Field data in.
Ranked fixes out.
VitalLens reads the Core Web Vitals your real visitors experienced, runs a full Lighthouse audit alongside it, and tells you which single fix moves the number most — ranked by impact against the effort it takes to ship.
Google grades the two separately, so pick the one you care about. Switching device on the report re-runs it either way.
- 5 scans a day, no account
- Mobile and desktop
- Field and lab in one report
A score is not an answer.
Most tools hand you one number from one simulated load. Your visitors are not a simulation — so we start with them, then use the lab run to explain what they felt.
What really happened
A 28-day rolling window of measurements taken on real Chrome visits to your URL — the same dataset Google uses when it decides whether your page passes. Real devices, real networks, the 75th percentile rather than a flattering average.
- Tells you whether you pass, and for whom
- Needs enough traffic before Google will report it
Why it happened
One controlled load on a throttled mid-range phone, with the full trace attached: which element was the largest paint, which request held it up, what shifted and by how much. Available for any URL, including the page you shipped ten minutes ago.
- Names the element, request and byte count to blame
- One machine, one moment — not your audience
When a URL has no field data yet, we say so plainly and give you the lab audit — no spinner that never resolves, and no invented number standing in for your visitors.
Five steps, and then again next week.
Performance is not a project you finish. VitalLens is built around the cycle teams actually run.
Measure
Scan any URL for field and lab vitals, on mobile or desktop.
Diagnose
Read a verdict in plain words, with the critical issues named.
Prioritize
Take a roadmap ordered by impact against effort, not by audit ID.
Monitor
Pulse re-scans on a schedule and emails you when a metric slips.
Communicate
Share a locked report link, or export text and PDF for the review.
Three things a speed test cannot do.
Running the audit is the easy part. These are the parts that change what your team ships next sprint.
Performance Pulse
Put a site on a schedule and stop checking manually. Pulse re-scans, compares against the last snapshot, and only writes to you when something actually regressed.
Competitor Eye
Pin the domains you are measured against and see the gap per metric — where you lead, where you lag, and what they are doing on the page that you are not.
AI Consultant
Every finding comes back as a fix you can hand to an engineer: the root cause in one paragraph, the change to make, and an honest note on what it will and will not buy you.
Against running PageSpeed by hand.
PSI is a good tool and we use its data. The difference is everything that happens after the number arrives.
| VitalLens | PSI by hand | |
|---|---|---|
| Data sources | Field and lab, one report | Field and lab, one page at a time |
| History | Kept per site, charted | Gone when you close the tab |
| Regressions | Scheduled re-scan, email alert | You have to remember |
| What to do next | Ranked by impact and effort | A list of audits to read |
| Competitors | Pinned and compared | A spreadsheet, if you keep one |
| Sharing | Locked link, text or PDF | A screenshot in Slack |
Before you run one.
Do I need an account?
Not to run your first scans — paste a URL and read the report. An account is what keeps the history, lets you put sites on a schedule, and pins competitors.
Where does the data come from?
Google’s Chrome UX Report for the field metrics and PageSpeed Insights’ Lighthouse run for the lab audit. We do not ask you to install a script or a tag on your site.
My URL has no field data. Now what?
CrUX only reports a page once it has seen enough Chrome traffic. When yours has not, we tell you instead of stalling, fall back to the origin-level data where it exists, and the lab audit still runs in full.
How is this different from PageSpeed Insights?
It reads the same sources. What it adds is memory — history per site, scheduled re-checks, competitor comparison, and a ranked fix list instead of an audit dump.
Can I share a report with someone who has no account?
Yes. Any report can be shared as a point-in-time link, or exported as plain text or PDF for the people who only ever want the summary.
Scan one page. See what your visitors actually get.
No tag on your site, no credit card. If the answer is boring, you have lost a minute.