Ecommerce Retargeting: Stop Paying to Reach People Who Were Never Going to Buy

Ecommerce retargeting is the practice of serving ads to people who have already visited your site, viewed a product, or started a checkout, with the goal of pulling them back to complete a purchase. Done well, it recovers revenue that would otherwise disappear. Done badly, which is most of the time, it burns budget on audiences who were never going to convert and annoys the ones who were.

The difference between those two outcomes is not your ad creative. It is how you define your audiences, how you sequence your messaging, and whether you are honest about what retargeting can and cannot do for your business.

Before getting into mechanics, it is worth grounding this in the broader funnel context. Retargeting sits at the bottom of the purchase funnel, where intent is highest and the cost of a poor decision is also highest. If you want to understand how retargeting fits into the full conversion architecture, the high-converting funnels hub covers the complete picture from awareness to purchase.

Why Most Ecommerce Retargeting Wastes More Than It Recovers

I have audited a lot of paid media accounts over the years, and retargeting is consistently the area where I find the most waste dressed up as performance. The numbers look good in the platform dashboard. ROAS is high. Click-through rates are healthy. And yet when you pull back and look at the business, the retargeting spend is not actually moving the revenue needle in any meaningful way.

The structural problem is this: retargeting platforms are very good at showing you conversions that happened after a retargeting impression. They are not good at telling you whether that conversion would have happened without the impression. When someone adds a product to their cart, leaves, and then comes back two days later to buy, the retargeting campaign will claim that conversion. But a meaningful portion of those people were going to come back anyway. They were just thinking it over.

This is not a new observation. Semrush’s breakdown of the conversion funnel makes the point that bottom-of-funnel activity tends to capture demand rather than create it. Retargeting is almost entirely a demand capture mechanism. That is not a criticism, it is just a fact that should shape how you budget for it and how you report on it.

When I was running iProspect and we were scaling the performance team, I made a point of running incrementality tests on every major retargeting programme before recommending budget increases. In roughly half of those cases, the incremental lift from retargeting was materially lower than the platform attribution suggested. Not zero, but lower. Sometimes significantly lower. The clients who were willing to look at that data honestly ended up with better-structured budgets. The ones who preferred the dashboard numbers kept overspending on retargeting at the expense of prospecting.

How to Build Retargeting Audiences That Actually Mean Something

The first thing to fix is audience segmentation. Most ecommerce accounts have one or two retargeting audiences: site visitors and abandoned carts. That is not a strategy, it is a default setting. The signal quality across those broad buckets varies enormously, and treating them the same way guarantees you will overspend on low-intent browsers and underspend on high-intent prospects.

Here is how I think about audience tiers for ecommerce retargeting:

Tier 1: Checkout abandoners

These are your highest-intent audience. They got to the point of entering payment details, or at least reaching the checkout page, and stopped. The reasons vary, from distraction to price sensitivity to genuine second thoughts, but the intent signal is strong. This audience deserves your highest bids, your most direct creative, and your fastest follow-up. Window of 3 to 7 days. After that, the signal degrades quickly.

Paid ads are only one part of this recovery sequence. The email channel often outperforms display for checkout abandonment, and the subject line strategy matters more than most people realise. If you are building out your abandoned cart recovery, the work on highest-performing subject lines for abandoned cart emails is worth reading alongside your retargeting setup, because the two channels should be coordinated, not competing.

Tier 2: Product page viewers with session depth

Someone who viewed a single product page for 8 seconds and bounced is not the same as someone who viewed three product pages, read reviews, and spent four minutes on the site. Both will end up in your “product viewer” audience if you are not filtering by session depth or page count. Separate them. The high-engagement segment is worth retargeting. The low-engagement segment is largely noise.

Tier 3: Category browsers

These people have shown category-level interest but not product-level intent. They are earlier in the decision process. Your retargeting message for this group should not be “buy now.” It should be doing the job that upper-funnel content does: building familiarity, demonstrating value, and earning the next click. Treat them like warm prospects, not hot leads.

Tier 4: Past purchasers

This audience is often lumped into retargeting when it belongs in a separate CRM-led retention programme. If someone bought from you 90 days ago, they do not need a retargeting ad reminding them your site exists. They need a reason to come back, which is a different message, a different channel mix, and a different budget allocation. Do not let your retargeting campaign take credit for reactivating customers who were already in your lifecycle programme.

