lala SEO For disciplined marketers

See what is wrong with your site in about a minute.

This is a real report — not a mock-up, and not our own site chosen for flattering it. Here is how it was produced:

  • We crawled all 39 pages urcalculator.com links to, in 20 seconds
  • Ran 101 checks, 48 of them quoting Google's own documentation
  • Went back to the live site and re-checked a sample of pages behind each finding

Free · no account required

The top of a real Lala SEO report for urcalculator.com: a triage score of 99 out of 100, counts of 39 pages, 0 critical, 2 warning and 2 info, a table of what was checked, and charts of response codes and time to first byte.
It continues well past this crop.

What every audit gives you

Verification

Every finding re-checked against your live page.

Most SEO tools never check whether the fault is real

So you spend a weekend fixing things that were never broken, while the one that is costing you traffic stays open.

  • Confirmed — we re-fetched the page and saw it again
  • Not verified — said plainly, never dressed up
A finding marked 'Confirmed against the live site' with the date and time it was re-checked.
The verification line, on every finding.

Triage score

One number that orders your work queue.

Forty findings in a list, and no idea which one matters

You fix the easy ones, feel productive, and leave the expensive fault open for another month.

  • Severity counts rank your queue before you start
  • Not a Google metric — the report says so itself
A report header showing a triage score of 99 out of 100, with counts of 39 pages, 0 critical, 2 warning, 2 info, and a 20 second runtime.
39 pages, two warnings, twenty seconds.

Coverage

What was checked, not only what failed.

A tool that only lists failures hides what it never looked at

You cannot tell "clean" from "not checked", so you pay someone else to check the same thing again.

  • Every area reports a verdict, pass or fail
  • With how many pages it covers
A 'What we checked' table listing Technical SEO, AI Overviews eligibility, crawl coverage and content opportunities, each with a verdict and the number of pages it covers.
Four areas, each with its own verdict.

Schema matches the page

Text in your markup, checked against the words on the page.

A validator says your markup is valid, not that it is true

Google's rule is that markup must match what readers see. A validator never opens the page, so it cannot catch a claim your page never makes.

  • Compared on words, never on formatting
  • A whole section at once
A finding headed 'Structured data claims text the page does not say', marked warning and affecting 3 of 3 pages. It quotes Google's policy, explains that formatting differences are not reported, and lists the three addresses where the markup and the page disagree.
The claim, and the addresses it was found on.

AI Overviews readiness

Google's published conditions, checked one by one.

Everyone is selling "AI optimisation" they invented

You buy schema Google never asked for, and still cannot be quoted because one directive on the page blocks your snippet.

  • Google's published conditions, quoted word for word
  • Not a score — nobody can measure being cited
An AI Overviews readiness checklist. Each condition is ticked and quotes Google's own wording, with a note that the source page is re-read word for word so the section cannot go out of date.
Re-read automatically, so it cannot go stale.

Speed

Response codes and time to first byte, from your pages.

A lab score will not tell you which pages are slow

Visitors leave before anything paints, and the crawler meets the same wait on every page it fetches.

  • Response codes across every page crawled
  • Time to first byte, bucketed, with the median
Two charts: a bar showing response codes across crawled pages, and a histogram of time to first byte with the median and slowest page labelled.
Server response only — the chart says so.

Progress

Every audit kept, so you can see what changed.

You fix things and never find out whether it worked

So the same fault comes back, or a deploy quietly undoes last month's work and nobody notices for a quarter.

  • Side by side with the two runs before it
  • Better or worse, called out per component
Three audits compared side by side, each with its score ring, and a status column marking each component as worse, better or unchanged.
Three runs, with the deltas called out.

Crawl scope

Whole site, one page, or one section.

An audit of the wrong pages tells you nothing

Crawl a whole shop when you changed one blog, and the finding you needed is buried under four hundred.

  • Whole site, one page, or one section
  • robots.txt obeyed by default
A crawl summary panel listing what the crawler reached, the robots.txt and sitemap it read, and the redirects it followed.
How the crawl was scoped, and what it met

Page evidence

Every page crawled, with the data behind each finding.

A finding you cannot check is a finding you cannot act on

Your developer asks which pages, you cannot say, and the ticket dies in the backlog.

  • Status, title, words, schema, canonical, response time
  • JSON export of every page crawled
A table of crawled pages, each row showing the URL with its status code, title length, word count, H1 count, schema types, canonical and response time.
The evidence behind every finding.

The 101 checks, and what each one costs you

Indexing

Canonicals, redirects, orphan pages and soft 404s.

Pages Google never sees

  • Canonicals and redirects — the wrong one hides a page
  • www, non-www, http, https — four sites competing as one
  • Orphan pages nothing links to
  • Soft 404s that look fine

Titles and snippets

What the searcher sees before deciding to click.

Pages nobody clicks

  • Titles measured in pixels, as Google cuts them
  • Descriptions cut before the offer is read
  • Headings that answer no question

Delivery

Core Web Vitals, compression and what Google cannot load.

Pages that cost you to serve

  • Core Web Vitals from real visitors
  • Missing compression — bytes paid for twice
  • Blocked resources Google cannot load
  • Domain expiry — the deadline that ends the site

Platform quirks

What Shopify and non-Latin scripts break for you.

Faults your platform creates

  • Shopify — the duplicate URLs it makes
  • Indic scripts — Hindi, Tamil, Bengali and six more

Keyword rankings

Every term the site ranks for, and the page that earns it.

You cannot fix a page when you do not know what it is for

So you rewrite a page that worked, and leave real demand unanswered.

  • Position, volume, difficulty, intent
  • The page that earns it
A keyword table for urcalculator.com: 196 keywords ranking, 0 on page one, 73,120 monthly searches, and rows showing each term's position, monthly searches, how often it is asked of an AI, a twelve-month demand sparkline, difficulty, intent and the page that ranks.
196 terms, 73,120 searches a month, AI demand included.

Backlinks

Referring domains, authority and spam score.

A big backlink number can be worth nothing

You quote the count to a client, and most of it is links Google already ignores.

  • Referring domains, authority and spam score
  • What to do about it — usually nothing, and we say so
A backlinks panel showing 314 referring domains, 318 backlinks, 0 broken links, authority 0 and a spam score of 43 marked 'mostly low-quality links', with an explanation that Google ignores such links rather than penalising them.
318 links, and the reason the number flatters you.

AI mentions

Whether ChatGPT and Gemini name you in their answers.

Answers are given about you without your page

ChatGPT and Google's AI answer your customers today, and nobody tells you whether you were in the answer.

  • Answers already given, not a question we asked for you
  • "Not found" means not found — never a failed check dressed up
An AI mentions panel reading 'Not found in any AI answer they have collected', with a note explaining this is a record of answers already given, so it means the site did not appear rather than that the check failed.
A null result, reported as a null result.

Competitors

Who shares your searches, and which are real rivals.

Half the sites you think you compete with are not rivals

So you benchmark against Wikipedia and never find the site your own size that you could beat.

  • Institutions and giants marked as such, not counted as rivals
  • Closest to your size named, because that is the winnable one
A competitors table listing sites that appear in the same searches, with canada.ca marked 'institution — not a real rival' and youtube.com marked 'everywhere — not a real rival', each row showing shared searches, their average position and how many searches they rank for.
canada.ca and youtube.com are not rivals. It says so.

What we will not tell you: where you will rank. Nobody outside Google can see how ranking is decided. We check what Google has published, and say plainly which findings we verified and which we did not.

Run a free audit See pricing