How to Set Up a Meta Ads Retargeting Funnel That Converts
META ADS · 10 MIN READ

How to Set Up a Meta Ads Retargeting Funnel That Converts

Retargeting is the easiest thing to make look profitable and one of the easiest to overspend on. Here is a structure that stays honest about what it is actually adding.

By Tariq Sallam·September 2026

Retargeting reports beautifully, because you are advertising to people who already showed intent. Some of them were going to buy anyway, and the campaign takes credit for it.

That does not make retargeting useless. Done properly it is one of the highest-return things in an account. Done as most small accounts do it, it is a way of paying to reach your existing pipeline. This is the structure I use and the tests I run to keep it honest.

Quick Info

Minimum traffic to bother
Roughly 2,000 monthly visitors
Budget share
10% to 20% of total spend for most accounts
Number of stages
Three is plenty
Frequency cap
Aim for under four impressions per person per week
Window
Seven to thirty days for most businesses
Essential test
Turn it off for two weeks and watch total sales

Do you have enough traffic?

Retargeting needs audiences large enough to deliver against. Below roughly two thousand monthly visitors, your pools are too small, delivery is inefficient and you are better off putting the whole budget into prospecting.

The exception is high-value sales, where reaching forty of the right people repeatedly is worth it. If your average sale is £8,000, small audiences are fine.

The audiences worth building

Build these in Audiences and keep them running so they populate. Three stages is enough; the six-stage funnels people diagram are mostly theatre.

Stage one: high intent

—Added to cart or started checkout, last 7 to 14 days.
—Visited a pricing or booking page, last 14 days.
—Started a form and did not submit.
—This is where most of the return is, and the audience is small.

Stage two: engaged

—Visited any page, last 30 days, excluding stage one.
—Watched 50% or more of a video.
—Instagram or Facebook engagers, last 30 days.
—Email subscribers who have not bought, uploaded as a custom audience.

Stage three: existing customers

—Past purchasers, for repeat purchase, upsell or review requests.
—Frequently the best return in the whole account and the most neglected.
—Exclude them from prospecting so you stop paying to acquire people you have.

Exclusions, which matter more than the inclusions

01Exclude converters from all prospecting and from stages one and two. Nothing wastes budget faster than advertising to somebody who bought yesterday.
02Exclude stage one from stage two, so the same person is not in both auctions.
03Exclude recent purchasers from repeat-purchase ads until your typical repurchase window.
04Exclude existing customers from lead generation entirely.

Getting exclusions wrong is the most common structural fault in retargeting, and it produces the experience everybody complains about: being followed by an ad for something they already bought.

Creative by stage

The mistake is running the same ad to all three. Each stage has a different reason for not having converted.

Stage one: remove the friction

They were close. Answer the specific objection: a guarantee, a delivery timeframe, a payment option, a testimonial about the thing they were worried about. Not a discount by default, because you will train people to abandon.

Stage two: give them a reason to return

They looked and left. Show proof, show the outcome, show a case study, or offer something smaller than the main purchase: a guide, a call, a sample.

Stage three: sell them the next thing

New products, complementary services, a referral request, a review request. Treat it as customer marketing rather than advertising.

Retargeting creative should acknowledge that they already know who you are. Repeating the introduction is the wasted opportunity.

Frequency, and how to stop annoying people

—Watch the frequency metric. Above roughly four impressions per person per week you are into diminishing returns and reputational cost.
—Cap the window. Seven to thirty days covers most businesses; ninety-day windows mostly serve ads to people who forgot they visited.
—Refresh retargeting creative as often as prospecting creative. Small audiences fatigue faster, not slower.
—Keep budgets proportionate. A large budget on a small audience produces enormous frequency and nothing else.

The single most common retargeting error is too much budget on too small an audience. It shows up as high frequency, rising costs and irritated customers.

Budget split

01Prospecting: 70% to 80%.
02Stage one and two retargeting: 10% to 20% combined.
03Existing customers: 5% to 10%, and worth more than that share suggests.
04Review the split monthly against blended cost per acquisition rather than per-campaign return.

When retargeting looks like it has a return of eight and prospecting looks like two, the instinct is to move the money. Do not. Retargeting has nothing to retarget if prospecting stops.

The honest measurement problem

Retargeting's reported return is inflated by people who would have converted anyway. Nobody can tell you by how much without a test, and the test is straightforward if slightly nerve-wracking.

01Turn retargeting off entirely for two weeks.
02Hold prospecting budget constant.
03Compare total sales or enquiries, not per-campaign attribution.
04If total volume falls, retargeting was adding demand. If it holds, it was claiming credit.
05Repeat annually, because the answer changes with your traffic mix.

I have run this test in several accounts. In some, turning retargeting off cost real revenue. In others, total sales did not move at all and we redeployed the budget. You cannot know which you are without looking.

Setting it up, in order

01Confirm the pixel and Conversions API are both firing, and that page-level events exist for your key pages.
02Build the three audiences and let them populate for a week.
03Build the exclusions before the campaigns.
04One retargeting campaign, one ad set per stage, three ads each.
05Set the budget at 10% to 20% of total spend.
06Check frequency after a week and cut budget if it is above four.
07Run the off test after a month of stable data.

Frequently Asked Questions

How much traffic do I need for retargeting?

Around two thousand monthly visitors for it to deliver efficiently, unless your sale value is high enough to justify reaching small audiences repeatedly.

What percentage of budget should go to retargeting?

10% to 20% for most accounts. Higher looks efficient in reporting and usually means you are buying conversions you already had.

How long should the retargeting window be?

Seven to thirty days for most businesses. Longer windows mostly reach people who have forgotten you and inflate frequency.

Should retargeting ads offer a discount?

Not by default. It trains customers to abandon and erodes margin. Answer the objection first and keep discounting for genuine stock or seasonal reasons.

Is retargeting still effective after privacy changes?

Yes, with smaller and less precise audiences than before. Server-side events via the Conversions API recover a meaningful share of the lost signal.

How do I know if retargeting is actually adding sales?

Turn it off for two weeks with prospecting held constant and compare total sales. It is the only reliable test available to a small account.

Before You Go

Three audiences, correct exclusions, different creative per stage, 10% to 20% of budget, and a frequency check. That is the whole structure and it takes an afternoon.

Then run the off test, because the reported return on retargeting is the most flattering number in an ad account and the least trustworthy. The budget guide covers where the rest of the money should sit.

Three stages. Correct exclusions. Test the off switch.

WRITTEN BY TARIQ SALLAM
Marketing Consultant. Entrepreneur. Content Creator.

I'm a marketing consultant, entrepreneur and content creator. I help businesses grow through practical marketing, websites, SEO, content and AI.

More About Tariq →

Keep reading