Suppression Lists: The Part of Retargeting Nobody Talks About

If audience building is the offensive side of retargeting, suppression is the defensive side, and most accounts do it badly. Suppression lists define who you exclude from your retargeting audiences. Getting this wrong means you spend money on people who have already converted, people who are clearly not going to convert, and people whose experience of your brand you are actively degrading by following them around the internet.

The suppression list you should always have running: recent purchasers. Anyone who has completed a transaction in the last 14 to 30 days should be excluded from your standard retargeting campaigns. This sounds obvious. It is not universally done. I have seen accounts where the retargeting campaign was serving ads to people who had bought the product three days earlier, effectively wasting money and creating a poor post-purchase experience at the same time.

Beyond recent purchasers, consider suppressing anyone who has visited your site more than a set number of times without converting. At some point, frequency is not helping. If someone has seen your retargeting ads 15 times and has not clicked, they have made their decision. Continuing to serve them ads costs you money and trains them to ignore your brand. There is a version of this that applies to broader demand generation strategy too. HubSpot’s overview of demand generation makes the point that not every visitor is a viable prospect, and your retargeting budget should reflect that reality.

Frequency Caps: The Number Most Accounts Get Wrong

Frequency capping is one of those settings that gets set once during campaign setup and then never revisited. That is a mistake. The right frequency cap depends on your audience tier, your creative refresh rate, and your window duration. A checkout abandoner might tolerate seeing your ad three times in 48 hours. A category browser who visited once should not see the same ad seven times in a week.

A rough framework I use as a starting point: tier 1 audiences, up to 3 impressions per day for the first 3 days, then step down to 1 per day for the remainder of the window. Tier 2 audiences, 1 to 2 impressions per day. Tier 3, no more than 1 per day and consider whether display is even the right channel for this group.

These are starting points, not rules. Test against them. But the principle holds: more impressions do not equal more conversions past a certain threshold, and the cost of overexposure is not just wasted spend. It is brand perception damage that does not show up in your retargeting dashboard.

Creative Strategy for Retargeting: What to Say and When

Retargeting creative is frequently the last thing people think about and the first thing they blame when campaigns underperform. The brief for retargeting creative is actually quite specific: you are not trying to introduce your brand, you are trying to resolve an objection or remove a friction point for someone who already knows who you are.

For checkout abandoners, the most effective creative tends to address the most common reasons people abandon: price uncertainty, delivery concerns, or just distraction. A clean product image, a reminder of what they left behind, and a clear path back to checkout. If you have a time-limited offer, this is the right place to use it, but use it sparingly. Training your audience to wait for a discount is a long-term margin problem.

For product viewers, social proof tends to work well: reviews, ratings, or a simple statement about how many people have purchased. You are not closing the sale yet. You are reducing the perceived risk of the next click.

For category browsers, think of your retargeting creative as content rather than advertising. What does someone in the early stages of this purchase decision actually need to know? That is your message. This is particularly relevant for categories with longer consideration cycles, where the decision is not made in a single session.

One thing I would push back on: the assumption that dynamic product ads are always the right format for retargeting. They are efficient, but they are not always effective. A dynamic ad showing someone the exact product they viewed is not always the most persuasive message. Sometimes a broader brand statement, a category benefit, or a piece of social proof will outperform a personalised product feed. Test both before committing your entire retargeting budget to dynamic formats.

Platform Mix: Where Retargeting Actually Works

The platform question in retargeting is more nuanced than most accounts treat it. Meta, Google Display, YouTube, and programmatic all have different audience quality characteristics, different attribution behaviours, and different creative requirements. Running the same retargeting campaign across all of them and comparing ROAS numbers is not a meaningful analysis.

Meta retargeting has become more difficult since the iOS privacy changes reduced the quality of pixel-based audiences. The workaround, using first-party data lists and server-side events, requires more technical investment but produces better results. If your Meta retargeting is still running on basic pixel audiences without any first-party data integration, you are working with a degraded signal.

Google’s retargeting options, particularly customer match and similar audiences built from first-party data, tend to perform well for higher-ticket items where the consideration cycle is longer. For lower-ticket impulse categories, the economics often do not support the CPM rates on Google Display.

For brands operating in a direct-to-consumer model, the platform mix question is closely tied to margin structure. If you are thinking through the broader question of how DTC economics affect your paid strategy, the comparison of direct to consumer versus wholesale is a useful frame for understanding where retargeting investment makes sense relative to your channel margin.

