EngineeringScore

Free Semrush alternative

Semrush is a search marketing platform and this is not one: no keyword research, no backlink index, no rank tracking, no competitor intelligence. What EngineeringScore does is narrower and finishes in about a minute: one URL, read once, for its headers, its cookies, its weight, its SEO tags and its accessibility. It stands in for one slice of Site Audit, on one page, and for nothing else on the platform.

What we don't replace

  • Everything Semrush is a platform for. Its own product menu lists Technical Site Health beside Keyword Research, Competitor Analysis, Backlink Analysis, Rank Tracking and Content Creation. Site Audit is one item in that list, and it is the only one this scan touches at all.
  • Crawling a site. The Semrush knowledge base describes Site Audit as a website crawler that checks up to 20,000 pages in a single audit, and 100,000 on its Business tier. We read the single page you name and follow none of its links, so a problem on a page you did not name is a problem we never saw.
  • Anything that only exists between pages: duplicate content across URLs, broken internal links, hreflang on an international site, sitemap and robots coverage. The Semrush Site Audit issues list carries a check for each of those, including sitemap.xml not found, robots.txt not found and broken internal links.
  • Watching a site's health move. The Semrush Site Audit feature page describes a single health score for tracking progress over time and thematic reports that group issues into focus areas, and its knowledge base describes scheduled auto re-crawls on a daily or weekly basis. Every scan here is a snapshot with nothing behind it to compare against.
  • Getting an issue list into someone else's workflow. The Semrush Site Audit feature page describes sending issues straight to a dev team through a Trello integration, and its knowledge base lists Excel, CSV and PDF exports from Site Audit.
  • Anything behind a login, a paywall or a geofence. We read one page as an anonymous visitor, signed out, so a page that only exists after sign-in is a page this scan never reaches.

What we replace

  • One page, examined once, across all of it: what the response headers say, whether HTTPS and the redirect behave, which cookies were set and how they are scoped, how heavy the page is and which class of asset made it heavy, what the SEO tags contain, and what axe-core found in the markup.
  • Getting an answer without setting a project up first. Semrush organises Site Audit around a campaign you create for a domain and re-crawl, on demand or on a daily or weekly schedule, as its knowledge base describes. Here there is no campaign, no account and no schedule: one URL goes in, one report comes back.
  • Checks that run against what Chromium ended up with, not against the HTML the server sent. Cookies written by script, requests fired after load and assets injected at runtime all count here, because they are what a visitor’s browser actually had to deal with. Semrush lists JS rendering for Site Audit as available with a Guru or Business SEO Toolkit subscription in its knowledge base.
  • An accessibility run on the page you scanned, in the same report as the SEO and header findings. The Semrush Site Audit feature page places accessibility management under Enterprise Site Intelligence, alongside JavaScript rendering and crawling for millions of pages.

Where Semrush is stronger

  • Semrush does the site-wide job and does it thoroughly, which is a different job from this one. Its knowledge base says Site Audit runs over 140 on-page and technical SEO checks, from duplicate content and broken links through to AMP, HTTPS implementation, hreflang, crawlability and indexability.
  • Semrush audits at a scale this scan does not approach: its knowledge base gives Semrush Site Audit a crawl budget of 100,000 pages per month on the Pro SEO Toolkit, 300,000 on Guru and 1,000,000 on Business, against our one page per scan.
  • Semrush runs Site Audit for an account with no SEO Toolkit: the Semrush knowledge base lists 100 pages per month at that level, which for a small site is the whole thing crawled rather than a single page of it. The same page says the tool itself comes with the SEO Toolkit or Semrush One, so that allowance is the floor, not a separate free product.
  • Semrush brings data no anonymous page load contains. Its Site Audit feature page describes health scores compared against your industry, and lists of pages blocking AI crawlers, naming ChatGPT-User, OAI-SearchBot, Googlebot, Google-Extended, Perplexity-User, PerplexityBot, Claude-User and Claude-SearchBot.
  • Semrush measures Core Web Vitals and we do not measure them at all. Its knowledge base says Semrush Site Audit analyses them on the first ten pages of a crawl that return 200 OK, a fixed set kept stable so the trend means something; our load-time figure is a single unthrottled cold-cache visit from one machine, which our own speed test page calls an ideal case rather than a simulation of a slower connection.

Where we're stronger

  • There is no project to create, no domain to add and no account to make. A public URL is the whole input; Chromium loads it on our side and the report is back inside a minute.

  • Findings arrive as prose rather than as a row you have to decode: what the thing is, why it matters, what to change, and a link on to a written guide for that check where a dedicated one exists.

  • A check we could not settle says so instead of quietly scoring, and nothing in that state moves the score in either direction. Fifteen checks can land there, spread across security, performance, accessibility, SEO and the redirect probe: a page whose asset sizes never arrived is not called light, and an accessibility engine that failed to run is not called clean. Named checks in named situations, incidentally, and not a habit the scanner has everywhere: a Referrer-Policy carrying a value we do not recognise passes on the header being present, since inventing a failure for a value this scanner cannot classify is the same error facing the other way. Against a platform export that distinction is the entire point, because a spreadsheet cell reading pass and one that ought to read we could not tell look identical in a column, and only one of them is safe to act on.

  • We publish the edges. Every one of our free tool pages carries a section on what that scan cannot see, check by check, including the ones where we test only that a header is present and never look at its value.

Sources

Every claim about Semrush on this page was read on from Semrush's own public pages. Their product changes and this page does not change with it, so treat the date as the claim's expiry, not its publication.

Ready to see where you stand?

Scan your site and get your Engineering Score with a prioritized roadmap in under a minute, no signup required.

Scan your site