In July 2026, a merchant gave a virtual try-on app a 5-star review after using it for two minutes. Both that review and a 3-minute one two days later still count, full weight, toward the perfect 5.0 the app shows every store owner. Shopify prints the tell on every review card, in small gray text almost nobody reads: "time using the app." We read that line on all 223 reviews of all 11 live virtual try-on apps on the Shopify App Store on 2026-08-30. Disclosure first: Ello is our app, and it has exactly 1 review (left after 20 days of use) — we're the least-reviewed vendor in this market publishing a piece questioning review counts, and you should notice that incentive. Our defense: every number below sits on a public listing page and can be recounted in an afternoon.
Share of reviews left within the reviewer's first 7 days of use: SS: AI Try On 91% (10 of 11), ANTLA 65% (33 of 51), Banuba 60% (3 of 5), Looksy 50% (3 of 6), GenLook 43% (15 of 35), TryPoint 42% (5 of 12), Icona 40% (6 of 15), Camweara 10% (6 of 59), Fittingbox 8% (1 of 12), Auglio 0% (0 of 16), Ello 0% (0 of 1).
| App | Reviews | ≤1 day | ≤7 days | >3 months | Shortest usage | Grade |
|---|---|---|---|---|---|---|
| Camweara | 59 · 5.0 | 1 | 6 | 18 | About 5 hours | A |
| Fittingbox | 12 · 4.7 | 0 | 1 | 6 | 4 days | A |
| Auglio | 16 · 3.5 | 0 | 0 | 10 | 8 days | A |
| GenLook | 35 · 5.0 | 2 | 15 | 5 | About 2 hours | B |
| TryPoint | 12 · 5.0 | 1 | 5 | 5 | About 22 hours | B |
| Icona | 15 · 5.0 | 2 | 6 | 1 | 11 minutes | C |
| Looksy | 6 · 3.8 | 1 | 3 | 1 | About 17 hours | C |
| ANTLA | 51 · 5.0 | 16 | 33 | 3 | 14 minutes | D |
| Banuba | 5 · 4.0 | 0 | 3 | 1 | 1 day | D |
| SS: AI Try On | 11 · 5.0 | 4 | 10 | 0 | 2 minutes | F |
| Ello (ours) | 1 · 5.0 | 0 | 0 | 0 | 20 days | n/a — 1 review |
Rubric: A = seasoned-dominated (≤15% first-week, real long-tenure cohort) · B = mixed with a meaningful over-a-month cohort · C = majority under 30 days · D = first-week-dominated (60%+) or major base flags · F = effectively zero tenure. Grades rate the review base's evidential value, not the app. Flags behind them, all public on the listings: ANTLA — 16 of 51 reviews inside the first day, including "14 minutes," "25 minutes," and six separate "About 1 hour" reviews; only 5 of 51 past a month; all 51 five-star. SS: AI Virtual Try On — reviews at 2 and 3 minutes of use, 10 of 11 inside a three-week July window, 7 of 11 from one country, longest tenure ever 13 days. Banuba — one five-star from a store named "My Store" (Shopify's default), and its only 1-star reviewer openly questioned the other reviews (Banuba replied publicly). TryPoint — the same store name behind two separate five-star reviews.
Every incentive points there — no conspiracy required. Review count and velocity feed App Store ranking; ranking drives installs; installs drive revenue. So apps prompt at the moment goodwill peaks: right after a smooth install. Shopify permits asking for reviews (it prohibits paying or pressuring for them), and no reader can tell from outside which path produced a given five-star. That's why tenure matters more than count: it's the one signal on the card that can't be asked for. We are NOT calling any app fraudulent — intent is invisible from public data, and a two-minute review can be sincere. We're saying a rating's evidential value is knowable, it varies wildly between apps showing identical stars, and our 2027 ranking now weights these grades: a D-grade base costs a position; the F-grade app isn't ranked.
Camweara (59 reviews across five years, one under a day, several past two years), Fittingbox (median tenure ~4–5 months), and Auglio prove reviews can age honestly. Auglio is the most instructive: the longest-tenured reviewer base in the audit AND the lowest rating (3.5) — because merchants who stay for years are the ones still around to report billing disputes and reliability failures. In this category, the apps with the most trustworthy-looking review bases have the lowest ratings, and several perfect 5.0s are built substantially on reviews left before the reviewer could have learned anything — exactly backwards from how buyers read stars.
We can't know intent from public data and don't claim to. What we measured (2026-08-30, all 223 reviews): 7 of 11 live apps had 40%+ of reviews from the first week, including 5-stars at 2, 3, and 14 minutes of use. Fast reviews can be sincere — but they can't speak to conversion, returns, quality, or reliability. Read the tenure line and weight accordingly.
The product's job is conversion lift and fewer returns, both unmeasurable inside a week — return windows haven't closed yet. A day-one reviewer can honestly rate onboarding, nothing more.
By tenure (2026-08-30): Camweara, Fittingbox, and Auglio — Auglio tellingly pairs the longest tenure with the lowest rating (3.5). Weakest tenure: SS: AI Try On (no reviewer past 13 days) and ANTLA (16 of 51 within the first day).
Asking is allowed; paying, incentivizing, or gating is prohibited. From outside you can't tell which produced a given review — which is why disclosed usage time is the more honest signal.