Most people who visit an online store leave without buying. Some were just browsing, but many were interested and got distracted, wanted to compare prices or weren't sure about shipping. Meta retargeting shows ads on Facebook and Instagram to people who already interacted with your store, so you can answer their doubts and bring them back. Done well, it's one of the most efficient parts of an ad budget. Done carelessly, it shows the same ad twelve times to the same person and takes credit for sales that would have happened anyway. This guide shows you how to set it up properly for a small online store.
In short: Build custom audiences from your pixel (product viewers, add to carts, checkouts) and from people who engage with your pages and videos. Speak to what stopped each group from buying, exclude recent buyers, keep frequency under control, and judge retargeting by extra sales, not by its own reported return. With a small budget, a separate retargeting campaign only makes sense once your store gets enough visitors.
How Meta retargeting works
Retargeting relies on custom audiences: lists of people Meta can match because they did something you can measure. The main sources are:
- Website visitors, tracked by the Meta pixel and the Conversions API: who viewed a product, added to cart, started checkout or bought. Meta's help article on creating a custom audience from website events shows the steps.
- Engagement on Facebook and Instagram: people who watched your videos, interacted with your page or profile, or opened a lead form.
- Customer lists: emails or phone numbers of past customers that you upload (with the right consent from your customers).
- Catalog activity: if your product catalog is connected, catalog ads can automatically show each person the products they viewed.
Everything starts with tracking. If your pixel doesn't record product views and add to carts correctly, your audiences will be empty or wrong. Check the Meta pixel and Conversions API setup first.
The retargeting audiences worth building
Not every visitor is equally warm. Group them by how close they got to buying, and give each group a message that fits:
| Audience | Look-back window (starting point) | What stopped them | What to show |
|---|---|---|---|
| Viewed a product, didn't add to cart | 14-30 days | Not convinced yet, comparing | The main benefit, real reviews, a demo video |
| Added to cart, didn't buy | 7-14 days | Price, shipping cost, timing | Shipping and returns info, the product they left |
| Started checkout, didn't buy | 3-7 days | Last-minute doubt, payment hassle | Reassurance: secure payment, delivery date, support |
| Engaged with your video or profile | 30-90 days | Doesn't know the products yet | Best-sellers, a clear offer |
| Bought in the last 30-180 days | Depends on your product | Nothing; they're customers | Exclude them, or cross-sell later |
The windows are starting points, not rules. A $15 impulse buy is decided in hours; a $900 sofa may take weeks. Website audiences can look back up to 180 days, and you can shorten them to fit your buying cycle.
Always exclude recent buyers from ads that push the product they just bought. It wastes money and annoys customers. If you sell consumables (coffee, skincare, pet food), build a separate "time to reorder" audience instead, timed to when they usually run out.
What to say at each stage
Retargeting copy should feel like a helpful shop assistant, not a stalker. Some examples for a made-up store selling leather bags:
- Product viewers: "Full-grain leather, stitched by hand, and it gets better with age. See why people keep theirs for years." (Only quote reviews you actually have.)
- Cart abandoners: "Your bag is still waiting. Free shipping over $80 and free returns for 30 days."
- Checkout abandoners: "Questions before you order? Delivery takes 3-5 days and our team answers within a few hours."
- Engaged video viewers: "From cowhide to finished bag in 6 steps. Shop the collection."
Avoid copy that makes people feel watched: "We saw you looking at…" is creepy. Talk about the product and the doubt, not about their browsing. For more examples, see our guide to Meta ad copy examples.
Retargeting in an Advantage+ world
Meta's automation has changed how retargeting works. In Advantage+ sales campaigns, Meta already shows ads to warm people when it predicts they'll buy, and your audiences act as suggestions rather than strict fences. So you have two practical approaches:
- Let one campaign do both. For small budgets, a single sales campaign that reaches new and returning people is often enough. Your job is to include ads that work for both: a strong introduction and a reassurance ad.
- Run a separate retargeting ad set with Advantage+ audience turned off, aimed at your custom audiences, and exclude the same audiences from your prospecting. This gives you control over messages and budget, at the cost of a smaller audience.
Our guide to Advantage+ sales campaigns explains the trade-offs in more detail.
How much to spend on retargeting
Retargeting audiences are small, so they can't absorb much budget. A practical way to size it:
- Estimate the audience. Say your store gets 6,000 visitors a month and 600 people add to cart. A 14-day add-to-cart audience might have around 300 people that Meta can match.
