Retargeting Ads for Dropshipping: When You Actually Have Enough Traffic to Run Them

Most dropshipping accounts turn retargeting on the same day they launch a prospecting campaign, usually because a tutorial said to and Meta's custom audience only needs 100 people to go live. Technically that's true. Practically, an audience that small just means the same 150 people see the same ad four times this week, click through again, and still don't buy — because they already decided the first time. Retargeting isn't broken at that size. It's just empty.

The real floor is traffic, not Meta's technical minimum. A store needs a rolling pool of people who've already shown intent — viewed a product, added to cart, started checkout — large enough that the algorithm can find a pattern in who converts and who doesn't, rather than burning through the same faces on repeat. Below that floor, retargeting spend behaves almost randomly. Above it, it's usually the cheapest conversion in the account.

What follows: the three problems that get wrongly blamed on retargeting, a reproducible way to size your own traffic pool before you commit budget to it, how to split that pool into segments that actually deserve separate treatment, a budget-allocation grid by traffic stage, and the frequency and creative rules that are different for retargeting than for a cold campaign.

What retargeting actually recovers — and three problems that get blamed on it instead

Retargeting recovers intent that already existed. It doesn't create it. Someone who looked at a product page, didn't convert, then saw a reminder ad and bought, was already most of the way there — the ad closed a gap, it didn't open one. That's the entire mechanism, and it's also why retargeting makes a weak store look stronger than it is for a few weeks: it's harvesting demand the prospecting campaigns already paid to generate.

Because ROAS on a retargeting campaign is almost always better than prospecting — sometimes three to five times better on the same spend — sellers reach for it as the fix for problems it was never built to solve. Three show up constantly.

• A product page that doesn't close the sale: if visitors are bouncing because trust signals, pricing clarity, or shipping info are missing, retargeting just shows them the same broken page again at a markup.
• Page speed killing the first visit before it had a chance: a slow mobile load loses people before they form any real intent to retarget against — there is often nothing worth recovering yet.
• The wrong landing page for the ad's angle: if the ad promises one thing and the page sells another, the mismatch is the problem on visit one. Showing the same mismatched page again on visit two just repeats the mistake at a lower cost per impression.

The traffic floor: how big the pool needs to be before retargeting is worth a campaign

Size the pool off your own analytics, not a platform minimum. Pull the count of unique visitors over a rolling 14-day window who viewed a product page but didn't purchase — Shopify Analytics or GA4 both report this directly. That number, not Meta's 100-person technical floor for a custom audience, decides whether a retargeting campaign has anything real to work with.

A useful floor to test against: roughly 1,000 to 1,500 people in that 14-day pool before a retargeting campaign holds a stable frequency for more than a few days on even a modest budget. Below that, the math turns against you fast — spend $20/day at a typical dropshipping CPM of $10-20, and the ad reaches somewhere around 150-250 people a day. Against a pool of 1,000, average frequency crosses 3 inside the first week; against 300, it crosses 3 in under two days. The algorithm isn't failing at that point — it's run out of new people to show the ad to and is just repeating itself.

• Days to hit a frequency cap ≈ (pool size × cap) ÷ daily unique reach — plug in your own numbers before setting a budget, not a borrowed one.
• Example: pool of 1,200, cap of 3, daily reach of 200 → roughly 18 days before frequency crosses the cap. Pool of 400 at the same reach → under 6 days.
• Limitation: CPM and daily reach swing hard by niche, geography, and season, so treat this as a planning check before launch, not a number to defend after the fact.

See the ads that are winning right now

Segmenting the pool: viewed, added to cart, and started checkout don’t deserve the same ad

Lumping every non-purchaser into one audience wastes the signal retargeting is supposed to use. Someone who viewed a product once and left is a different person, with a different objection, from someone who built a cart and abandoned it at shipping cost. Treating them identically means showing a cart-abandonment message to someone who never had a cart, which reads as confusing rather than persuasive.

