Review, claim and data checks for Gen Z brand loyalty research methods

What to take away

A loyalty study can fail on how it recruits, claims and handles data even when the method is sound. Run these checks on your own plan, and on a vendor's proposal, before any money moves. The FTC pages are US rules; the W3C and NIST documents are voluntary guidance.

Put three questions to the vendor in writing

Ask who gets contacted, what they are given, and what happens to their answers. A vendor that cannot reply in writing has not thought it through, and that silence is itself your answer.

The FTC's guide on soliciting and paying for online reviews tells marketers not to ask for reviews only from customers they think will leave positive ones. It also warns that some platforms bar reviews from people who received an incentive, even when the incentive is disclosed.

If the study seeds product with creators, the FTC Endorsement Guides (16 CFR Part 255) and the endorsement rules for creators apply on top.

See which check each method sets off

Each way of collecting loyalty evidence puts a different check first, so read the table against the method you chose rather than against the whole list.

Teen respondents need a consent check of their own. COPPA, the Children's Online Privacy Protection Act, covers children under 13, and where COPPA stops explains why it is not the whole answer for sixteen-year-olds.

Awareness and engagement are not loyalty. If your deck calls a high engagement rate loyalty, you have measured attention and renamed it. Say plainly which one the number is.

Substantiate every claim before it goes out

The FTC's policy statement on advertising substantiation says advertisers need a reasonable basis for objective claims before they are disseminated, not after a complaint arrives.

That applies to a claim like "Gen Z switches brands twice as often" the same way it applies to a product claim. A research finding is still an objective statement once it appears in a press release.

Before any figure leaves the building, write down the claim, the evidence behind it, the date of that evidence, and the qualification you will attach. If you cannot fill all four, cut the claim.

This is a US standard. Other markets set their own, so check the rule where the campaign runs and take advice from qualified counsel there rather than treating this as a universal test.

Map the data flow, not the notice

Draw what you collect, where it lands, who can see it, how long it stays, and what happens when someone asks you to delete it. Then check that drawing against the law that governs the market.

The W3C Privacy Principles set out shared concepts for people who design systems that handle personal data, and they warn against pushing privacy work onto the individual. They are principles, not law. In practice the burden sits with you, not with a sixteen-year-old who did not read the consent screen.

The NIST Privacy Framework starting guide describes a voluntary process for identifying privacy risk, assigning owners and recording responses. It is a management aid, not legal clearance.

Which instrument you chose decides how much of this applies, because the research tools differ in what they collect.

A worked example: the loyalty claim that did not survive review

A team ran a survey of 400 respondents aged 18 to 24, found 71% said they would buy again, and published that Gen Z brand loyalty is up. A review found three problems.

The sample was a convenience panel, not a defined population. The question asked about intent, not behavior.

And the incentive was a discount code, which the FTC guidance treats as a relationship that needs disclosing.

The corrected version reports the intent figure as intent, names the panel and its limits, discloses the incentive, and pairs it with twelve months of repeat-purchase data from the same cohort. If the two disagree, the behavior wins.

Common questions

What should a vendor brief say about incentives?

Name every reward, say whether it depends on a positive answer, and require disclosure wherever a response is shown. An incentive tied to a good rating is the first thing to remove.

Does the substantiation standard apply to campaigns outside the US?

Not as such. It is a US standard, and other markets set their own, so check the rule where the campaign runs and take advice from qualified counsel there before you publish.

What if the survey and the purchase data disagree?

Report both and let the behavior lead. If you still need to choose the instrument, pick the method before the vendor and write the question first.