Here is what usually happens when a shopper taps an ad for an online store. They are on their phone, half-interested, thumb already moving. They tap. A browser opens inside the app, your site starts loading, a cookie banner slides up, a pop-up offers ten percent off, and somewhere behind all of that is the product they tapped on. You paid for that tap. Most of the time, the shopper is gone before the page settles.
Meta built an ad format for exactly this problem, and almost nobody uses it. It is called a collection ad. It opens a full-screen storefront inside Facebook or Instagram, lets the shopper browse your products without waiting on your site, and only sends them out to you once they have picked something. It takes a little more setup than a single image, which is most of the reason it gets skipped.
It also just got harder to find. Meta is rolling out a new layout in Ads Manager, and for a lot of accounts collection no longer appears where it used to, in the Ad setup section. It now lives further down, in Ad creative, under Format display options. If you went looking for it recently and decided it had been retired, it hasn't. It moved.
The Ad Format Almost Nobody Runs
Most small stores run the same three shapes of ad: a single image, a video, and a carousel. That is not a bad lineup. But all three share one weakness, and it is the step after the tap. Every one of them hands the shopper straight to your website, which means every one of them is only as good as your mobile site on a slow connection.
That is a bigger problem than most owners realize. We have written a whole piece on how every extra second of load time costs you customers, and another on why the site that looks great on your laptop is not the one your customers see. Fixing your site is always the real answer. But while you are fixing it, you are still paying for every tap that lands on it.
A collection ad changes where the shopper spends the first thirty seconds after the tap. Instead of waiting on your site, they are browsing a page Meta hosts, built from your own catalog, that opens as fast as anything else in the app. By the time they reach your website, they are arriving at a specific product they have already chosen, not at your homepage wondering where to start.
What a Collection Ad Actually Is
In the feed, a collection ad looks like one large image or video with a row of smaller product images underneath it. Meta requires at least four products or images. It only runs on mobile placements, never on desktop, which is fine, because that is where the problem it solves lives.
The important part is that the two halves of the ad do different things when someone taps them:
- Tap the cover image or video, and a full-screen Instant Experience opens. This is the storefront: your cover at the top, your products underneath, the shopper scrolling and tapping around without ever leaving the app.
- Tap one of the product images, and the shopper goes straight to that product's page on your website.
So the ad gives the shopper two routes. The ones who already see what they want take the short one. The ones who are only curious take the long one and browse first. Either way, the shopper picks the route, not you.
The Instant Experience itself is built from a template. For a store with a product catalog connected to Meta, the one you want almost every time is Storefront, which pulls your products straight from the catalog. Lookbook tags products inside lifestyle photos, but it does not run on Instagram at all. Digital circular is built for pushing people to a physical store and does not run in Instagram Reels. Customer acquisition works without a catalog, which is how a business with no product feed can still run the format.
Your Brand Video and Your Catalog Photos Are Weak Alone. Together They Sell.
Think about the two kinds of images most online stores have lying around.
The first is catalog photography. A product on a white background, shot straight on, one after another. Those photos are honest and useful. They answer "what exactly am I buying?" But nobody ever stopped scrolling for a candle on a white background. On their own, they are a spec sheet with pictures.
The second is brand content. A video of the product in someone's life, a founder explaining why they started, a clip of the thing being used in the backyard at sunset. That is what makes a stranger care. But it rarely tells anyone what to buy or what it costs. The viewer finishes the video interested and with nowhere obvious to go.
A collection ad stacks the two. The cover does the persuading, and the products underneath answer the question the cover raises. That combination is the real reason to use the format, and it should shape what goes in the cover. Do not put a fifth product shot up there. Put the reason to care: the product in use, the problem it solves, the moment someone would reach for it. Let the tiles below handle the "which one, and how much."
If you already have video that performs in your other ads, start there. It has already proven it can stop a thumb. And if your catalog has video attached to products, that is worth a look too. We covered why most catalogs still carry no video at all, and the collection storefront is one of the places it pays off.
When It Beats Sending Shoppers to Your Site, and When It Loses
A collection ad is not better than your other ads. It is better at one job. Here is how to tell whether that job is yours.
It tends to win when:
- Your mobile site is slow or cluttered. This is the biggest reason. The storefront opens inside the app, so the shopper's first impression of your range is not held hostage by your theme, your apps, and your pop-ups.
- People browse before they buy. Apparel, home goods, gifts, anything where a shopper wants to see ten options before settling on one. A storefront is a shelf, and shelves work for browsers.
- The price allows a fairly quick decision. A shopper can go from a feed ad to a product page in two taps. That suits products bought on impulse or after a short look, not products that need a week of research.
It tends to lose when:
- You sell one hero product. There is nothing to browse. A single strong ad pointed at a great product page will almost always do better.
- The purchase needs your full page to close. Expensive, technical, or high-trust purchases need reviews, comparisons, guarantees and room to explain. A storefront gives the product a tile and a price. That is not enough room to make a four-figure case.
- You sell a service. You can technically run a collection ad without a catalog, using the customer acquisition template. But for most service businesses, a real landing page does that job better, for the same reasons we laid out in why Meta's instant form gets you cheap leads and a landing page gets you customers.
Your Product Set Is the Ad. Build It Like One.
