Retargeting
Retargeting is paid advertising shown only to people who have already interacted with you — visited a page, used the app, or appeared on a list you uploaded. A tag on your site drops those visitors into an audience, and your ads are then served to that audience instead of to strangers. It is the cheapest-looking line in a media plan and the one an early-stage startup is least likely to be able to run.
What has to be true before an ad can be retargeted
Retargeting is three moving parts, and founders usually picture only the third. A tag or pixel fires on your site and records that a browser was there. Those records accumulate into an audience list. A campaign then targets that list — but only once the list is big enough for the platform to serve against it, which is where the plan tends to stop.
- The platforms enforce a floor. Google Ads states that a remarketing list needs at least 100 active visitors or users from the last 30 days before it will serve on Display, Search or YouTube. Uploaded customer lists need 100 active matched users for lists uploaded or refreshed after 1 February 2024, and 1,000 for older ones. Meta and the other networks set floors of their own; check the one you intend to use rather than assuming Google's number travels.
- "Active in the last 30 days" is a rolling window, not a total. A list is not a bucket that fills up. People age out of it, so a list that cleared the floor in a good week can drop back under it in a quiet one and quietly stop serving.
- The list is built by JavaScript. Anything that does not run your tag never enters the audience — which is a real limit, and occasionally a favour.
- Consent and browser rules shrink it again. Tracking-prevention defaults and consent choices mean the audience you can actually build is smaller than the traffic you can see in your own logs.
- Retargeting is a channel, not a strategy. It can only re-reach demand that something else created. If nothing is creating demand, there is nothing to re-reach.
Why a young startup usually cannot fill the list
A hundred people in 30 days sounds trivially achievable until you notice what it is counting: not pageviews, not sessions, but distinct identifiable users who are still identifiable when the ad would serve. Most pre-product-market-fit sites do not clear it, and the ones that do often clear it with the wrong hundred.
- Much of your traffic is not a person at all. When we audited our own origin logs over a three-hour window, the large majority of requests came from datacentre ranges with spoofed or empty user agents. The tag-based nature of a retargeting list is the one place that helps you: the same JavaScript gap that hides bots from analytics also keeps most of them out of the audience. See how much of your website traffic is bots for how to tell the difference.
- A bounce is on the list too. Our own paid-landing telemetry recorded a signup who left 4.06 seconds after arriving. Someone who was gone in four seconds carries no more intent on their second visit than on their first, and you are now paying twice to reach them.
- Small lists are expensive per outcome. Retargeting looks cheap because the click-through rate is high; on a list of a few hundred, high rates are calculated on numbers too small to decide anything, and you learn nothing you can act on.
- Frequency turns against you fast. The smaller the audience, the more times each person sees the same ad on the same budget. At a few hundred people, a modest daily spend becomes an irritant rather than a reminder.
What to do with the signal instead, early on
The intent data is real even when the audience is too small to advertise to. It just pays better somewhere other than a second impression. Our own account runs no retargeting campaign at all, and the deliberate alternative has been more useful than a thin audience would have been.
- Feed the bidding rather than chasing the visitor. We upload offline conversions back to Google Ads — a click-keyed action that fires when a signup becomes genuinely engaged, kept as a secondary, observation-only conversion so it informs reporting without hijacking bidding. The auction learns which clicks became real users, which improves who arrives next rather than paying again to re-reach whoever already left.
- Fix the landing page before you buy the second visit. If the first visit failed for a reason you can name, the second one will fail the same way. Retargeting spend is the most expensive possible way to avoid rewriting a headline.
- Know your consideration gap. Retargeting earns its place when there is a genuine delay between first look and decision — a considered purchase, a demo request, a plan comparison. For an impulse signup it mostly buys impressions you would have had anyway.
- Keep the labels clean now, so the option exists later. Consistent UTM parameters and a working conversion setup are what make a retargeting decision measurable the day your audience finally clears the floor.
- Decide with [ROAS](/glossary/return-on-ad-spend), not click-through rate. Retargeting always wins on click-through rate. That is the metric it is least entitled to be judged on, because the audience was selected for already knowing you.
FAQ
- Is retargeting the same as remarketing?
- In practice, yes. Google's own products say remarketing, most of the ad industry says retargeting, and some people reserve remarketing for email follow-up to an existing list. Nobody will misunderstand you either way — but read whichever word your platform uses in its own documentation, because the eligibility rules attach to the product name, not to the industry term.
- How many website visitors do I need before retargeting is worth it?
- Google's technical minimum is 100 active users in the last 30 days for Display, Search and YouTube. That is the point at which an ad can serve, not the point at which it is worth serving. A useful working threshold is higher: enough traffic that the list stays above the floor in a slow week, and a product with a real gap between first visit and decision.
- Does retargeting still work now third-party cookies are restricted?
- It works, with a smaller and shorter-lived audience than it had a few years ago. The mechanics have moved toward first-party tags, uploaded customer lists and platform-side modelling. The practical effect for a small advertiser is that lists take longer to fill and expire faster, which sharpens the same question of whether you have the volume to bother.
- Should I retarget people who bounced in seconds?
- Exclude them if your platform lets you. A visitor who left almost immediately did not fail to convert for lack of a reminder, and including them inflates your audience with the people least likely to return. If excluding them leaves you under the serving floor, that is the honest answer to whether you should be retargeting yet.
An AI growth team that runs this for you
AgentCeres is a managed AI marketing team — you approve what ships. 14-day free trial, from $39/month.