Skip to main content

Analytics

Analytics is the data side of Discover — Google Analytics 4, Search Console, and your live experiments in a single workspace-scoped view, with AI-generated explanations layered on top of every chart so you don't have to hunt for what changed and why. It lives at /<workspace>/analytics.

📷 Screenshot: Analytics page with the six tabs across the top and the date-range / sync / settings controls on the right.

The page is gated to Discover-tier plans and shown by default to the Marketing role; other roles see it once they pin it.

Connecting an account​

The cog icon at the top right opens Analytics Settings, where you connect your GA4 accounts and pick which one is active. You can attach multiple accounts (e.g. one per property) and switch between them from the same selector — every tab on the page re-queries against the new account immediately.

If the page comes up empty the first time, that's the cue: open settings, connect GA4, and pick the account you want to look at.

Date range and sync​

To the right of the tab strip:

  • Date range — defaults to the last 30 days. Quick presets cover today, yesterday, 7d, 30d, 90d, year-to-date, plus a custom range.
  • Sync — pulls the latest data from GA4 / GSC. Tabs auto-load on first visit; the sync button is for forcing a fresh pull when you know something just changed (e.g. a campaign just launched).

Tabs​

Summary​

The default tab. A grid of headline metrics — total users, sessions, page views, average time on page, bounce rate — followed by a line chart of the dominant metrics over the selected range. Below the chart, AI insights highlight the largest week-over-week movers and explain why (a referring channel surged, a top page lost traffic, a new audience cohort appeared). When the dashboard is opened from a feature-detail view, an additional page insights block scopes the same metrics to the URL the feature touched.

Usage​

Top pages by users, sessions, and engagement. Useful for "which content is doing the work" — and the click-through to a page row opens its individual time-series alongside the page-level AI insight.

Audience​

Demographic and behavioral cohorts: country, device, browser, new vs returning. Same chart-plus-insight pattern.

Attribution​

Where users came from before they converted: referring channel, source/medium, campaign, landing page. The AI insight here calls out shifts in your channel mix (e.g. organic share dropped 8%, paid search filled in) which is the most common Discover signal for "what to invest in next."

Experiments​

Manage A/B tests against your live site. Each experiment lists its hypothesis, variants, sample size, current uplift, and confidence; click into one for the detailed results panel. Experiments deep-link by ID — ?tab=experiments&experimentId=… — so you can paste an experiment URL into Slack and land everyone on the same view.

Google Search Console: top queries, click-through rate, average position, indexing health. Use this to spot queries you're ranking for but losing clicks on (high impressions, low CTR), which usually feeds straight into a Roadmap item.

What flows out of Analytics​

  • Insight cards can be turned into Roadmap items in one click — the relevant metrics and the AI's explanation are attached as context for whichever agent picks the item up.
  • Experiment results can graduate a winning variant into a feature spec, or kill the experiment and document why.
  • Pinned charts travel with you — pin one to a chat to ask follow-up questions ("why did mobile bounce rate spike on Tuesday?") with the data already in context.