Core Web Vitals come in two flavours, and conflating them is how agencies end up arguing with clients about numbers. Crawl Cove keeps them separate and labels each honestly: lab data is synthetic and on-demand, field data is real users from the Chrome UX Report. Both live on the Pages / URL inspector, per exact URL.
Note
What you'll need: the Playwright chromium binary for the Lighthouse lab
(npx playwright install chromium, run once per machine), and a free CrUX API key for
field data. Everything else on the page works before either is set up. ~10 minutes.
Lab vs field — what each actually tells you
Lighthouse lab (synthetic). A single controlled run against one URL, on demand — you click the lab button on the Pages inspector and Crawl Cove runs Lighthouse in a clean chromium environment. It is never run during a crawl, so it never slows an audit or skews politeness. Lab data is great for diagnosing why a page is slow and for testing a fix before it has enough traffic to show up in field data.
CrUX field (real-user). Aggregated, anonymised measurements from real Chrome users over the trailing 28 days, served by Google's Chrome UX Report API. This is what Google actually experiences — and what feeds page-experience signals. Low-traffic URLs simply won't have a CrUX record, and Crawl Cove caches that honest "no data" answer rather than inventing a green score.
Tip
Use them together: field data tells you which pages real users struggle on; lab data lets you reproduce and fix the problem on demand, then re-run to confirm — without waiting 28 days for the field window to catch up.
Enabling the Lighthouse lab (chromium, once per machine)
The lab reuses the same Playwright chromium binary that powers the optional Tier-2 rendered crawl pass. Install it once:
npx playwright install chromium
Without it, the lab panel shows "Lighthouse skipped — install chromium" and the crawl still completes on Tier-1 data — nothing crashes. After installing, the lab button on the Pages inspector runs Lighthouse against that exact URL on demand.
Enabling CrUX field data (free API key)
Field data needs a Chrome UX Report API key. It's agency-global and separate from Search Console — one key serves every client. Exact steps:
- In Google Cloud Console, use the same project you set up for GSC. Go to APIs & Services → Library and enable the Chrome UX Report API.
- Go to Credentials → Create Credentials → API key. Then restrict the key to the Chrome UX Report API so it can't be misused elsewhere.
- In Crawl Cove open Settings and paste the CrUX API key. It's stored encrypted via safeStorage.
Field data caches for 24 hours per (URL, form factor) — desktop and mobile are tracked separately — so repeated inspections don't re-hit the API. Until the key is set, the field panel reads "CrUX API key not set" and, again, everything else on the page works normally.
Heads up
Restrict the API key to the Chrome UX Report API in step 2. An unrestricted key is a secret you don't want leaking, even though Crawl Cove only ever stores it encrypted and never logs it.
Reading the two panels on a client call
Open the Pages inspector for the URL in question. The CrUX field panel shows the real-user verdict (or an honest "no data"), and the lab button gives you a fresh synthetic run to dig into. Because both sit beside the page's findings, GSC metrics, and index status, you can answer "why is this page slow, and is it costing us traffic?" live, in one place — using only free data sources.
Next
- Inspecting a URL — the Pages inspector where both CWV panels live, alongside findings, queries, and index status.
- Running your first audit — install chromium once to unlock both the lab and the Tier-2 rendered crawl pass.