Weekly ecommerce tips, deals & news.
Retargeting is advertising aimed at people who have already visited your store. Instead of finding new strangers, you follow up with someone who showed interest and left.
Think of a shop assistant who remembers you looked at a jacket last week. They greet you differently from a first-time visitor. That memory is the whole idea.
Most visitors leave without buying. Retargeting exists because that first visit rarely ends in a sale.
A small piece of code on your site records that someone visited. That record goes into an advertising audience.
When the same person browses elsewhere, the ad platform recognises them. Your ad appears instead of somebody else’s.
The recognition used to rely on third-party cookies. It increasingly relies on logged-in accounts and server-side signals instead.
That shift matters for planning. Audience sizes are smaller than they were, and they take longer to build.
Static retargeting shows the same creative to everyone in the audience. It is usually a brand message or a general offer.
Dynamic retargeting shows the exact products someone viewed. It pulls images and prices straight from your catalogue.
That requires a properly structured product data feed. AdTribes has a walkthrough for creating a Google remarketing feed.
Feed quality decides ad quality here. A missing image or stale price becomes a broken advert.
Most store owners assume showing the exact viewed product must work best. The research says otherwise.
A field experiment in the Journal of Marketing Research found dynamic ads less effective than generic equivalents. That was the average result across the study. It is the opposite of the industry assumption.
The effect reverses for people further along in deciding. Once someone has been reading reviews, the specific ad stops underperforming.
So match the ad to the stage, not the click. Early browsers respond to the brand, and late deciders respond to the product.
Treating all past visitors as one audience is the most common waste. Their purchase intent varies enormously.
Someone who bounced off the homepage in four seconds is barely a prospect. Someone who reached the payment step almost bought.
Build tiers by depth of visit and spend accordingly. Cart abandoners justify a real budget, whereas casual browsers do not.
Exclude recent buyers as a matter of routine. Advertising a product to the person who just bought it wastes money and irritates them.
Build a separate audience for them instead. Past buyers respond well to accessories and refills, just not to the thing already in their hallway.
Retargeting has a reputation problem, and it is earned. Everyone has been followed around the internet by one pair of shoes.
Cap how often each person sees your ad. A handful of impressions per week is plenty for most stores.
Set an end date on membership too. Chasing someone for ninety days after a single visit is rarely productive.
Watch the click-through rate decay curve. When it flattens, the audience has stopped noticing you.
If you only build one audience, build this one. Cart abandoners are the closest thing to a queue of near-buyers.
Cart abandonment averages 70.22% across e-commerce. That is a very large pool of people who nearly paid.
Many of them left for fixable reasons. Unexpected shipping costs and slow checkouts account for a great deal of it.
So the ad should answer the objection, not just repeat the product. Free delivery or a returns promise often does more than a discount.
Retargeting depends on tracking, and tracking now depends on permission. That reshaped the tactic considerably.
Visitors who decline cookies cannot be added to an audience. Banner design has a measurable effect on how many do decline.
One study found only 11.8% of consent banners met minimal legal requirements. Fixing yours is worth doing regardless of the audience cost.
Data customers hand over directly is unaffected. Zero-party data and email lists keep working when cookies stop.
Imagine a WooCommerce store called Selby Cycles, selling mid-range road bikes and accessories. Bikes are a considered purchase, so nobody buys on a first visit.
Selby starts with one audience containing every visitor from the last 30 days. Everyone sees a dynamic ad for whatever bike they viewed.
Results are mediocre and the cost per sale is high. Existing customers keep seeing bikes they already own.
Complaints appear on social media about the ads being everywhere. That is a frequency problem, not a targeting one.
They split the audience into three tiers. Cart abandoners, product viewers, and shallow visitors each get their own treatment.
Cart abandoners see the specific bike with a delivery reassurance. These people were close, so the detail helps.
Shallow visitors see a brand ad about their fitting service instead. Nothing product-specific, because they have not decided anything yet.
Recent buyers are excluded entirely, and frequency is capped at five per week. Membership expires after 21 days.
Total ad spend falls, because the shallow tier gets a much smaller budget. Sales attributed to retargeting rise.
Their ROAS improves mainly through removing waste. Nothing about the creative got cleverer.
Site-wide conversion sits near the 2.03% benchmark, while the cart abandoner tier converts several times higher. That gap is the argument for segmenting.
The complaints stop as well, which nobody measures but everyone notices. Capped frequency turns a pursuit back into a reminder.
Selby also tags every ad link with UTM parameters. That gives them a reading independent of the ad platform.
Both chase people who did not buy. The difference is whether you know who they are.
Retargeting works on anonymous visitors. You never learn their identity, and you pay per impression or click.
Email needs an address, so it only reaches people who gave you one. Sending costs almost nothing by comparison.
An abandoned cart email usually outperforms an equivalent ad. It arrives in a personal space and costs a fraction as much.
Use ads to reach the anonymous majority and email for the identified minority. Wishlist remarketing sits neatly in the second group.
Match it to your buying cycle rather than a default. Cheap impulse items justify a few days, while considered purchases justify weeks.
Beyond about a month, returns drop sharply for most stores. Long windows mostly inflate spend.
Watch what happens to customer intent over time. Someone who browsed a month ago has usually bought elsewhere or moved on.
Only once you have enough traffic to build a usable audience. Below a few thousand monthly visitors, the lists are too small to work.
Email is usually the better first investment. It costs less and you keep the list whatever the ad platforms do next.
Fix your product pages before either one. Sending paid traffic back to a weak page just costs more per failure.
Be sceptical of the platform’s own reporting. It naturally claims credit for sales that were already going to happen.
Run a holdout group that sees no ads and compare. Pair that with a sensible attribution model rather than last click alone.
Retargeting works best when it behaves like a polite reminder rather than a pursuit. Segment by how close someone came to buying and cap how often they see you. Resist assuming the exact product ad is always right. Measure it against a holdout group, because the platforms will happily take credit for sales you had already won.
Copyright © StoreOwnerTips.com. All Rights Reserved.
