What it was. An online survey of people who resell clothing, run by Fleek and recruited through Fleek's own channels and reseller communities. Participation was voluntary, and a screening question routed non-resellers out.
Fieldwork. 27 August – 8 September 2026, a 13-day window. Median completion time was 11.7 minutes.
How the sample resolves
| Stage | Responses |
|---|
| Responses received | 341 |
| Internal test entries removed | −2 |
| Screened out as a non-reseller | −1 |
| Abandoned part-way | −1 |
| Completed responses analysed | 339 |
| of which, currently reselling | 328 |
| of which, no longer reselling | 11 |
The four bases used in this report
Different questions were asked of different groups, so the base changes between charts. Every chart states its own.
| Base | n | Used for |
|---|
| All completed responses | 339 | Motivations, demographics and business shape. Anything asked of everyone |
| Current resellers | 328 | Platforms, live selling, sourcing and frustrations. Anything about how they trade today, so it excludes the 11 who no longer resell |
| Current resellers in the UK and EU | 264 | The platform figures in Sections 1 and 2. US respondents are excluded because Poshmark and Mercari barely exist outside North America, and including them understates the European picture |
| Ban stories | 95 | The reasons sellers were given for a ban or restriction. 92% came from UK and EU respondents, so that split is effectively already a UK and EU figure |
Where respondents were based
United Kingdom 182 · United States 52 · France 33 · Germany 13 · Ireland 6 · Belgium, Italy and Portugal 5 each · Romania, Canada and Pakistan 4 each · Netherlands, Finland, Poland, Spain and Sweden 3 each · Austria, New Zealand and Norway 2 each · Czech Republic, Lithuania and Australia 1 each, making 22 sovereign countries. Scotland, Wales and Northern Ireland are counted within the United Kingdom, as are respondents who gave only a county or town. One further respondent was in the Channel Islands (a Crown Dependency, so not counted as a separate country) and one gave only “Europe”.
Generational fingerprint bases. 18–25 n=150 · 26–45 n=121 · 46–60 n=38, all current resellers.
How to read the charts
Three shapes appear in this report, and only one of them totals 100%.
- Single select. One answer per person, so it totals 100%. Example: the one platform that earns them most.
- Multi-select. Several answers per person, so it totals more than 100%. Example: which platforms they sell on, or why they resell, where up to three reasons were allowed.
- Independent bases. Each row is its own group and its own 100%. Reading down a column shows how something changes across groups; it is not a split of one total. Example: the full-time rate by length of time reselling, or the ban rate by platform.
Derived measures
- Monthly sales were collected in bands, so medians use band midpoints: under £250 → £125 · £250–750 → £500 · £751–2,000 → £1,375 · £2,001–5,000 → £3,500 · £5,001–10,000 → £7,500 · over £10,000 → £12,500. The top band is capped, so any total built from it is conservative.
- Generational fingerprints index how often each age group gave a reason against the average for all current resellers: 100 = average, 181 = 81% more likely, 60 = 40% less likely. The 26–35 and 36–45 bands were merged because their answers were near-identical. Respondents over 60 (n=7) and those who preferred not to give an age (n=12) are included in all totals but are not shown as groups, because the bases are too small to index.
- Win rate is the share of a platform's users who name it their single biggest earner, counting only resellers who sell on at least one other platform. It is the fair comparison, because only sellers with an alternative are included.
- UK and EU covers respondents who gave a UK nation, a UK county or town, or an EU member state. Everyone else sits outside it, including the US, Canada, Australia, New Zealand, Norway and Pakistan.
Precision and weighting
This is a non-probability sample. Respondents opted in rather than being drawn at random from a known population of resellers, so a margin of sampling error does not apply and none is quoted. Standard significance testing assumes random selection, so we have not run it, and no difference in this report is described as statistically significant. Differences are reported as they were observed, with the base beside them. No weighting has been applied: every figure is an unweighted count of respondents, so the sample's own shape, including its UK weighting, carries straight through to the numbers.
Limitations, stated plainly
- The sample is self-selecting, recruited through Fleek and reseller communities. It is not a random or quota sample of all resellers, and it over-represents the UK (54% of respondents) and the platforms Fleek's audience uses. Read it as a detailed picture of an engaged reselling population, not a census of the market.
- Every figure is self-reported, including sales, income share and ban histories. None is independently verified.
- Where a base falls below about 30 we say so beside the figure, and it should be read as direction rather than precision.
- The survey did not ask when respondents expect to go full time, so this report makes no claim about timeframes.
- Fieldwork ran over 13 days in late summer, and the seasonal questions reflect that moment.
Quotes and consent
The survey's final question asked “Would you like to be featured in the report?” Respondents chose one of four options, and we used them as follows:
| What they chose | n | How we used it |
|---|
| “Quote me from what I've written” | 98 | The only group named in this report, and then by first name and country only |
| “Happy to be interviewed” | 143 | Quoted anonymously. Agreeing to an interview is not the same as agreeing to have a written survey answer published under your name |
| “Anonymous only” | 84 | Quoted anonymously |
| “No thanks” | 14 | Not quoted anywhere in this report, named or anonymous |
No surname, handle or shop name appears anywhere, and no respondent is identified by city. Quotes are lightly edited for length and clarity; meaning is never changed.
