Shopify FAQ analytics tells you which answers shoppers bother to open. Product FAQs for AI Crawler records impressions, interactions, and an interaction rate, with a daily graph and a leaderboard of FAQ groups. Use a group that is seen and rarely opened as a rewrite candidate. An open is interest in the question. It is not an order, and the catalog does not turn it into revenue.
The dashboard shows total impressions, total interactions, and the interaction rate as a percentage. A historical graph plots the daily series. A table ranks the top 10 FAQ groups by impressions and shows each group's interaction rate. Lifetime counters sit on the group. Daily rows feed the graph. Read both: a lifetime total can be dominated by an old campaign, and the daily line shows whether a rewrite changed anything after you saved it.
What an impression counts
An impression is recorded once per distinct FAQ group shown on a page load. The client deduplicates that count. If the same group is on the page once, that load contributes one impression for the group, not one impression per question inside it. A shopper who refreshes can count again, because it is a new page load. A shopper who merely scrolls is not a separate impression event beyond that load.
The group has to render to be counted. A product with no assigned FAQ does not produce an impression for a group that is not there. A shop-wide group counts when the page block shows it. Display still comes from the metafield. The analytics call is the request the storefront makes to log the event. Seeing a large impression number means the block was on pages people opened. It does not mean they read every answer.
What an interaction counts
An interaction is recorded when a shopper opens an accordion item or reveals more items with Load More. Those are the two events the catalog names. It does not say that a hover, a search keystroke, or a completed purchase is an interaction. The interaction rate is interactions relative to impressions, shown as a percentage on the dashboard and on each group in the top-10 table.
A high impression count and a low interaction rate can mean the questions are obvious, the answers are already visible in another template, or the headings are too vague to earn a click. Grid/cards and Boxed show more text at once than a closed accordion, so compare groups that use the same template before you decide one topic is "boring." A low rate on Accordion may be a heading problem. The same words on a card layout may not need an open at all.
- Impression: one per FAQ group per page load, deduplicated in the browser.
- Interaction: an accordion open or a Load More reveal.
- Interaction rate: the percentage the dashboard shows from those two counts.
- Leaderboard: top 10 groups by impressions, with each group's rate.
A rewrite loop
Sort your attention by the leaderboard, not by the FAQ you enjoyed writing. Take a group with impressions and few interactions. Read the questions the way a shopper would. If the heading repeats the product title, replace it with the objection: shrinkage, waterproofing, ship time, compatibility. Put the fact in the first sentence of the answer. Save the group so the metafield updates. Then watch the daily graph after the change, not the lifetime total alone. The lifetime counter still includes every day before the rewrite.
Also look at questions support still receives. Analytics cannot see the email that was never avoided. If shoppers open "What is the fabric?" and the inbox still asks about shrinkage, the answer they opened did not contain the shrinkage fact. Add it. If almost nobody opens a question and nobody emails it either, the question may not belong on the page. Delete it, or move a store-wide rule into the global group so the product list stays short.
What this dashboard does not measure
It does not measure orders, refunds, or support-ticket volume. A group can be opened often because shoppers are anxious, which is a reason to clarify the answer, not a reason to call the FAQ a conversion tool. It does not score whether the answer is true. There is no accuracy grade. A popular answer can still be wrong, and a wrong answer that people open is the urgent edit.
Search filtering is not listed as its own analytics event. You learn that someone opened a row or revealed more rows, not which word they typed. If you need the words, read the support inbox and the search terms you already collect elsewhere. The FAQ dashboard stays on impressions, interactions, the rate, the daily graph, and the top 10 groups. Creation limits are a billing check, separate from this report, and the catalog does not publish the monthly number.
Give a rewrite a few days of traffic before you judge it. The daily series is there so a single busy afternoon does not define the group. If impressions stay steady and interactions rise after the new heading, the question was the problem. If both fall, the block may be lower on the page than it was, which is a theme change rather than a wording failure. Note the date you saved the group so you know which part of the graph is the new copy. The listing is Product FAQs for AI Crawler.
Questions about FAQ analytics
Does an open mean the shopper will buy?
No. An interaction is an accordion open or a Load More reveal. Orders are not part of this report.
Why is the rate low on a card layout?
Interactions fire on accordion open and on Load More. A layout that already shows the answer can draw fewer opens. Compare like with like.
How many groups are on the leaderboard?
The dashboard table ranks the top 10 FAQ groups by impressions and includes each group's interaction rate.
The report is part of Product FAQs for AI Crawler. Pick one group the leaderboard shows people see, rewrite the question they are not opening, and check the daily graph after you save.