The Maze: Shopify is changing the way it counts store visits. Its session measurement update started on September 21, with rollout running through September 23. Merchants may see different traffic and conversion figures even when the underlying business has barely moved. The change focuses reporting on human activity and captures shopping journeys that older measurement could miss. The announcement gives no country-specific limit. For anyone judging a campaign or store redesign this week, the measuring stick deserves a look before the celebration.
A better conversion rate can arrive without another sale. Shopify’s update changes sessions and the rates calculated from them, including cart, checkout and bounce metrics. A session is a visit grouped from a shopper’s activity; change which visits count and the percentage that converts can move. Orders, sales and customer totals are unaffected by this measurement change. That distinction matters when a team is trying to explain a sudden performance jump. A reporting shift alone cannot establish that an advert brought better buyers or a checkout redesign worked.
The count can move in either direction. The revised measurement retains some journeys without a standard pageview, such as a shopper following a cart link straight to checkout. That can add sessions. Filtering identified bots can remove sessions; keeping continuous activity together across midnight UTC can also reduce the count. Sessions still end after 30 minutes without activity. These changes pull in different directions, so there is no universal uplift to promise. Store-level effects depend on the journeys and automated activity previously included.
Cleaning a report does not lock bots out of a store. Shopify’s filtering separates automated traffic for analysis while useful crawling can continue to support search discovery and link previews. Its classifier is conservative, and some sophisticated automation may remain undetected. “Human” is therefore a reporting category, not a guarantee that every included visit came from a person. Merchants gain a clearer way to inspect activity; they do not acquire a new security barrier through an analytics setting.
September comparisons need a visible break in the series. Historical records are not deleted or reprocessed, but filters can change how old sessions appear. Matching filters before and after the rollout still leaves a changed counting method. Bot classification also has an October 7, 2025 boundary, making older comparisons harder. Supported reports allow filter adjustments; some surfaces do not. Custom reports and apps using session data warrant review, while most merchants need no technical implementation work. The practical next step is to establish a post-update baseline and keep comparisons explicit.
Why it matters: Performance meetings turn measurements into budget decisions. If teams treat a new conversion rate as proof of a better campaign, they risk rewarding a change in counting. Mark the rollout in reporting, compare sales and customer totals, and ask which session definition a dashboard uses. Then judge subsequent campaigns against the new baseline. The useful gain is clearer measurement; any claim of better trading still needs evidence from the business itself.
