Crawl every page. Fix what costs traffic first.
Prism fetches and Lighthouse-scores every page, then sorts every finding by what it is actually doing to your traffic: broken pages still taking clicks, traffic that stopped and can be recovered, upside the site is sitting on, and plain housekeeping.
See what broke, and how much of the site it touches.
Every page gets fetched and scored, not a sample, so a finding always comes with the share of pages carrying it.
- Issues
- Every finding grouped by severity, with the share of pages affected.
- All Pages
- Every crawled page with its status, scores and findings in one table.
- Content & On-Page
- Thin content, near-duplicates, and titles or descriptions that are missing or reused.
The top of the list is the work that pays.
Fix First classifies the same findings by what each one is doing to your traffic. Connect Search Console and every finding shows what it is doing to real clicks.
- Classified by effect on traffic
Findings sorted by what each is doing to the clicks behind them, not by how alarming they look.
- Broken while earning
Pages still taking clicks while returning an error or carrying noindex, with the click counts.
- Severity still ships
Issues keeps the classic ordering beside it, so both questions have a report.
Severity is not the same thing as cost.
A crawler grades a finding by what kind of problem it is. Fix First asks what the finding is doing to the traffic behind it. A critical error on a page nobody visits is housekeeping.
By severity
- Missing titlecritical
- Client errors (4xx)high
- Slow pageshigh
- Missing canonicalhigh
- Thin contentmedium
By effect on traffic
Broken while earning
broken pages still taking clicks
← Client errors (4xx) · high
Recover
traffic that stopped, recoverable
from traffic history, not from the crawl
Earning despite issues
upside the site is sitting on
← Slow pages · high← Thin content · medium
Tidy
housekeeping, no traffic behind it
← Missing title · critical← Missing canonical · high
Run it again and see what the release changed.
Every run does the same work on every page, so two crawls are comparable.
- Crawl Comparison
- Two runs side by side: pages, health score and every issue type, diffed.
- Live Monitor
- Watch a crawl work through the site while it runs.
- Performance
- Lighthouse's four scores on every page, so a slow template is a list of URLs.
Fourteen reports off one crawl.
The whole suite, grouped by what you open each report to find out. One crawl feeds all fourteen, and a crawl costs no credits.
Defined in the glossary: Fix First, X-Robots-Tag, internal PageRank.
What the crawl found
The crawl as a record, from the whole site down to a single page.
- Overview
- The health score and the category scores under it, the status codes the crawl met, and the issues affecting the most pages.
- Issues
- Every issue type grouped by severity and category, each with the number of pages it affects and a panel saying what it is and why it matters.
- All Pages
- Every crawled page in one table, with far more columns than fit on a screen and presets to bring in the rest.
- Page Analysis
- One page read all the way down: its findings, the queries bringing it clicks, its markets and devices, every internal link pointing at it with the anchor text used, and how it has changed across recent crawls.
What to do first
The findings ordered by what each one is costing you.
- Fix First
- The same findings as Issues, banded by their effect on traffic: broken while still earning, recoverable, earning despite issues, or tidy. Findings that could not be priced are listed rather than quietly dropped.
- Coverage Gaps
- The crawl and Search Console read together: earning pages nothing links to, ranking pages missing from the sitemap, and internal equity pointed at pages nobody arrives at.
- Crawl Comparison
- Two crawls side by side: the issues introduced, the issues fixed, and how the health score moved between them.
Report by report
Each area of the crawl in full.
- Content & On-Page
- Word count and readability distributions, title and description lengths against their bands, the thin pages, and image weight by page and by format.
- Internal Linking
- Click depth against inlinks, orphan pages, least-linked pages, broken internal links with the pages they come from, the anchor text you use most, and your internal PageRank leaders.
- Link Opportunities
- Ranked internal links worth adding, anchor text by target, the pages leaking the most followed equity, and the external domains receiving it.
- Performance
- Average Lighthouse scores, Core Web Vitals page by page, performance against load time, the audits offering the biggest savings, and the slowest pages in the crawl.
- Technical
- Indexability, canonical tags, sitemap coverage, redirect chains, schema types and the markup that fails validation, hreflang, security headers, page weight, server and protocol fingerprint, where crawl budget goes, and a live robots.txt check of which AI crawlers are allowed to read the site.
The crawler itself
What it is doing now, and what it is allowed to do.
- Live Monitor
- The running crawl as it happens: what it is fetching, how much queue is left, and the URLs that failed.
- Crawl Settings
- Page limits, the URL source, JavaScript rendering, robots.txt handling, subdomain scope, and the crawler identity every request will present.
- reports over one crawl
- 14
- reports over one crawl
- Lighthouse scores on every page
- 4
- Lighthouse scores on every page
- properties on the free plan
- 10
- properties on the free plan
- credits used by a crawl
- 0
- credits used by a crawl
Elsewhere in Prism
One report with a page of its own
Frequently asked questions
Why does Fix First order things differently from Issues?
Issues groups findings by severity, which says what is broken. Fix First classifies the same findings by what each is doing to your traffic. A broken page still taking clicks comes first; a critical error on a page nobody visits is housekeeping.
What does Fix First need to classify by traffic?
Search Console connected. With it, findings are classified by their effect on real clicks. Without it, you get a severity list, plainly labelled.
Can I see what a release changed?
Crawl Comparison diffs two runs and shows the movement in page count, health score and every issue type between them.
The rest of the platform
- AI Visibility Monitor brand mentions across ChatGPT, Claude, Gemini and Perplexity: prompts on a schedule, every answer kept in full, every cited source listed. 18 reports.
- Search Performance 17 reports over your Search Console data: queries in striking distance, cannibalised pages, position changes, month on month, and the day Google swapped the page that ranks.
The first Fix First list usually pays for the afternoon.
Point Prism at a site and it crawls every page, scores each one, and shows what each finding is doing to the traffic behind it.