Most people who visit your online store leave without buying, and plenty of them meant to come back. Retargeting ads for ecommerce stores exist to win that second chance. They show past visitors the products they viewed while those people scroll Facebook, Instagram or YouTube. Done well, retargeting recovers sales at a sensible cost. Done carelessly, it pays for orders that were coming anyway and annoys everyone else.
Yet many stores set it up once and never revisit it. This guide shows you how to build audiences by intent, cap frequency, segment by cart value and measure the sales your ads actually caused.
How Retargeting Ads Work, and Why the Reports Flatter Them
It starts with tracking code. The Meta Pixel and the Google tag record events such as product views, add to cart, checkout started and purchase. Each platform then builds lists from those events, and you decide which list sees which ad. Dynamic ads go further and pull the exact products a shopper viewed from your catalog. On Google, that feed lives in Merchant Center, and our guide to running Google Shopping campaigns covers setting it up.
Here’s the catch. The people on these lists already liked your products enough to visit, and many would return on their own. However, the platforms still count their orders as conversions. Depending on your attribution settings, they may even credit sales to people who saw an ad but never clicked. So retargeting almost always looks brilliant in Ads Manager, whether or not it’s truly profitable.
Build Your Audiences by Shopping Intent
A single “all visitors” list treats a quick home page glance the same as a half-finished checkout. Instead, sort visitors by the deepest step they reached:
- Product viewers never added to cart. Remind them lightly with bestsellers or reviews.
- Cart adders stopped before checkout. Show the exact products, plus shipping and return details.
- Checkout starters are your hottest and smallest group, so keep their window short.
- Past buyers belong in a separate campaign for add-ons or new arrivals. Often, automated email flows for repeat customers reach them for far less.
Match the window to your buying cycle
A shopper choosing a $25 phone case usually decides within a day or two. By contrast, a family pricing an $1,800 sectional in Paramus might compare stores for weeks. So check how long orders typically take after a first visit, and set windows to match. A reasonable starting test is 3 to 7 days for checkout starters and 14 to 30 days for product viewers.
Exclusions matter as much as targeting
First, remove anyone who bought. Nothing irritates a customer faster than an ad for the blender they unboxed yesterday. That means your purchase event must fire reliably, and every audience must exclude recent buyers. Also consider leaving out visitors who bounced within seconds.
Frequency Caps: How Often Your Retargeting Ads Should Run
Frequency is the average number of times one person sees your ads in a period. Too low, and shoppers forget you. Too high, and your brand feels like a stalker. Google lets you cap impressions per person in Display and video campaigns, although some automated campaign types manage frequency for you. Meta offers hard caps only with certain objectives, such as reach. So for most sales campaigns there, you control frequency through budget, audience size and window length.
That budget link trips up many small stores. Say your 7-day cart audience holds 800 people and your ads cost about $10 per thousand impressions. A $50 daily budget buys roughly 5,000 impressions, or about six ads per person, every day. In other words, size the budget to the audience, not a round number.
As a starting point, aim for a couple of impressions a day on your checkout window and a few per week for product viewers. When frequency climbs while click-through rate falls, the audience is tired. Refresh the creative, shorten the window or trim the budget.
Segment Ecommerce Retargeting Ads by Cart Value
A $40 cart and a $400 cart both count as one “add to cart,” yet the second can justify far more recovery spend. If your pixel passes the cart value with each event, Meta and GA4 audiences can both filter on it.
Start with your own margins. Suppose a $400 order leaves you $150 in gross profit after product cost and shipping. Spending $30 in ads to win it back is a good trade. On the other hand, a $40 order that leaves $12 can’t carry that spend. Then build a few tiers around your average order value:
- Low-value carts: short window, small budget, a simple reminder and no discount.
- Mid-value carts: standard window, dynamic product ads, reviews and your return policy.
- High-value carts: longer window, bigger budget and ads that answer worries about delivery, warranty and returns.
On Google, you can also report order values so Smart Bidding optimizes for conversion value in each auction, instead of treating every sale as equal. That matters most when your prices range widely.
Be careful with discounts
It’s tempting to hang a coupon on every ad. However, shoppers catch on, and some learn to abandon carts on purpose and wait for the code. Instead, save incentives like free shipping for high-value carts, and test them against a version with no offer. Keep every claim true, too. The FTC’s advertising guide for small businesses is clear that you need proof before an ad runs, so “only 2 left” has to be accurate.
Ads are also only one recovery channel. Email and text reminders cost far less per touch, and our guide to abandoned cart recovery for online stores covers their timing and consent rules.
Measure Incremental Lift, Not Platform ROAS
The real question is simple: how many of these orders would you have gotten without the ads? That gap is incremental lift. In ecommerce, retargeting ads are especially prone to claiming sales they didn’t cause. After all, they follow people who were already close to buying.
The cleanest answer comes from a holdout test. You split a comparable audience in two, show ads to one group and compare purchase rates. Meta and Google both run conversion lift studies for eligible advertisers. Smaller accounts can run a rougher version:
- Geographic holdout: if you ship nationally, pause retargeting in a few matched states for four weeks, then compare sales trends.
- On and off test: pause retargeting for two or three weeks in a steady season, then compare cart recovery with the weeks around it.
Picture a gift shop in Red Bank that runs a holdout on its cart audience. The group that saw ads placed 100 orders, and Ads Manager credits the campaign with all of them. Meanwhile, the matched group that saw no ads still placed 70. The ads really caused about 30 sales, so the true return is less than a third of what the dashboard showed. That may still be profitable, but now you know. Use your store’s own order data, and rerun the test after big changes.
Keep Your Retargeting Ads Legal and Honest
Retargeting runs on personal data, so privacy rules apply. New Jersey’s data privacy law took effect in January 2025. It lets residents opt out of targeted advertising and requires covered businesses to honor universal opt-out signals, such as browser privacy settings. It applies to companies handling data on at least 100,000 state residents a year (not counting data used only to process payments), or 25,000 if they earn money selling data.
Many small stores fall below that line. Even so, post a clear privacy notice, use a consent banner that actually works, and ask your attorney whether the law reaches you. Also check platform policies before retargeting sensitive products, since Google and Meta both restrict personalized ads tied to health.
Retargeting Ads for Ecommerce: Quick Answers
Is retargeting the same as remarketing?
Mostly, yes. Google has long called it remarketing, and some marketers also use that word for email reminders.
How much traffic do retargeting ads need?
Your audience can only be as big as your traffic. With a few hundred visitors a month, your lists stay tiny, and platforms may not serve ads to them at all. In that case, spend on attracting visitors first. Main Street shops still building their site can start with our guide to launching an online store as a New Jersey retailer.
Where to Start This Week
You don’t need a new platform to fix retargeting. Just confirm that your cart and purchase events fire with order values, split your lists by intent and value, set frequency limits and plan one holdout test. Those four steps cost nothing but time.
If you’d like a second set of eyes, Samaroo Solutions is based in northern New Jersey and works with online stores across the state. Our digital advertising team can audit your tracking and build retargeting ads for ecommerce that fit your margins. When you’re ready, contact us to talk through your store.
