Lighthouse vs a conversion audit
Last updated:
A Lighthouse 100 is a speed and stability score. It is not a conversion. Core Web Vitals measure load, interactivity, and layout shift — not whether a paid click can see the offer and one primary action. Use Lighthouse for performance. Use SiteRevenue for the money path. Use both. SiteRevenue does not collect Core Web Vitals and will not invent them. Sample · free grade.
The dated argument is Lighthouse 100, conversion 0. This page is the evergreen comparison: when to run which tool, and what we refuse to fake.
Comparison table
| Lighthouse / Core Web Vitals | SiteRevenue | |
|---|---|---|
| Job | Is this page a reasonable citizen of the web? | Can a paid click convert on this public URL? |
| Scores | Performance 0–100; LCP, INP, CLS | Letter grade (A healthier than F); leak score (higher = more leaks); dimensions 0–10 (higher = healthier on that slice) |
| Sees | Timing, shift, unused JS | Hero, CTA, offer, proof, form, trust, mobile path in public HTML |
| Evidence | Lab or field traces | Page captures (fail-open), ranked findings, action box |
| Artifact | A shareable 100 | A file of record you can re-run — MD / JSON / PDF when the full job produces them |
| Re-run | A new lab or CrUX sample | A new immutable job on the same site → v1 Grade D → v2 Grade B |
| CWV | Measures them | Never invents them. No fake “Speed: 94” on the scorecard |
| Ads / SEO crawl | Not a campaign audit | Not an Ads login, not a technical SEO crawl |
A page can ship a 100 because the hero is a compressed image and almost no copy. That is a green lab and a dead fold. A 60 on performance with a clear offer + one CTA can still take the sale. Both can be true. They are not the same score.
Common gotchas
- Treating 100 as “the landing is done.” Dual equal pills, a slogan H1, and a form that switches the job do not turn LCP red.
- Lab on the wrong URL. Operators test the cached homepage on a fat pipe, then send spend to a builder URL with chat widgets and a hero taller than the viewport. The run was honest about the wrong page.
- Stripped tags in the lab, live tags on the URL ads hit.
- Assuming we ran PageSpeed Insights. We did not. A conversion finding about a clipped sticky or a hover-only nav is a path issue, not a fabricated vital.
- Inventing a speed number inside a conversion report so the scorecard looks complete. We will not do that. The mobile conversion guide is explicit.
- Skipping the fold sprint to chase a last four Lighthouse points on a page that still cannot answer what is this, who is it for, what do I do next?
How SiteRevenue differs
SiteRevenue scores the public URL you paste. Optional desired outcome and competitor URLs frame the job (full reports fetch those public rival pages; fail-open). We do not connect Google Ads. We do not scrape CrUX. We do not emit Core Web Vitals.
A full $19 page report is a conversion file:
- Letter grade + leak score (do not treat the leak number as a health ring)
- Dimension scorecard — new full jobs emit Hero, CTA, Message match, Offer, Proof, Form, Trust, Mobile (Message match omitted only when there is no ad copy). Older jobs stay omit-safe
- Page captures — fail-open; a missing shot is not a failed audit and not a fake Lighthouse clip
- “Fix these 3 this week” action box (effort S/M/L) + typically 5–8 ranked findings
- Dollar ranges are prioritization estimates, not guarantees
The free teaser is a grade, leak score, executive summary, and top 3 leak titles. Unlocking starts a new full job that captures the public page again. It does not reuse the teaser scrape. Session reports are noindex. Share links are read-only, not a public gallery.
Read the stack: How to read a revenue audit. Phone-fold rules: Mobile conversion.
When to use both
They are complementary. Collapsing them into one vanity 100 is how operators get a green dashboard and a red CPA.
SiteRevenue first when you pay for traffic and the fold might be the leak:
- Make the job obvious on the phone fold — offer + one primary, in HTML.
- Align the arriving promise with H1, CTA, and form eyebrow — message match.
- Ship the action box, then re-audit the live URL for a letter-grade strip. Empty re-runs waste a credit.
Lighthouse / CWV next on the same URL ads or search actually hit, once that path exists:
- Layout shift covering the CTA
- A hero image that is the LCP and also hides the offer
- Interaction delay on the real submit
Use both every week you buy clicks. A faster mismatched page is a faster bounce. A converting page that jumps 40px under the thumb still wastes mobile spend.
Public pages only. Not a chatbot. Not an accessibility certificate. Agent job webhooks are later — poll if you automate.
Related guides
- Lighthouse 100, conversion 0 — dated argument (not this table)
- Mobile conversion — phone fold; no invented CWV
- Conversion leak checklist — Mobile items 23–25
- What is a website revenue audit?
- How to read a revenue audit
- Overnight teardowns vs a file in this session — 24h PDF vs minutes
- Comparisons · All guides
Start with a free grade
Paste the public URL visitors actually land on (not only the lab homepage) at siterevenue.ai. No card. See the open sample (illustrative wireframes, including a clipped mobile bar a 100 does not catch). Full file: Pricing.