4 things to check before trusting Gen z shopping habits statistics

What to take away

Gen z shopping habits statistics get quoted in decks because they are memorable, not because anyone checked them. Four checks catch most of the bad ones. Run all four before a figure reaches a budget line.

Check 1: who was actually asked

Start with the sample frame, because it decides what the number can describe. Pew Research Center defines Gen Z as the 1997 to 2012 birth years, roughly ages 13 to 28 today. A study reporting "Gen Z" across that whole range has pooled a middle schooler with a 28-year-old renter.

Name the recruitment source. GWI and YouGov draw on standing survey panels, so their Gen Z is people who join panels. A TikTok Shop buyer survey reaches TikTok Shop buyers. Piper Sandler's Taking Stock With Teens surveys about 13,500 US teens, average age near 16, twice a year in schools.

Then check what the methodology page discloses. AAPOR's Transparency Initiative and ESOMAR's guidance for sample buyers both ask for recruitment method, field dates, sample size, weighting and completion rate. A study that hides those five cannot carry a generational claim.

Check 2: what counted as a purchase

Four different measures hide behind the word purchase. Stated intent comes from surveys. Receipt panels record what panelists actually bought. Card data records aggregated card spend.

Circana, formed from IRI and NPD, and NielsenIQ's consumer panel run receipt panels. Mastercard SpendingPulse and Bank of America's Consumer Checkpoint report card data. A Shopify merchant's order log or a TikTok Shop seller dashboard records one seller's transactions.

Adobe Analytics measures online transactions across a large slice of US retail sites, which Adobe describes as more than a trillion retail-site visits. That is observed behavior, but it covers online sales only and only the sites in the sample.

These sources produce different numbers for the same question, and none is wrong on its own terms. The error is citing a stated-intent figure as a sales figure. Write down which of the four you hold before you compare it to anything. If a vendor will not say, assume stated intent and discount accordingly.

Check 3: who paid for the study

Vendor research is common and sometimes good. What matters is whether the funding shaped the question.

A report from Klarna, Afterpay or Zip will find that buy-now-pay-later use is growing among young shoppers, because the data is their own checkout traffic. Deloitte's Global Gen Z and Millennial Survey, fielded with an independent research firm for roughly 23,000 respondents across 44 countries, carries a different sample.

The FTC guidance for marketers using reviews covers fake feedback, selective requests, conditioned incentives, hidden relationships and paid ranking. The FTC's rule against fake and incentivized reviews took effect in October 2024. The same integrity questions apply to research you cite: who paid, what was asked, and what was left out.

Name the funder in the same sentence as the number. Readers can weigh it themselves.

Check 4: can you reproduce it

A statistic you cannot rerun on your own customers is a hypothesis. Turn it into a question you can field: same wording, your list, your traffic, a stated date.

Qualtrics, SurveyMonkey or a Google Forms link can carry the question. Cint, Prolific and Dynata rent panel sample if you need people outside your list.

Record the source date and limits beside every figure you keep. When a number changes, you need to know whether the market moved or the study did. Statista aggregates other people's research, so cite the original study, not the Statista page.

The W3C explanation of information and relationships makes a useful parallel point for reporting: structure that exists visually should also exist in the underlying data. Apply that test to your own dashboard before you trust it.

Where the checks fail in practice

Keep the prior version of any figure you correct. When a number is replaced, the old one usually survives in a deck somewhere, and you will need to trace where it went.

A worked correction

A team cites a figure that 68% of Gen Z shoppers abandon carts over shipping cost, the kind of claim Baymard Institute researches. The methodology page shows the sample came from one retailer's email list.

The correction has three steps. Relabel the figure as that retailer's customers, not Gen Z. Field the same question on your own list. Compare the two before either goes into a budget.

Then trace the correction through every downstream output: the deck, the media plan, the creative brief, the agency brief. A corrected number that stays in the old deck is not corrected.

Common questions

What is the first thing to check in a Gen Z shopping statistic?

The sample frame. Find out how respondents were recruited and on what dates. If the study will not say, the number cannot carry a generational claim.

Are vendor-funded statistics usable?

Yes, with the funder named. The funding does not make a study false, but it shapes which questions get asked. Put the sponsor next to the figure every time.

How do I test a statistic on my own customers?

Field the same question, with the same wording, to your own list or site traffic. Record the date and the sample. Compare the two figures before either one reaches a budget.

What do I do when a figure turns out to be wrong?

Preserve the old version, note who used it, and correct every asset it reached. The CISA business-system logging guidance explains why organizations retain and review event records; the same habit applies to a figure you have published.

For procurement records, the CISA software acquisition fact sheet covers development practice, supply-chain exposure and vulnerability management.

Neither decides your local question, and neither replaces Gen Z shopping habits factors or the worked examples you build from your own data.