Retargeting is useful only when the first visit left recoverable value
Retargeting—also called remarketing—uses a platform-defined audience of people who have already interacted with the business. That may include website visitors, app users, video viewers or customers who supplied information. The business then pays to reach some of them again.
The attraction is obvious: these people are familiar with the offer. The danger is equally important. Familiarity does not prove suitability, consent, unmet intent or incremental value. A platform may report an efficient sale that would have happened through direct, organic, email or sales activity anyway.
My verdict is: use retargeting as a bounded recovery system, not a default tax on every visitor. It is most defensible when a valuable buying journey needs more than one interaction and the second message removes a real obstacle. It is weak when the business merely repeats the same promise more often.
This is a different decision from asking whether the business is ready for paid advertising, when to stop paid ads or how long paid advertising takes. Here, the question is whether a specific share of prior demand deserves another paid contact.
The Retargeting Recovery Value Gate
I would require six connected answers before reserving budget for retargeting.
Who is worth returning?
Separate valuable buying behaviour from accidental, job-seeker, support and poor-fit visits.
Why did they stop?
Name the missing proof, timing, stakeholder, price clarity, reassurance or next step.
Can the platform serve it?
Check consented, matchable audience size and delivery eligibility before planning reach.
Should this person be included?
Apply the current privacy, platform and sensitive-category rules in every served market.
What new value appears?
Offer evidence, comparison, timing, objection handling or a relevant continuation—not repetition.
What changed because of it?
Test qualified demand, contribution and assisted progress against a credible comparison.
Google's current help documentation calls remarketing audiences “your data segments” and says they may include previous website, app, video or customer interactions. It also publishes minimum active-user thresholds that differ by network. That means raw site traffic is not the same as an addressable audience, and a small audience may never support the intended programme.
Judge recovered value, not reported efficiency
A retargeting report often looks strong because the audience already knows the business. That is not enough. The owner needs the value that would not have arrived without the extra paid contact.
For ecommerce, compare incremental orders, contribution after media, returns and new-versus-existing customer value. For B2B and service businesses, compare suitable enquiries, accepted opportunities, sales progression, expected contribution and the cost of delayed decisions. Keep the definitions separate; a cheap lead and a profitable order are not interchangeable.
| Reported result | Commercial question | Owner interpretation |
|---|---|---|
| Low cost per conversion | Would these buyers have returned without the ad? | Efficient reporting is not yet incremental growth |
| Many repeat visitors | Were they suitable prospects, customers, staff or support users? | Audience quality must precede scale |
| More form fills | Did fit, acceptance and pipeline progression improve? | Judge sales value, not form volume |
| More recovered baskets | Did contribution after discount, returns and media improve? | Recover margin, not headline revenue |
This scenario demonstrates the questions. It is not a benchmark, forecast, client result or proof that retargeting caused growth.
Privacy belongs inside the investment decision. The UK Information Commissioner's Office says its updated guidance on storage and access technologies was finalised on 29 April 2026, and that technologies used for online advertising require consent. Google separately requires privacy-policy disclosures for its data segments. Requirements vary by jurisdiction, platform, audience and data type; obtain appropriate legal advice rather than treating this guide as compliance approval.
Decide where retargeting belongs—and where it does not
| Evidence pattern | Verdict | Next move | Avoid |
|---|---|---|---|
| Enough suitable visitors; known hesitation; valid permission | Test | Use one bounded audience, one new message and a commercial outcome | Retargeting every visitor with the same ad |
| Long sale; multiple stakeholders; useful proof exists | Support | Sequence evidence that helps the buying group progress | Calling views or clicks pipeline |
| High-intent abandonment; healthy contribution | Recover | Remove the obstacle, exclude purchasers and test incremental contribution | Using discounts as the automatic answer |
| Small or ineligible audience | Delay | Improve demand, consented data and owned follow-up first | Broadening into unrelated people to force delivery |
| Weak offer, confusing page or slow sales response | Repair | Fix the constraint before buying another exposure | Paying to repeat a broken journey |
| Sensitive data, unclear permission or policy risk | Stop | Get specialist legal and platform guidance | Assuming technical availability means lawful use |
If visitors arrive but do not buy, use the traffic-to-revenue diagnosis. If enquiries are suitable but stall, use the qualified-lead momentum guide. If the audience itself is wrong, start with customer-segment selection.
The Visit-to-Incremental-Value Evidence Loop
A useful programme closes the gap between a prior interaction and a business result without confusing correlation with causation.
Define recoverable demand
Name the actions that signal suitability and the people who must be excluded.
Find the stopping point
Use page, CRM, sales and customer evidence to identify what prevented progress.
Confirm eligibility
Validate consent, disclosure, platform rules, audience size and data handling.
Add a useful next message
Resolve the hesitation with proof, clarity or a relevant continuation of the journey.
Connect to business stages
Track suitable enquiry, opportunity, purchase, contribution and exclusions beyond the click.
Test incremental value
Use holdouts, experiments or matched comparisons where feasible, then keep, narrow or stop.
Review Google Ads management, Meta Ads management, conversion tracking, landing-page work, case evidence, evidence standards and Thomas's operating model.
Practitioner note: I would ask what valuable action happened before someone entered the audience, what prevented progress, what the next ad genuinely adds and how the business will detect value that would not otherwise have occurred. If those answers are missing, more exposure is not yet a strategy.
Sources and evidence notes
Sources and current search results were checked on 14 October 2026. Search prioritisation is qualitative; no unverified search volume, universal budget split, conversion-rate benchmark, forecast or client result is used. The Retargeting Recovery Value Gate, Visit-to-Incremental-Value Evidence Loop, decision matrix and illustrative example are original ThomPerformance practitioner analysis.
- Google Ads Help: about audience segments
- Google Ads Help: how your data segments work and current eligibility thresholds
- Google Ads Help: privacy-policy disclosures for data segments
- UK ICO: guidance on storage and access technologies, finalised 29 April 2026
- UK ICO: how the rules apply to online advertising
Frequently asked questions
What are retargeting ads in plain English?
Retargeting shows advertising to people who previously visited, used, bought from or otherwise engaged with a business. Platforms may call these remarketing or first-party data audiences. The business still needs an eligible audience, a valid data basis and a useful reason to re-engage.
Are retargeting ads worth it for a small business?
Sometimes. They can help when the business attracts enough suitable visitors and buyers need more than one interaction. They are less useful when eligible audiences are too small, the offer is weak, sales follow-up fails or most visitors were never likely customers.
How much should a business spend on retargeting?
There is no universal percentage. Fund enough exposure to test one defined audience and message without starving new-demand activity. Cap the test by acceptable customer economics, audience size, buying cycle and the value of the action—not by a generic budget split.
How long should retargeting continue?
Match the eligible window to the real buying and replacement cycle. A short-consideration purchase may need days; a complex B2B decision may need longer. Exclude customers or closed opportunities when the message no longer helps, and stop when incremental value does not justify continued exposure.
Do retargeting ads require cookie consent?
Rules vary by jurisdiction and technology. In the UK, ICO guidance says storage and access technologies used for online advertising require consent. Obtain current legal advice for the markets you serve, make choices clear and do not treat a platform setting as proof of compliance.
Paid-growth diagnostic
Recover valuable demand without paying for noise
Share the buying journey, eligible audience, sales cycle and outcome data. I'll identify whether retargeting should recover a specific hesitation, support a longer decision, remain a small test—or wait until the offer, data, page or sales process is stronger.
Discuss a retargeting growth diagnosis