The paid acquisition benchmarks for DTC brands are also worth reviewing before you set retargeting budgets, because the cost-per-acquisition norms vary significantly by category and average order value, and what looks like a healthy retargeting ROAS in one category can be a loss-making number in another.

Attribution Honesty: What Your Retargeting Numbers Are Actually Telling You

This is where I tend to have the most difficult conversations with clients. Retargeting attribution is almost always overstated, and the reason is structural: retargeting campaigns target the people most likely to convert anyway, and then take credit when they do.

The only way to know the true incremental value of your retargeting is to run a holdout test: take a portion of your retargeting audience, do not serve them ads, and measure the conversion rate difference between the exposed group and the holdout group. The difference is your true incremental lift. Everything else is correlation dressed up as causation.

Most platforms will not encourage you to run holdout tests, for obvious reasons. But the ones that do support it, Meta’s conversion lift studies and Google’s geo-based incrementality testing among them, are worth using before you scale retargeting spend significantly. The results are often humbling. They are also more useful than any dashboard metric.

I judged the Effie Awards for several years, and one of the things that process reinforces is the difference between activity that looks like it is working and activity that demonstrably moved a business outcome. A lot of retargeting sits in the former category. That does not mean it has no value. It means the value is smaller than the platform claims, and your budget allocation should reflect that.

Forrester’s analysis of lead nurturing makes a related point about bottom-of-funnel investment: the assumption that more touchpoints always improve conversion rates is not supported by the evidence. The same logic applies to retargeting. More impressions, more platforms, more creative variants do not automatically produce better outcomes. Precision does.

Retargeting Across Business Models: CPG, DTC, and Marketplace Contexts

Retargeting strategy is not one-size-fits-all, and the business model you are operating in has a significant effect on how you should approach it.

For CPG brands selling through ecommerce channels, the retargeting logic is complicated by the fact that you may not own the conversion point. If someone views your product on your brand site and then converts on a retailer’s site, your retargeting pixel does not see that conversion. This is a structural measurement problem, and it affects how you should interpret your retargeting performance data. The CPG ecommerce strategy piece covers this in more depth, including how to think about attribution across mixed-channel environments.

For brands going through an ecommerce platform migration, retargeting audiences are one of the most frequently overlooked casualties. When you move platforms, your pixel history resets, your audience pools drain, and your retargeting performance will dip for weeks or months while the new pixel accumulates data. Build that into your migration timeline and your performance expectations.

For financial services and marketplace businesses, retargeting operates under additional regulatory constraints, and the audience segmentation logic is different. The considerations around financial marketplace positioning include compliance requirements that affect what you can say in retargeting creative and to whom you can say it. This is an area where the default ecommerce retargeting playbook needs meaningful adaptation.

Measuring Retargeting Performance Without Fooling Yourself

Given everything above about attribution, what metrics should you actually use to evaluate retargeting performance?

Start with view-through conversion windows. Most platforms default to a 1-day or 7-day view-through window, meaning a conversion is attributed to retargeting if someone saw your ad within that window, even if they never clicked it. This inflates retargeting performance significantly. I would recommend setting view-through windows to 1 day maximum for retargeting, and relying primarily on click-based attribution for your primary performance metric.

Next, look at your retargeting audience overlap with your organic and direct traffic. If a large proportion of your retargeting conversions are coming from people who were already visiting your site directly, you are paying to intercept traffic that was coming back anyway. This is the clearest sign that your retargeting spend is over-indexed.

The metric I find most useful is cost per incremental conversion, which requires a holdout test to calculate properly. Short of that, look at your overall business conversion rate against your retargeting spend as a proportion of total paid budget. If retargeting is consuming 40% of your paid budget but your overall conversion rate has not moved materially when you have scaled it up or down, that is a signal worth investigating.

HubSpot’s guidance on conversion optimisation makes the point that improving on-site conversion rates often has a larger impact on revenue than increasing retargeting spend. If your site converts at 1.5% and your competitor’s converts at 3%, no amount of retargeting will close that gap. The funnel fix comes before the retargeting fix.

Retargeting is one component of a broader funnel architecture. If you are building or auditing that architecture from top to bottom, the full framework on high-converting funnels covers how each stage connects and where retargeting sits relative to the other investment decisions you need to make.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions