Technical SEO is unglamorous and almost invisible to the business owner paying for it. Nobody notices a correct canonical tag. Everybody notices a new homepage design. That asymmetry is exactly why these faults survive for years on otherwise well-run websites.
Below are the ten we find most often in audits, roughly in order of how much damage they do. Several of them we have found on our own site over the years, which is how we know how easily they hide.
1. Every page pointing its canonical at the homepage
The canonical tag tells Google which URL is the authoritative version of a page. When a template hard-codes it, every page ends up declaring that the homepage is the real version of itself.
The effect is brutal: you are explicitly instructing Google not to index your service pages, your pricing page and your blog. A site with thirty pages behaves like a site with one.
How to check: open any inner page, view source, search for rel="canonical". The URL should be that page’s own address. If it says the homepage, you have found your problem.
2. Non-existent URLs returning 200 instead of 404
Many PHP and framework sites use a catch-all rewrite that serves the homepage for any unmatched path. The page looks fine to a human, but the server returns a success status.
Google calls these soft 404s. They create an unlimited supply of duplicate URLs, waste crawl budget, and drag down the site’s overall quality assessment.
How to check: visit yoursite.com/this-page-does-not-exist-12345. You should see a genuine error page, and the status code should be 404, not 200.
3. Duplicate titles and meta descriptions
If your services page and your contact page carry identical titles, Google has no signal for which answers which query. Both are weakened.
How to check: Google Search Console does not report this well. Crawl the site with Screaming Frog’s free tier and sort by title. Any repetition is a fault.
4. Redirect chains and mixed hostnames
A site should resolve to exactly one canonical hostname. When http, https, www and non-www all serve content independently, you have four versions of every page competing with each other.
Worse is the chain: http://www redirecting to http, which redirects to https://www, which redirects to https. Each hop loses a little and slows the page.
How to check: type all four variants into a browser. Each should land on the same final URL in one redirect.
5. Core Web Vitals ignored
Page experience is a genuine ranking factor, and in India it matters more than global averages suggest, because a large share of traffic arrives on mid-range Android phones over mobile data.
The usual culprits are the same everywhere: uncompressed hero images, render-blocking CSS, four font files where one would do, and layout that jumps as elements load.
How to check: PageSpeed Insights, but read the field data section rather than the lab score. That is what Google actually uses.
6. No structured data
Schema markup does not directly raise rankings, but it governs whether you appear as a rich result — star ratings, FAQ dropdowns, business hours in the sidebar. Those dramatically affect click-through rate on the same position.
For a local service business the minimum set is LocalBusiness or ProfessionalService, Service markup on each service page, and FAQPage where you genuinely answer questions.
How to check: Google’s Rich Results Test.
7. A broken or lying robots.txt
We have seen a single stray character block an entire site. A common one:
Disallow: / thank-you.html
The space after the slash makes the rule malformed. Strict parsers can interpret it as Disallow: / — blocking everything. Equally common is a robots.txt left over from a staging site, still disallowing the whole domain months after launch.
How to check: yoursite.com/robots.txt, then paste it into Search Console’s robots.txt tester.
8. Orphan pages and shallow internal linking
A page linked from nothing is a page Google struggles to find and has no reason to value. Internal links are how authority moves around your site, and most sites route all of it into the navigation and none of it into the pages that need to rank.
How to check: crawl the site and look for pages with zero inbound internal links. Then check whether your most commercially important page is linked from more than just the menu.
9. Cloaking left behind by a previous agency
This one is rarer but far more dangerous. A rewrite rule detects Googlebot by user agent and serves it different content from what humans see.
It is sometimes installed with good intentions and sometimes not. Either way it violates Google’s spam policies and risks a manual action that removes the site from search results entirely.
How to check: read your .htaccess file and look for any rule referencing Googlebot or bot user agents. Then use the URL Inspection tool in Search Console to view the page exactly as Google renders it, and compare with your browser.
10. No analytics, or analytics nobody reads
Not strictly a ranking factor, but it is why the other nine go unnoticed for years. Without Search Console and GA4 configured, a traffic collapse is invisible until enquiries dry up a quarter later.
How to check: view source and search for gtag or googletagmanager. If neither appears, nothing is being measured.
A 20-minute self-audit
- Check the canonical tag on three inner pages.
- Visit a deliberately invalid URL and confirm it returns a real 404.
- Test all four hostname variants for a single clean redirect.
- Read robots.txt line by line.
- Read .htaccess for any bot-specific rules.
- Run the homepage and one service page through PageSpeed Insights.
- Run both through the Rich Results Test.
- Confirm Search Console and GA4 are both installed and receiving data.
Most sites fail at least three of these. The good news is that technical faults, unlike content and authority, are usually fixable in days rather than months — and the recovery in impressions often shows up within two or three weeks of a recrawl.
We run this audit properly and send you the full findings, free, whether or not you go on to work with us. Request an audit.