Skip to main content

Running your first scan

A scan is Site Health's core primitive. It runs Google Lighthouse against every published page of a Webflow site and stores the results.

Trigger a scan​

From any site detail page:

  1. Click Run Scan in the header
  2. Choose a strategy:
    • Mobile — faster, covers the CWV metrics that affect Google ranking
    • Desktop — useful if your site is primarily used on desktop
    • Both (recommended) — scans both strategies, takes ~2x as long
  3. Click Start scan

Watch the progress​

A toast in the bottom-right shows live progress via Server-Sent Events:

  • Pages scanned / total
  • Estimated time remaining
  • Option to email me when it's done (handy for big sites)

Scans typically take:

Site sizeMobile onlyBoth strategies
5 pages~1 min~2 min
20 pages~3 min~6 min
50 pages~7 min~14 min
100+ pages~15 min~30 min

Exact time depends on Google's PageSpeed API load. Site Health rate-limits at 3 req/sec to stay within quota.

What gets scanned​

  • Every published page in your Webflow site (published pages come from Webflow's sitemap)
  • Including CMS collection template pages — we resolve them to a real live item URL so Lighthouse scans actual content, not an empty template
  • Excluding password-protected, 404, or draft-only pages

What gets collected​

Per page, per strategy:

Scores (0-100)​

  • Performance
  • SEO
  • Accessibility
  • Best Practices

Core Web Vitals (lab + real-user)​

  • LCP (Largest Contentful Paint)
  • CLS (Cumulative Layout Shift)
  • INP (Interaction to Next Paint)
  • FCP (First Contentful Paint)
  • TTFB (Time to First Byte)
  • TBT (Total Blocking Time)
  • Speed Index
  • Time to Interactive
  • Chrome UX Report (CrUX) data for LCP/CLS/INP — real user 75th percentile

Diagnostic data​

  • Top 5 Lighthouse opportunities with estimated savings
  • Per-page LCP element (which element is the slowest contentful paint)
  • CLS culprits (which elements cause layout shift, with scores)
  • Top 15 third-party scripts by blocking time + transfer size
  • All failing accessibility audits with severity
  • All failing Best Practices audits
  • All failing SEO audits
  • Performance diagnostics (DOM size, bootup time, etc.)
  • Stack packs (framework detection: Webflow, React, Next.js, etc.)
  • Final screenshot of the page

After the scan completes​

  • Toast notification with the final Health Score
  • Dashboard + site detail instantly reflect the new data
  • Trend charts update (after 2+ scans exist)
  • If any regression was detected → regression alert email fires
  • If the Health Score dropped below your alert threshold → alert email fires
  • If any alert rules triggered → alert email fires

What if a scan fails?​

Scans can fail for a few reasons:

  • Google rate limited us — we retry 3 times with exponential backoff automatically
  • Your site is down / unreachable — the scan marks as failed and you can retry
  • Worker restart — mid-scan job got interrupted; we'll retry with idempotent cleanup

Failed scans show a red Failed badge in the scan history. Click Run Scan again to retry — each scan is independent.

Next: set it on autopilot​

Now that manual scans work, schedule them so you never need to trigger one again: