Docs · 01

Connecting your data

NavSignal builds its map by joining what Google Search Console and GA4 already know about your site with a crawl of the pages themselves. This is what connecting each one actually requires, what NavSignal does with it the first time, and when a warehouse connection like BigQuery enters the picture.

What connecting requires

Both Search Console and GA4 connect through Google's own OAuth screen, the same consent flow you'd see connecting any Google tool. NavSignal requests only thewebmasters.readonlyand analytics.readonlyscopes: read-only for both, which permit pulling the clicks, impressions, position, and session data those accounts already report, but never permission to change a setting or a report in either product. If you manage properties for multiple clients, you grant access per property the same way you would for any other read-only integration.

What happens on the first audit

Once a site is connected, an audit runs in a fixed order: NavSignal crawls your site first, so every page has a link graph, a word count, and its reachability from the rest of the site before anything else is layered on. It then pulls in keyword ranking data and runs the five pattern detectors against whichever sources have data for this run. Orphan-page and keyword-cannibalization already fire this way today, from the crawl and keyword data every audit gathers. Ranking-opportunity and the two engagement patterns are built to read Search Console and GA4 data the same way, but the pull that writes a customer's own Search Console and GA4 numbers into a run has not shipped yet, so connecting either source does not yet light up its pattern; nothing else about the run fails or waits on a pattern with no data. On a connected workspace, the last step writes a plain-language action brief for each finding before the run is marked complete.

What BigQuery adds, and why only Full needs it

Free and Lite are designed to need no warehouse setup at all: once the GA4 pull ships, both tiers will read entirely through the GA4 Data API, the same read-only reporting layer as your GA4 connection above. BigQuery is a separate, optional connection: it queries your own GA4 BigQuery Export, the raw event stream Google can export into your warehouse. That export is granular enough for NavSignal to measure the actual page-to-page path a visitor took, not an estimate reconstructed from session totals. That measured journey-edge map, and the one issue pattern that depends on it, are Full-only, because reading it requires a Google Cloud project and a dataset the two lighter tiers were never asked to require.

Full details on what NavSignal does and does not do with any of this access are on the Trust page.