I'd log the exact query, engine, location, device, date, and the URL cited. The same query can return different answers, so one screenshot is a snapshot rather than a stable ranking signal. For a store, separate buyer-intent queries from informational ones and track qualified organic sessions, product-page add-to-cart, and revenue per session by group. A citation is useful context, but it isn't a conversion win by itself. If a competitor is cited, note the specific question and proof their page answers. Then make the equivalent product detail easy to verify on your own page and watch the relevant funnel step after it gets recrawled.

Before blocking, compare server logs with Shopify or analytics sessions and look for patterns: repeated hits to the same endpoints, no normal page flow, and no add-to-cart or checkout activity. A user-agent string alone is easy to spoof, so verify known crawlers rather than trusting the label. Then use the CDN or WAF to challenge suspicious traffic and watch whether performance improves, while keeping the search crawlers you want. I’d also segment suspected bot traffic out of reporting before judging conversion rate.