Starting September 21 and rolling out store by store through September 23, Shopify began changing how Analytics counts a "session." Shopify's own documentation says the change can affect sessions, conversion rate, add-to-cart rate, reached-checkout rate, checkout conversion rate, bounce rate, pageviews per session, online store visitors, and searches — most of the numbers a merchant checks every day. Orders, sales, and customer counts are not affected.
What actually changed: previously, a session was cut off the moment the clock hit midnight UTC, no matter what the shopper was doing. Now, a session keeps running as long as the shopper stays active. A visit that skips a pageview entirely — say, clicking a cart link straight into checkout — now counts as a session too. Shopify also tightened how it filters out bot and system traffic. Shopify's help page states outright that the numbers can look different "without necessarily meaning that your traffic, orders, or customer behavior changed," and tells merchants to treat post-update data as their new baseline rather than compare it directly to what came before.
Why this is worth noting
This site covered two similar-looking stories before: on August 24, "Google briefly underreported Search Console impressions and clicks due to a logging bug," and on August 27, "YouTube redefined what counts as a view." This one is neither — it's not a logging error like Google's, and it's not a deliberate loosening of a revenue-eligibility definition like YouTube's. Shopify is redrawing the actual boundaries of what a session is — when a day ends, whether a pageview is required, what counts as a bot — in the name of accuracy. But the symptom looks identical in all three cases: a store owner opens their dashboard one morning and the numbers have simply moved.
If you run a store, it's tempting to see this week's conversion rate or bounce rate swing and immediately assume you did something wrong. But this particular swing comes from a change in how a session itself gets counted, not from any change in what shoppers actually did.
What's certain and what isn't — Certain: Shopify changed how it measures sessions over a three-day rollout from September 21–23, and its own documentation lists sessions, conversion rate, add-to-cart rate, reached-checkout rate, checkout conversion rate, bounce rate, pageviews per session, visitor counts, and searches as metrics that may look different. Orders, sales, and customer counts are unaffected (checked September 23, 2026). Uncertain: Shopify hasn't published how much any individual store's numbers will move, or in which direction — a store with heavy bot traffic might see visitor counts drop, while a store with many shoppers active past midnight might see session counts drop and pageviews-per-session rise. Beyond that, it's guesswork.
If your site gets 100 visitors a week
This project runs an information site, not a store, so it tracks search impressions and visitor counts rather than sessions or conversion rate. But the underlying principle doesn't care which platform you're on. As of August 3, 2026, this site's Search Console impressions stood at 65 with zero clicks — at that scale, even a sudden fifty-percent swing is impossible to tell apart, on your own, from a change in how something got measured. So the practical takeaway is the same regardless of tool: when a metric moves without explanation, the first move isn't to guess at a cause and start editing your pages — it's to check that tool's changelog or announcements page first. Sometimes, like Shopify here, the platform tells you upfront that the measurement itself changed. Sometimes, like Google's Search Console bug, it's quietly logged as a known issue. Either way, that's the place to look before you start suspecting yourself.