Device comparison
When a crawl runs across desktop, mobile, and tablet, the Devices view compares scores and issues between them.
When a crawl covers more than one device, the Devices view puts them side by side. You can see where a page behaves differently on mobile than on desktop, which matters because Google indexes and ranks the mobile version. A page that scores well on desktop but drops content, links, or structured data on mobile is being judged on the weaker render.
Spider can crawl in desktop, mobile, and tablet modes in one pass. Mobile is always the lead: it's crawled first because Google indexes mobile-first. Desktop and Tablet are added to compare for indexing gaps. Each device runs as its own crawl, so one audits while the others queue for a free slot.
What runs once vs per device
Not everything is device-specific. Site-level foundation is shared across every device; the on-page audit runs separately for each viewport.
| Layer | Scope | Examples |
|---|---|---|
| Site architecture | Once per crawl, shared by all devices | SSL, robots.txt, sitemaps, llms.txt, crawlability |
| On-page / page score | Per device | Content, headings, links, schema, images, Core Web Vitals |
This split is why a site can show one Site Architecture score but different page scores per device. The foundation is sound, yet the mobile render drops content the desktop render keeps.
What it compares
- Category score deltas — each category scored per device, with the difference between renders highlighted so parity gaps stand out
- Mobile-specific issues — findings that appear only on the mobile render, such as content or links present on desktop but missing on mobile
- Performance comparison — Core Web Vitals side by side, where mobile typically lags desktop and is the version Google measures
The deltas to act on are the large ones. Small score differences between renders are normal; a category that is several grades worse on mobile usually means content or markup is being dropped from the render Google actually indexes.
How each device is crawled
Each device uses its own user agent and viewport: Chrome Mobile for the mobile
render, Chrome Desktop for desktop. If a page or robots.txt returns a 403,
Spider retries once as the device-specific Googlebot before reporting it blocked.
These defaults are set per project and can be overridden under Advanced →
Per-device user agent & viewport. See The crawler for
the full option set.
Multi-device crawling needs a paid plan. The free Scout tier is mobile-only, so device comparison is a Tracker/Leader feature. See Plans & tiers.