About
How these tools are built.
MarkFluent builds free tools for paid growth and organic search. There is no account, no email gate, and no attempt to turn a score into a sales call. The tools exist to make a campaign or a site easier to judge before money is spent.
The SEO audit
A static crawl with stated coverage.
The audit starts at the address you enter, reads robots.txt, and follows same-origin links to a maximum of 250 pages and five clicks deep. It checks status codes, indexability, titles and descriptions, content depth, internal links, structured data, and PageSpeed for the entry page. Unknown categories are excluded from the weighted total rather than scored as zero, and every report states what it reached and where it stopped.
A real audit is a server job. The parsed snapshot is stored for seven days; raw HTML is not retained. Read the audit privacy note for the data flow, or run the audit yourself.
The calculator
Decimal maths in your browser.
The advertising profitability calculator runs entirely in the tab. It uses the four campaign modes, keeps blank fields unknown, and never converts currencies — the INR and GBP switch changes the display symbol only. Summary and PDF exports are generated locally, and the inputs are not sent to a server.
The calculation guidedocuments the formulas, the contribution treatment, and the attribution limits. Thecalculator is free to use.
What we do not claim
Estimates, not guarantees.
The audit does not predict rankings, traffic, or revenue. The calculator models the figures you enter; it cannot see platform attribution, incremental lift, or cash timing. Neither tool replaces analytics, Search Console, or a conversation about the business. Both state their assumptions next to the result so a number is never mistaken for proof.
The tools are built and maintained by MarkFluent. For the wider practice, seemarkfluent.co.uk orbook a call.
Method notes
How the tools stay honest.
The calculator carries a formula version, and its guide shows the definitions, the worked vectors, and the cases where a metric is unavailable. The audit publishes its coverage, stop reason, and the rules behind each score. When a signal cannot be observed, the report says so instead of substituting a sample. Those choices make the output less dramatic and more useful.
Nothing here is a ranking guarantee, a traffic projection, or a substitute for first-party data. A score can show where a crawl found friction; it cannot show demand, margins, or whether a page will convert. The calculator can show contribution at the figures entered; it cannot see platform attribution or incremental lift.
The pages load no third-party analytics or advertising pixels. The calculator keeps its inputs in the tab. Audit jobs keep a parsed snapshot for seven days and discard raw HTML. The audit privacy note and thecalculator privacy policyhold the detail.
Sources are deliberately narrow: the target's own HTML and robots.txt, Google PageSpeed Insights for the entry page, and the figures entered in the calculator. No backlink, keyword, or traffic dataset is purchased. The copy and rules are reviewed when a rule or formula version changes; the current calculator formula is 1.0.0. Last reviewed 19 November 2026.
How the audit score is built.
Each reached page is checked against the rules that apply to its type. The categories are crawl and indexability, metadata, content depth, internal linking, structured data, and performance for the entry page. A rule with no eligible evidence is excluded rather than scored as a failure, so a site with blocked pages can still receive a score on the categories that were observable. Coverage is the share of scheduled pages that were evaluated; a partial crawl can produce a score, but the report states its denominator and stop reason.
The result is a weighted readiness heuristic, not a ranking model. It does not include backlinks, search demand, competitor positions, or conversion data. Two sites can share a score and need different work; read the rule list and the affected counts, not just the headline number.
Reproducing a result.
Run the same URL with the same limits and the static HTML should produce the same rule findings. PageSpeed is fetched live and can vary between runs, and a page that needs JavaScript to render may look thinner than it does in a browser. Snapshots expire after seven days, so export the PDF if you need a fixed record. For a comparison, both sides run sequentially under the same page, depth, and time budgets.
A useful review changes one thing at a time: fix the largest critical or high finding, re-run, and compare the affected counts rather than chasing the headline score. If the report is partial, raise the page or time budget before treating the result as a full picture.