Retargeting is the easiest campaign in any account to launch and the easiest one to get quietly wrong. It looks like it is working because the numbers flatter it, and it feels like it is working because everyone in the business keeps seeing their own ads. At The Growth Bully, a Malta performance marketing agency, the retargeting campaign is usually the first place we look when an account reports a strong return that the bank balance does not agree with.
The failure is almost never the idea. Following up with someone who already showed interest is sound. The failure is treating every visitor as one audience, showing them one message, and leaving it running until the brand becomes wallpaper.
Why does retargeting stop working?
Because the audience is treated as a single group when it is really several. Someone who bounced off the homepage in four seconds and someone who abandoned a filled-in enquiry form get the same ad for the same number of weeks. One of them was never a buyer. The other needed a reason and got a reminder.
That is what makes retargeting feel like stalking rather than selling. The person is not annoyed that you followed up. They are annoyed that the follow-up carried no new information, so the tenth impression asked exactly what the first one asked and answered none of the reasons they hesitated.
What is the difference between retargeting and remarketing?
In practice, very little, and the distinction matters less than the platforms suggest. Retargeting usually describes paid ads served to people who visited or engaged. Remarketing often describes contacting a known list, by email or by matching that list back into an ad platform. Both are warm audience work and both fail the same way.
What matters is the source of the audience and how recent it is. A list of last week visitors and a list of customers from two years ago behave nothing alike, so they should never share a campaign, a message or a budget line.
How should a retargeting sequence be structured?
By what the person actually did, in order of intent, with a different job for each stage. The sequence is not a funnel diagram for its own sake. Each segment has one question it needs to answer, and if the ad does not answer that question it is spending money to be seen rather than to move somebody forward.
The structure that holds up across accounts, from coldest to warmest:
- Shallow visitors. Landed and left quickly. Low intent, so keep the budget small and the message about the problem rather than the purchase.
- Engaged visitors. Read a service page or several pages. They understand what you do, so show proof and the reason to choose you.
- High intent, no action. Hit a pricing, contact or booking page and stopped. Something specific blocked them. Answer the objection directly.
- Abandoned enquiry or cart. Started and did not finish. The shortest window, the highest budget share, and the most direct call to action in the whole account.
- Existing customers. A different campaign entirely, for the next purchase or the referral, and usually excluded from everything above.
Exclusions do more work here than targeting. If each segment is not excluded from the ones below it, everybody drifts into the highest budget group and the sequence collapses back into the single blunt campaign it was meant to replace. The same discipline underpins how we build warm audiences on Meta and in Google Ads.
How often should someone see a retargeting ad?
Less often than the platform will happily deliver, and with a ceiling agreed before launch rather than after a complaint. Small markets reach saturation fast, because the pool of people who visited your site in a given week is finite and the budget will keep buying impressions long after everyone in it has seen the ad several times.
The practical rule is to set a frequency cap per segment, review it weekly, and treat a rising frequency with a flat conversion count as a signal to refresh creative or shrink the window. Rising frequency is not extra pressure on the same buyer. It is the same budget buying worse and worse impressions.
How long should a retargeting window stay open?
Long enough to cover the real decision period for what you sell, and no longer. A seven day window suits an impulse purchase and starves a considered one. A one hundred and eighty day window on a low value product mostly pays to advertise to people who bought elsewhere months ago.
Set the window from your own sales cycle, not from the platform default. If enquiries usually convert within a fortnight, a month of retargeting is generous and anything past that is sentiment. If the sale takes a season, the window should stretch but the frequency has to come down to match, or you will exhaust the audience long before they are ready.
Does retargeting work on a small audience?
It works, but only if the expectations and the budget are scaled to the size of the pool. In a market the size of Malta, a warm audience is often measured in hundreds rather than tens of thousands, which means a modest daily budget can cover it many times over in a single week.
The mistake is funding retargeting as a growth channel. It converts demand that already exists, so it cannot create more of itself. When the warm audience is small, the honest answer is to spend the extra money on the cold campaigns that fill it, and let retargeting stay the cheap, disciplined layer that closes what those campaigns started. How that split gets decided is the whole subject of splitting a budget across channels.
How do you know retargeting is actually adding sales?
By checking whether the account improved, not whether the campaign reports well. Retargeting claims conversions from people who were already going to convert, which is why it almost always shows the best return in the account and almost never deserves it. The check is straightforward:
- Look at total enquiries and total cost per acquisition at account level before and after the campaign launched.
- Pause the highest spending segment for a defined period and watch what the account does, not what the campaign does.
- Confirm the conversion action being counted is a real enquiry rather than a page view or a soft engagement, which is where most of the inflation hides.
None of that is possible without measurement you trust, which is why conversion tracking comes before any warm audience work. The rest of the Meta side is covered in what actually drives results on Meta, and the reasons warm traffic still fails to close are in why leads do not convert.
Retargeting done properly is quiet. It runs on a small share of the budget, speaks differently to each group, stops when the window closes, and never has to be defended. If you want to know whether yours is closing demand or just billing for it, get a straight read on your account and what it would take to fix.