Public · Errata

Errata

Corrections, removals, and data-quality improvements. We log anything that materially changed what we showed you, so a number you saw last month can be re-read with the right context.

See also: methodology · changelog · compliance · freshness.

• 2026-05-26scores

  Opportunity score labels recalibrated

  We re-tuned the word labels behind every opportunity and vulnerability score to match the real spread of the market, so the top of the scale is reachable and a strong category reads as strong. The underlying numbers didn't change only the words we attach to them got more honest.

• 2026-05-25categories

  Country names removed from the category list

  A handful of country and region names had slipped into the category rankings as if they were app categories. We now filter the full set of country names out, so the category leaderboard only shows real app categories.

• 2026-05-24pain themes

  Foreign-language complaints excluded from pain themes

  Some non-English reviews were producing meaningless complaint themes (stray words that aren't really complaints). We tightened the language filter so pain themes are built from English reviews only, giving cleaner, more readable themes.

• 2026-05-20categories

  Two near-identical categories no longer collide

  Two categories with almost the same name ('Drag and drop editor' and 'Drag-and-drop editor') were being merged into one page. They're now kept separate so each category's numbers are correct.

• 2026-05-15reviews

  Review 'helpful' counts confirmed unavailable

  Shopify doesn't publish how many people marked a review as helpful, so that figure isn't available to us. We removed every place it had been referenced to avoid implying we have data we don't.

• 2026-05-10vulnerability

  More platform-owned apps excluded from vulnerability signals

  Apps published by large platform owners (e.g. Microsoft Clarity, Intuit Mailchimp, Pinterest, TikTok) aren't realistic challenge targets for an independent builder. We expanded the list of platform owners we exclude so those don't crowd out genuine opportunities.

• 2026-05-05scores

  Apps with too few reviews no longer rank

  Apps with only a handful of reviews don't have enough signal to score reliably. We now hold them back from the opportunity and risk rankings so unproven micro-apps don't appear as if they were established players.

• 2026-04-28pricing

  Duplicate pricing plans collapsed

  Some listings repeated the same plan more than once (for example a monthly and an annual variant of one tier), which inflated plan counts. We now count distinct plans only, so pricing comparisons reflect what a merchant actually chooses between.

• 2026-04-20history

  Seven years of history added (since 2018)

  We backfilled category history all the way to 2018, so growth arcs and the 'what launched / what faded' views reflect the full arc of the market rather than just the most recent months.

• 2026-04-14pricing

  Pricing read more reliably from listings

  Free-text pricing like 'From $19/month' or '$0.10 per order' was occasionally misread. We improved how we interpret it, so pricing headroom and benchmarks are more accurate.

• 2026-04-08categories

  Field-label noise removed from categories

  Stray labels that aren't real categories (formatting fragments, product codes, delivery-time strings) had leaked into the category list. We filter them out so the rankings only contain genuine app categories.

• 2026-04-02reviews

  Support-response timing normalized

  Developer replies were occasionally dated inconsistently, which skewed how responsive a developer looked. We normalized those dates so support-responsiveness signals are fairer.