QSortby Guide
Results

Intent — what shoppers came for

What shoppers came for in the range you pick — buying a gift, replacing something, close to buying, browsing, treating themselves — how each kind behaves, what is trending with each, when each peaks, and what Intent AI is actually doing for your store.

Last updated · Nov 7, 2026

Intent suggest is invisible from the storefront by design — every shopper sees a slightly different order, so there is nothing to look at. Results → Intent is where it becomes visible: what your visits were actually read as, how each kind of shopper behaves, and whether the layer is earning anything.

Full on Growth. Pro sees the what shoppers came for reading — the layer runs in shadow there, classifying and reporting without touching your storefront. Recording runs on every plan, so the day you upgrade the history is already there. (Every new install runs on the Growth feature set for its first 10 days — see Your first 10 days.)

What shoppers came for

The share of visits in the range read as each of the five intents, with the count of visits read against visits in total. A visit is a 30-minute run of behaviour, and it is read once its behaviour points clearly at one intent — so a short visit, or one that stays ambiguous, is not read yet and shown in grey rather than forced into a bucket. If 90% of your read traffic lands in one intent, your intent keywords are probably too generic for your catalogue.

Under it, how each kind of shopper behaves: visits, product views, carts, purchases and revenue per intent. This is the table that tells you whether Close to buying is really closing, and how much a Browsing visit is worth to you.

Do the tags match what shoppers buy?

An order is evidence about the reading. A visit read as Buying a gift that ended in a gift-set order agrees; one that ended in a restock does not. Where enough checked visits exist — at least thirty — this card says how often the reading agreed with what the order showed, per intent. Treating themselves is read mostly from price and is deliberately left out of the headline number; it is a weak proxy, and the page says so.

Which products are rising with each kind of shopper: what gift-buyers are moving toward this week that they weren’t last week. New and Rising appear once the period before this one was read too, so they take a range or two to show up on a fresh install.

When it peaks

Visits by day across the range and by hour of day, coloured by intent, in your store’s own time zone. A store whose gift-buying clusters on weekday lunchtimes and whose browsing clusters on Sunday evenings has two audiences, and that is worth knowing before you write copy or schedule a popup.

What Intent AI is doing

The honest status card, per intent:

  • How products are chosen — By rules: a product’s intent tags against the shopper’s intent until a model has earned its place. When learned picks are on for an intent, it says so, and says what the check showed: that before going live the model beat the tags at putting what shoppers went on to click first, on visits it had not seen, by how much, and how sure that is. A model in training says it is in training and does not rank yet.
  • For You serves — whether intent reaches your storefront at all. It does so through a Personalized collection with Rank by intent on; the card tells you when there isn’t one, or when there is but the switch is off.
  • Measured lift — the result of the on/off test, or Not measured yet. The test switches intent off for half your shoppers on For You and every offer that ranks by intent, for a length you set, and reports how much more (or less) often the intent group bought, with a range. See Experiments.

Each section carries a short AI summary of what it is showing — a starting read, not a verdict.

Why it might be empty

  • A brand-new install has nothing until the theme app embed has been collecting for a while.
  • Traffic that lands, buys fast and leaves is often never read clearly, and shows as not read yet.
  • Visitors who haven’t granted analytics consent send almost nothing, so they can’t be read either. See the note in Set up QSortby.

If the page says no visit in this range was read yet, it is usually one of those three rather than a fault.

Menus are tucked away for bigger images — hover the edges to peek them, or Focus to pin. Click any image to zoom.

Popular

Set up QSortby — two steps Getting started Auto-sort a collection Auto-sort Create a personalization rule Personalization Upsell offers — the six touch points Upsell Troubleshooting Help
↑↓ to navigate ↵ to open esc to close