI can match your products to the right users.

I open Brandsearch Brand Analysis for each competitor I'm tracking and go to the AI-Radar tab. It shows the brand's extracted USPs, target keywords, audience pain points, and core benefits — all pulled from their actual copy and landing pages.

This is what Meta's AI is reading too.

A brand with clear positioning — "premium protein for men over 40 who hate chalky shakes" — gives the algorithm a clean signal. It knows exactly which users to show that product to.

A brand with generic positioning — "high-quality supplements for everyone" — gives Meta nothing to work with. The algorithm can't differentiate it from 500 other supplement brands.

I compare 3-4 competitors in AI-Radar side by side. The one with the sharpest positioning and the highest ad spend is the one most likely to dominate Meta AI Shopping in that category.

Here's what I found checking three cookware brands. Two of them positioned around "professional-grade" and "chef-approved" — nearly identical messaging. The third focused entirely on "the last pan you'll ever buy" — durability and warranty. That third brand had 30% more active ads and rising traffic. They owned the durability angle. The other two were fighting over the same lane.

If you find a positioning gap like that in your niche, that's your entry point for Meta's AI. A clear, differentiated message gives the algorithm something to categorize.

Build Your Signal Profile Before They Lock the Category

Meta AI Shopping is early. The carousel positions aren't locked. But they will be.

Once Meta's algorithm identifies the top 3-5 brands in a category, those brands generate the most clicks. More clicks mean more data. More data reinforces their position. It's a flywheel. Early movers compound their advantage.

Here's what I'm doing based on the competitor research:

Matching creative velocity. If the top brand in my niche launches 12 new creatives per month, I need to be at 10+ minimum. The algorithm needs fresh data. Recycling the same three ads won't cut it.

Sharpening positioning. I rewrote my landing page copy to target a specific audience segment — not "everyone who wants X" but "operators who need X because Y." That's the kind of clear signal Meta's AI can actually use.

Increasing EU spend. The EU Adspend data showed me my competitors are spending 3x what I am in key markets. I don't need to match them immediately, but I need to close the gap enough that Meta's algorithm considers my brand in the same tier.

Tracking weekly. I check my saved competitors in Brandsearch every Monday morning. If a new brand enters the top 5 by spend or winning ad count, I need to know immediately — not three months from now.

Diversifying traffic sources. Don't rely 100% on Meta paid. Run organic content on TikTok and Instagram. Build your email list. The brands Meta's AI will favor are the ones with strong signals across the whole ecosystem — not just a big ad budget on one platform. A brand with 35% paid, 30% organic, and 20% direct traffic looks like a real brand to the algorithm. A brand with 90% paid traffic looks fragile.

The brands that build these signals in Q2 2026 will have a structural advantage by Q3. The ones that wait will be playing catch-up against an algorithm that already has its favorites.

The 20-Minute Weekly Monitoring System

Meta AI Shopping isn't a one-time research project. The signal landscape shifts every week as brands scale, pause, or pivot.

Here's the routine I use:

1. Monday (10 min): Open Brandsearch Discovery. Filter to your niche, Phase: Winning, sorted by Total Adspend (EU). Check who's in the top 5 this week. Note any new entrants or brands that dropped off.

1. Thursday (10 min): Open your saved competitors in Brandsearch Brand Analysis. Check each brand's AI-Radar tab for positioning changes. If a competitor just sharpened their messaging around a specific pain point, that's a signal — they're preparing for AI-driven discovery.

Save notes in your Swipe File folder. Over a month, you'll have a clear picture of which brands are building Meta AI Shopping signals and which are standing still.

I keep a folder called "AI Shopping Signals" with the top 5 brands in my category. Every Monday I check their ad count, spend trajectory, and newest creatives. Takes 10 minutes. The patterns become obvious after 2-3 weeks — who's scaling, who's stalling, who just launched a new angle worth studying.

The operators who track this now will know exactly where they stand when Meta expands the program. Everyone else will find out when their organic reach drops and they don't understand why.

The Bottom Line

Meta AI Shopping changes how products get discovered. Instead of paying for every impression, brands can get recommended to 3 billion users — if they've built the right signals.

You can't apply. But you can prepare.

The method:

1. Track competitor ad spend — Brandsearch Discovery with EU Adspend sort shows who's feeding Meta the most data
2. Identify who's scaling — Brandsearch Discovery Phase: Winning filter shows which brands have creatives that work
3. Reverse-engineer positioning — Brandsearch Brand Analysis AI-Radar tab shows what Meta's AI reads about each brand
4. Build your own signals — match creative velocity, sharpen positioning, close the spend gap

The algorithm already has favorites. Find out who they are before your category gets locked.