Teens, Tokens, and Location

A click-through checkbox on a landing page is weak evidence of parental consent. It does not establish that a parent, rather than the teen, clicked; it does not record what the parent was shown; and it does not create a verifiable record. Verifiable parental consent generally requires a method proportionate to the risk - a signed form, a credit-card or government-ID check, or a knowledge-based verification call - and the method should be documented. For a study collecting geolocation and household income from teenagers, the risk is not trivial, so the lighter methods are hard to defend.

The teen's own assent is also required, and it should be a separate step that explains, in plain language, what is collected, who sees it, and that they can stop. The current two-click design collapses parent and teen into a flow that a 12-year-old can complete in thirty seconds.

On the age signal: if device and response patterns suggest some respondents are under 13, treat those records as ineligible rather than as data. Children's privacy rules in many jurisdictions impose strict consent requirements below 13, and continuing to hold the data after suspicion arises is difficult to defend. Flag, quarantine, delete, and adjust screening - for example, an age gate before the survey and a consistency check on birth year against device signals.

Apply data minimisation and purpose limitation. Geolocation is being collected for a retail activation, not to answer the research question about play habits; that is a marketing use dressed as research. The defensible options are to drop it, or to collect coarse location (city or postcode district) with explicit, separate consent and a clear explanation that it will be used to plan store events. Precise coordinates from teenagers, retained indefinitely, create re-identification and safety risk that the study does not need.

Household income is sensitive but common in audience research; the issue is whether a teenager can answer it accurately and whether it is necessary. Ask it as a banded, optional question, explain why it is asked, and do not force an answer. If the client only needs a proxy for spending, a simpler measure may do.

For both, make consent granular: separate tick-boxes for the core survey, for coarse location, and for any future contact. Bundling everything into one "I agree" invalidates the consent for the optional elements. And state retention: location data should have the shortest retention of anything in the study.

Start with the platform rules, because they bind the recruitment channel regardless of what the client wants. Major platforms restrict how certain categories can be targeted to users known to be minors, and some prohibit using interest categories based on engagement with child-directed content to build audiences. Using "engaged with toy content" as a targeting criterion to reach teenagers is exactly the kind of construction that draws enforcement action and account suspension.

The practical route is age-neutral, interest-light recruitment: broad geographic targeting, creative that describes the study rather than the product, and no use of child-directed content categories to build a custom audience. If the platform offers a certified teen-targeting mode with limited categories, use it.

Separately, tell the client that the research sample must not be shaped by product engagement. Recruiting only teens who already engage with toy content produces a biased sample and undermines the study's claim to represent play habits. The targeting request is both a compliance risk and a methods problem, and the methods argument is often the one that persuades a marketing team.