About WebDiagnosis
Built because qualifying a prospect list is the slowest hour of the week
One scanner, a narrow promise, and a rule that every published number arrives with its denominator. This page says what it does, what it will not claim, and where the figures come from.
Who builds it
Hasnain Mehdi
Founder, WebDiagnosis
Works in search: technical and on-page SEO, content publishing, and the off-page work that produced this tool. WebDiagnosis exists because qualifying a prospect list was the one part of a link campaign that nothing on the market did well and everybody did by hand.
He also runs SEO Maxing, an SEO practice, and operates content sites of his own. The editorial and Search Console process this site publishes under came out of running those, not out of a course.
He wrote the scanner, and hand-checked the 1,001-row study fifty rows at a time before a single figure from it was published.
The job it does
The reading, not
the judgement.
WebDiagnosis started as a private script for work that was being done by hand, one browser tab at a time.
A purchased prospect list arrives, and before anything useful can happen someone opens every URL, scrolls to the bottom, looks for a comment form, views source to see what the existing comment links carry, notes the Domain Rating, and decides. Two to four minutes a page when nothing goes wrong.
On a thousand-row list, most of that work returns a no. That is the part worth automating: not the judgement at the end, which still needs a person, but the reading that gets you to it.
What it does differently
A narrower claim,
and the only useful one.
Most tools here describe domains, because domain data is what can be bought in bulk.
It opens the page you gave it
Not the domain, not the homepage. The exact URL, reported as it is right now. That is a smaller claim than a backlink index makes, and it is the one that changes what you do next.
The scope is deliberately small
Comment-link opportunity, on pages you already hold. It does not find prospects, and it does not qualify guest posts, directories, forums or editorial placements.
Two more tools are being built
Ordering a list by relevance, and checking months later whether placed links held. Both are listed as planned with no date. Nothing here describes them in the present tense.
Where the numbers come from
Measured here,
including the bad news.
Everything published on this site comes from running the shipping scanner on real lists and writing down what happened.
That last one is the useful half, and it is not flattering to the metric most of this industry sorts on: on that list, Domain Rating ordered the prospects exactly backwards. It is on the homepage rather than buried. Read the study and its method, or check the working in the methodology.
Operating rules
The five rules the
product is built on.
These decide what gets built and, more often, what does not.
The exact page, or nothing
Domain metrics describe a domain. Whether a comment can be left is a property of one page and one template. If we cannot read the page, we say so instead of substituting the domain average.
Deterministic, not predictive
Regular expressions, CSS selectors and arithmetic. No model sits in the scoring path, which is why the same page scores the same way every run and why the weights can be printed in the interface.
Unknown is a result
A check that found no evidence reports Unknown. Filling that gap with a platform default would put an assumption in a column that looks like an observation, and you could not tell them apart afterwards.
Every number carries its denominator
A percentage without its sample size, its date and its limitation is not a finding. On this site they travel together, including when the number is unflattering.
It reads, it never acts
No comment is posted, no account is signed into, no CAPTCHA is answered. The tool is a reader. That boundary is not a feature gap, it is the design.
Corrections
If a figure here is wrong,
say so.
Corrections are made on the page with the date of the change, rather than quietly. That has already happened once.
Write to support@webdiagnosis.net. Support, corrections and data requests all reach the same inbox. There is no editorial team; where a claim is not first-hand, the page says so, and every page names who wrote it.