With the Storefront template, you do not upload the products in the ad one at a time. You pick a product set from your catalog, and the set decides what can appear. That makes the product set the most important creative decision in the whole ad, and it is the one most owners make without thinking.
The common mistake is choosing "All products." It is the default and it is the easy choice. It is also how a store ends up advertising a gift card, a replacement part, a clearance item it is trying to get rid of at cost, or a product from a completely different line than the one the cover video is about. Meta will show whatever the set allows. If it is in the set, assume it will be in the ad.
This is not an argument for a tiny set. Meta's own guidance is to use a large product set, more than 50 products where you can, so that dynamic ordering has room to show each shopper the items they are most likely to want. It also notes that out-of-stock items drop out of the set automatically. Size is fine. The problem is products that do not belong together or should not be advertised at all.
So build the set like a shelf, not a warehouse. That is the same idea we used for collection pages on your own site: one coherent theme, deep enough to browse. A few rules that hold up:
- Match the set to the cover. If the cover video is about summer outdoor furniture, the set is outdoor furniture, not the whole store.
- Leave out anything you would not pay to advertise. Gift cards, parts, samples, low-margin filler, and items you are clearing at a loss all come out.
- Let Meta order it, unless you have a reason not to. "Order dynamically" is Meta's recommended setting. It lets the delivery system pick which products appear in the ad, ranked by popularity or likelihood of purchase. Hand-pick the order only when you are running a specific promotion and need specific products in front.
Two small settings are worth checking before you publish. First, the storefront's button. It should point to a page on your site that shows more of the same products, like a category or collection page. Leaving it on a generic default, or pointing it at your Facebook page, wastes the shopper at the exact moment they are ready to leave the app. Second, add URL parameters to that destination, so your own analytics can tell collection traffic apart from everything else.
Meta Will Count a Click That Never Left Facebook
This is the part that trips up almost everyone who tests the format, and it can make a collection ad look far better than it is.
In Meta's own definition, link clicks counts clicks "to destinations or Instant Experiences, on or off Facebook properties." Read that twice. When a shopper taps the cover and the storefront opens, Meta counts a link click. Nobody has reached your website. Nobody has seen your product page. They are still inside Facebook, and your report already says they clicked.
The result is predictable. Put a collection ad next to your single-image ad, and the collection ad will often show a higher click-through rate and a cheaper cost per link click. It looks like a winner. Some of that is real interest. Some of it is simply that opening the storefront counts as a click, when on a normal ad the click only counts once someone heads for your site.
The numbers that tell you the truth are the ones measured at your door:
- Outbound clicks counts only clicks that take someone off Facebook's properties. That is the shopper actually heading to your site.
- Landing page views counts the times someone actually loaded your page. It needs the Meta Pixel installed, and it is the closest thing to "they arrived."
- Purchases and cost per purchase are, as always, the numbers that decide.
Meta also gives you two metrics that exist only for this format, and they are useful for diagnosing the storefront itself. Instant Experience view time is the average number of seconds people spent inside it. Instant Experience view percentage is how much of it they saw on average, based on how many components they scrolled through. If people open the storefront and bail after a few seconds having seen very little, the problem is the storefront, usually a cover that promised something the products underneath do not deliver. If they browse deep and still do not click out, the problem is the products or the prices, not the format.
None of this is special to collection ads, but the gap between the headline number and the real one is wider here than anywhere else in your account. We have written about which four numbers actually decide a Meta ad and about why the sales Meta reports rarely match your bank account. Both apply here, only more so.
Test It Against Your Best Ad, Not Against Nothing
The worst way to try a collection ad is to launch a brand-new campaign around it, watch the click-through rate, and call it a success. You would be comparing it to nothing, on a metric that flatters it.
The better way is to put it where your current winner already runs. Add it as one more ad in an ad set that is already working, using the same audience, the same budget and the same optimization. Let Meta split delivery between the collection ad and the ads it is already competing with. Then compare on what matters:
- Cost per purchase. Did it buy customers at the same price, or cheaper?
- Average order value. Browsing a storefront sometimes leads to bigger baskets, because shoppers see more of the range. Sometimes it does not. Check your own numbers rather than assuming.
- Outbound clicks and landing page views, not link clicks, if you want to compare how much traffic it actually sends.
Give it enough spend to reach a real number of purchases before you judge it, the same as any other new ad. And be honest about what a loss means. If the collection ad loses to your single-image ad, that is useful information about your catalog and your buyer, not a failure of the format. Some stores sell one thing people already know they want, and for them the short path is the right path.
If it wins, the storefront is worth maintaining. Refresh the cover when it tires, keep the product set clean as your catalog changes, and check the storefront metrics now and then to see whether people are still scrolling.
A Storefront Is Only as Good as What You Put on the Shelf
Collection ads are not a trick. They are a reasonable answer to a real problem: a shopper who taps your ad on a phone deserves to see your products before they see a loading screen. Used for the right catalog, with a cover that sells the reason and a product set you would be proud to show, they can carry that shopper from curious to chosen before they ever reach your site.
Used carelessly, they will show your gift cards to strangers and report clicks that never happened. The difference is the setup, and the setup is an afternoon's work.
This is the kind of thing our Meta ads team sets up and tests for online stores every week. If you would rather spend that afternoon running your business, we are glad to take it. And if you would rather build it yourself, everything you need is above. Either way, go take a look at where your ad taps are actually landing.