- Decide how often they should see an ad. Two to three times a week is plenty for most products.
- Do the math. If a thousand impressions cost around $15 (a hypothetical CPM) and you want 300 people to see an ad about 3 times a week, that's 900 impressions a week, roughly $14 a week, or $2 a day.
That's the point: with a few thousand visitors a month, a dedicated retargeting campaign can only spend a few dollars a day before it starts hammering the same people. If your total budget is $30 a day, giving retargeting $15 would push frequency up fast. Many small stores do better by spending most of the budget on reaching new buyers and keeping retargeting lean. If you're still planning your overall budget, see how much to spend on Meta ads.
Watch frequency closely
Frequency is the average number of times each person saw your ad in the period. Because retargeting audiences are small, frequency climbs fast. When the same person sees the same ad too often, results fall and comments get grumpy.
There's no official limit, but as a working rule, check retargeting ad sets weekly: if frequency is above 3 to 4 in a week and click-through rate is falling, refresh the ads or lower the budget. Our guide to ad fatigue and frequency covers the signs and fixes.
Don't let retargeting take credit for free sales
Retargeting reports often look amazing: a high "sold per $1 of ads" and a low cost per sale. Part of that is real. Part of it is people who were coming back anyway, saw an ad on the way, and got counted as a retargeting sale within Meta's attribution window.
Ways to keep yourself honest:
- Compare total store sales, not only Meta's reported sales, before and after you change retargeting spend.
- Pause retargeting for a week (if you can afford the test) and see how much total revenue actually changes.
- Use Meta's lift or A/B test tools when your account is large enough.
- Check new versus returning customers in your store's reports.
Our guide to the Meta ads metrics that matter explains how to read reported results against your real orders.
What if…? Common retargeting problems
What if the retargeting ad set barely spends? The audience is probably too small. Lengthen the look-back window, combine audiences (viewers plus add to carts), or fold retargeting into your main campaign.
What if frequency hits 8 in a week? Lower the budget, add new ads, or widen the audience. A small audience can't take a big budget.
What if your custom audiences show "too small" or "not ready"? Check that the pixel is firing the right events and that enough people have visited. Very new stores may need a few weeks of traffic first.
What if you sell health products or other sensitive categories? Meta restricts some uses of website data and some audience options in sensitive categories, and ads must never imply that you know something personal about the viewer. Keep the copy about the product, not the person.
What if it's Black Friday? Warm audiences matter most in big sales periods, because people browse for weeks before buying. Build them early and plan your messages; see our Black Friday Meta ads playbook.
How Meta Ads Copilot Max helps with retargeting
Meta Ads Copilot Max reads your campaigns, ad sets and ads from Ads Manager and flags problems that hit retargeting hardest: high frequency (it suggests rotating creatives or widening the audience), budgets that don't spend (often a sign the audience is too small), and cost per sale above your margin. Results follow each campaign's objective, so a retargeting ad set optimized for purchases is judged on purchases.
When you need fresh ads for a warm audience, Copilot writes three versions from your product page, keeps your banned words out, and Claude in Chrome (required) builds them in Ads Manager, paused, waiting for your approval. You can also add campaigns to an exclusion list so Copilot and Claude never touch them.
You can try Meta Ads Copilot Max free for 7 days. Start your free 7-day trial and get your key by email.
Frequently asked questions
What is retargeting on Meta?
Retargeting means showing ads to people who already interacted with your business: website visitors, people who added to cart, video viewers, page engagers or past customers. On Meta it's done with custom audiences built from your pixel, your pages and your customer lists.
How long should my retargeting window be?
Match it to how long people take to decide. For low-price items, 7 to 14 days for cart abandoners is a reasonable start. For expensive products, 30 days or more can make sense. Website audiences can look back up to 180 days.
Should I run a separate retargeting campaign?
Only if your store gets enough visitors to build audiences of a useful size. With a small budget and a few thousand monthly visitors, a single sales campaign with ads that work for new and returning people is often simpler and just as effective.
Why does my retargeting campaign show a much better return than prospecting?
Warm audiences buy more easily, and some of them would have bought without the ad. Retargeting also depends on prospecting to fill the audience in the first place. Judge both together, using total store sales.
Do I need catalog ads for retargeting?
No, but they help stores with many products, because they show each person the items they viewed. For a store with a few products, regular image or video ads aimed at your custom audiences work fine.