Three segments cover almost every dropshipping store, each with its own lookback window and message:

• Viewed, no add to cart — 14-day window: the objection is usually trust or price, not logistics. Lead with reviews, a guarantee, or a direct price comparison.
• Added to cart, no checkout started — 7-day window: narrower, because cart intent decays faster than page-view intent. Lead with urgency on stock or a time-limited offer, not a generic reminder.
• Checkout started, not completed — 3-day window: the highest-intent and smallest segment, usually lost to a shipping-cost surprise or a payment-method gap. A direct nudge showing the actual cart contents outperforms a generic ad here almost every time.

Budget split: how much of total spend goes to retargeting, by traffic stage

The right share of total ad budget for retargeting tracks the size of the non-converting pool, not a fixed rule of thumb copied from a course. Here's a grid worth testing against your own numbers:

• Under 500 non-converting visitors per 14 days: skip retargeting for now — the pool is too thin to hold a stable frequency, and the budget is better spent growing prospecting traffic until it clears this floor.
• Between 500 and 2,000: 10-15% of total spend, one broad segment rather than three — splitting a pool this size three ways leaves each slice too small to spend meaningfully.
• Between 2,000 and 8,000: 15-25% of total spend, split into two segments — viewed-only and added-to-cart — with the checkout-abandon segment folded into the cart segment until it is large enough to stand alone.
• Above 8,000: 25-35% of total spend, all three segments run separately with the lookback windows from the section above.

Frequency caps and creative rules that don't carry over from prospecting

A prospecting campaign can lean on the algorithm to manage frequency because the audience is large enough to self-regulate. A retargeting audience usually can't — the pool is small enough that Meta's delivery system will happily serve the same person the same ad six times in a week if nothing stops it. Set a manual frequency cap around 3-4 per week rather than trusting automatic delivery here, especially on the tighter, higher-intent segments above.

Creative needs a different job, too. A prospecting ad has to introduce the product cold. A retargeting ad already has an audience that's seen it — repeating the same hook just restates something they already decided not to act on. Creative that references the previous visit implicitly works better: a review-led angle for the viewed-only segment, a price or shipping-cost clarification for cart abandoners, a direct checkout nudge for the checkout-started segment. Running one static creative across all three segments for weeks is how a small retargeting pool fatigues inside days instead of weeks — rotate at least two concepts per segment so the same 1,000 people aren't staring at the identical image on every visit.

When retargeting won’t fix the problem — and where to find the angle that will

If a product's prospecting campaigns are already struggling — weak CTR, CPA well above target, a thin add-to-cart rate — retargeting won't rescue it. It can only harvest demand that prospecting already created, and a campaign that isn't creating much demand gives retargeting almost nothing to work with. Fix the upstream problem — creative, offer, or product — before assuming a retargeting layer will patch it.

Where retargeting creative does need outside input is the angle itself. Trackira's ad library shows how long a competing store has kept a specific follow-up angle running across Meta, TikTok, Pinterest, and Google Shopping — a faster read on what actually converts a second-visit audience than guessing from inside one account alone. It doesn't replace the traffic-floor math above; it just saves the guesswork on what to put in front of the pool once it's big enough to bother with.

How many website visitors do I need before retargeting ads are worth running?

Pull your own 14-day count of unique visitors who viewed a product but didn't buy, from Shopify Analytics or GA4. Below roughly 500-1,000 in that window, a retargeting campaign usually can't hold a stable frequency for more than a few days on any meaningful budget — the pool is just too small. Above it, retargeting is usually the cheapest conversion source in the account.

Is Meta's 100-person custom audience minimum enough to run retargeting ads?

Technically yes, practically no. A 100-person audience is enough to launch a campaign, not enough to avoid showing the same handful of people the same ad multiple times a week. Use your own non-converting visitor count, not Meta's technical floor, to decide when retargeting is worth a real budget.

Should cart abandoners and product-page viewers get the same retargeting ad?

No — they have different objections. Someone who only viewed a product is usually unsure on trust or price; someone who abandoned a cart already decided to buy and hit a friction point, often shipping cost. Segmenting into at least two groups with different lookback windows and messages consistently outperforms one blended audience.

What percentage of ad budget should go to retargeting for a dropshipping store?

It depends on the size of your non-converting visitor pool, not a fixed percentage. Roughly 10-15% once you clear 500-2,000 visitors in a 14-day window, scaling to 25-35% past 8,000. Past about 35%, a store is usually re-harvesting demand it already paid to create rather than growing net sales.