No. SleekRank reads from your data source. If your sheet pulls in sample results from a metasearch via a script on a schedule, those flow through on the cache cycle. SleekRank does not scrape engines or call flight APIs directly. The pattern is a separate import job that updates the sheet, and SleekRank renders whatever is in the source after a cache flush.

Both page groups read from the same engines sheet. The pairs page group joins two rows at render time using the slug pair from a pairs sheet. A row edit propagates to wherever the engine is referenced after the cache cycle, including every pair page where the engine is product_a or product_b. The data layer enforces consistency that manual page editing cannot.

Define another page group with a different URL pattern and source, and use the same engines sheet as a join. A region page group at /flights/region/{slug}/ can filter engines by region_strength column. A per-route page group at /flights/{origin}-to-{destination}/ can join to a separate routes sheet and recommend engines that cover that route best.

No. The verdict is whatever is written in the sheet. SleekRank does not write content, it injects content. For longer verdicts that exceed a sheet's column comfort, store them in a separate JSON file keyed by engine slug and join at render time. The verdict text is yours, the render and routing layer is the responsibility of SleekRank.

Yes. Map an image URL column to og:image with the meta type. Each engine page can render a custom social card via that mapping. For dynamic per-engine OG images that overlay the engine name, coverage stats, and a price-alert badge over a styled background, pair with SleekPixel which renders OG images from data on the fly.

Update the supported_airlines and missing_airlines columns in the sheet. Every page that references the engine, including per-engine, every pair page where it appears, and any region roll-up, reflects the coverage change after the cache window. Coverage changes are the most common drift source on manual reviews, since contract changes rarely surface in editorial workflows.

Store mistake_fare_events as a JSON array keyed by date, route, and resolution. Render it as a list on the per-engine page via list mapping, and join it on pair pages to show comparative mistake-fare track records. A separate page group can target /flights/mistake-fares/ as a chronological feed driven from the same source.

Yes. Add a column for predicted_accuracy_window with a percentage and a sample size, plus a JSON time series of historical predictions versus actuals. Render the headline number on the per-engine page and the time series via a chart library, joined at render time, so readers can compare prediction confidence across engines on the pair pages.