Facebook Ads Retargeting Strategy: The Complete Playbook

Retargeting is the campaign type that should be embarrassingly profitable — and yet most brands run it badly. Either they push too much budget at a tiny audience and burn it out, or they treat retargeting as a single ad set with one generic creative. Both approaches leave money on the table.

Here's the retargeting playbook that I'd run for any ecommerce or service brand serious about getting more out of warm traffic.

What retargeting is (and isn't)

Retargeting is showing ads to people who already know you. They've visited your site, watched your video, engaged with your Instagram, opened your email — something. The whole point is to re-engage existing intent, not generate new awareness.

It is not a substitute for prospecting. If your retargeting numbers look great but you're not running cold acquisition, you're not really doing paid media — you're harvesting the work other channels did. As soon as those channels slow down, retargeting dies.

The five retargeting audience types (in order of intent)

From coldest to warmest:

Page/Instagram engagers — interacted with your social profiles in the last 30-90 days
Video viewers — watched a percentage of your video ads (50%+ usually)
Site visitors — landed on your site without taking deeper action
Add-to-cart non-purchasers — clear intent, but didn't finish
Initiate checkout non-purchasers — highest intent, just lost them at the last step

Each of these needs different messaging and different budget. Treating them as one bucket is the most common retargeting mistake I see.

Segment 1: Page and Instagram engagers (lightest touch)

These people barely know you. Maybe they liked a post or saved a Reel. Treat them like a warm-up audience, not a hot one.

Creative: Brand introduction. Founder story. "Why we built this." Educational content about the problem you solve.
CTA: Soft. Learn More, Watch Now, Read Story.
Budget: Small. Maybe 10-15% of your retargeting spend.
Frequency cap: 3/week max.

Segment 2: Video viewers

If someone watched 50% or more of your video ad, they were genuinely interested. Use that signal.

Creative: Different angle from the original video. Don't show them the same ad again — show them the next chapter.
CTA: Mid-strength. See Pricing, View Collection, Get Demo.
Budget: 15-20% of retargeting spend.
Frequency cap: 4/week.

Segment 3: Site visitors (no purchase, no add-to-cart)

These are window shoppers. They browsed but didn't commit. The biggest mistake here is treating them like cart abandoners — they're not. They didn't get that close.

Creative: Product education. Reviews. Comparison content. Use cases.
CTA: "See All [Products]," "Read Reviews," "Browse Collection."
Budget: 25-30% of retargeting spend.
Frequency cap: 4/week.

Segment 4: Add-to-cart non-purchasers

Now we're talking. They picked something up and put it back down. They're undecided.

Creative: Trust signals. Reviews. Free shipping reminder. Returns policy. Maybe a soft incentive ("Free shipping at checkout").
CTA: Direct. "Complete Your Order," "Back to Cart."
Budget: 20-25% of retargeting spend.
Frequency cap: 6/week.

Segment 5: Initiate checkout non-purchasers

The hottest audience you have. They typed in their name, maybe their email, maybe their card — and bailed. Often a technical or pricing concern killed the deal.

Creative: Urgent and direct. "You left this behind." Discount code (if margin allows). Customer service reassurance. Limited stock messaging (only if true).
CTA: "Finish Checkout," "Claim Your Order."
Budget: 15-20% of retargeting spend.
Frequency cap: 8/week (the audience is small, so the absolute number of impressions is still low).

How long should retargeting windows be?

Default windows in Meta are 30/60/90 days. The right window depends on your purchase cycle.

Impulse purchases (< £40 ecommerce): 14-30 days
Considered purchases (£40-£500): 30-60 days
High-ticket purchases (£500+): 60-180 days
B2B leads: 90-180 days

Don't blanket-use 30 days because it's the default. Look at your actual customer journey data — how long between first site visit and purchase? — and set windows to match.

Frequency hygiene is everything

The number one reason retargeting goes wrong is unmanaged frequency. The audience is small, the budget is fixed, and Meta has to keep showing the same ads to the same people. Frequency creeps from 3 to 6 to 12, and your warm audience starts hating you instead of buying from you.

Monitor frequency weekly. If it's above your cap for any segment, do one of three things:

Refresh creative. New ads, same audience.
Expand the audience. Lengthen the lookback window.
Lower the budget. Yes, lowering retargeting budget often improves retargeting performance.

Don't forget exclusions

Every retargeting ad set should exclude past purchasers (unless it's a winback campaign, which is a different beast). This is so basic and so often missed. People will rage at you for showing them ads for the product they bought yesterday.

Also exclude employees, current customers (in the case of a one-time-buy product), and anyone you've already sent an aggressive retargeting message to in the last 14 days.

A clean retargeting structure

For most accounts under £2,000/day, this is the structure I'd use:

Campaign: Retargeting

Ad set 1: Site visitors (last 30 days), excluding ATC and past purchasers

Ad set 2: ATC + Initiate Checkout (last 14 days), excluding past purchasers

Ad set 3: Page/Instagram engagers (last 60 days), excluding everyone above

Three ad sets. 3-4 ads each. Frequency capped. Budgeted appropriately. That's a complete retargeting setup that will outperform 90% of what you see in real accounts.

Letting AI manage your retargeting layers

Keeping this many segments balanced manually is a part-time job. Pix-Vu handles retargeting segmentation, frequency caps and creative refreshes automatically using your pixel data. £99/month — much cheaper than the budget you waste re-spamming exhausted audiences.