There's a specific phone call we get about twice a month. Leads have quietly dried up, nobody can explain it, nothing broke, nobody changed anything. They pull the site up on their phone while we're talking and it looks exactly like it did the day it launched.
That's the part worth noticing. In almost every one of these cases the site really is fine - it just isn't what's ranking.
Nobody tells you this when you buy a website: there are two versions of it. The one you published, and the one Google built. Google crawled your pages, kept some, picked its own favourites among the duplicates, ignored a few of your instructions, never found a couple at all, and stored the result. That second version is what competes in search results.
Most audits check the first version. Every failure below lives in the gap between the two.
1Google is overruling your canonical tags
Canonical tags tell Google which version of a page is the real one. Most people implement them once, see them in the page source, and consider it done.
But a canonical tag isn't an instruction - Google treats it as a hint, one signal among several. Your internal links are another. So are your sitemap and your redirects. When those disagree with your canonical tag, Google resolves the conflict itself, and it doesn't ask.
The page you wanted to rank then simply isn't in the index. A different one is. You'll keep seeing your preferred URL in your CMS, your sitemap, your page source - everywhere except the place that matters.
In Search Console, open Indexing → Pages and look for Duplicate, Google chose different canonical than user. Then run any affected URL through the URL Inspection tool, which shows "User-declared canonical" and "Google-selected canonical" side by side.
When those two lines disagree, you've found a page that can't rank no matter how good it is.
2You have pages nothing links to
An orphan page is a page on your site with no internal links pointing at it. Nothing in the nav, nothing in a footer, nothing in a related-posts block. You can reach it if you type the URL. Nobody else can find it by browsing.
They accumulate quietly. A page gets pulled from the nav during a redesign. A service leaves the dropdown but the page stays live. A location page gets built for a campaign and never linked. Six months later a dozen pages exist, load perfectly, and are invisible to your own site structure.
If there are no links, we won't find the URL, robotted or not.John Mueller, Google
Here's where most advice goes wrong: people assume the XML sitemap solves this. It doesn't. Google's own team has described internal links as the primary way URLs get discovered, with sitemaps second. A sitemap tells Google a URL exists; internal links tell Google it matters. A page sitting in your sitemap with zero internal links usually gets found and then parked - in Search Console, Crawled – currently not indexed. Google's way of saying: we looked, we weren't convinced.
Crawl your site with Screaming Frog, then compare the crawl against your sitemap and your Search Console URL list. Anything that appears in the latter two but not in the crawl is orphaned.
3Your most important content only exists after JavaScript runs
This one is almost impossible to spot by eye, because in a browser everything works.
Google crawls in two passes. The first fetches the raw HTML your server sends. The second - rendering - executes JavaScript later, in a separate queue, when resources allow. Anything that only exists after scripts run is invisible to that first pass, and the second isn't guaranteed to be timely or thorough.
For local businesses it hits in a predictable place: the details that make a location page local. Addresses pulled from an API. Phone numbers injected by call tracking. Service lists loaded at runtime. Reviews in a third-party widget. On screen the page is rich and specific. In the raw HTML it's a heading and some boilerplate.
Links are the worse version of this. If your navigation or your "areas we serve" list is built by JavaScript, Google may not follow those links at all - and every page beneath them inherits problem #2 through no fault of its own.
Right-click any important page and choose View Page Source - that's the raw HTML, not Inspect Element, which shows the rendered version. Search it for your phone number, your address, a service you offer. If they're not there, Google's first pass can't see them either.
4Your location pages are the same page wearing different hats
If you serve multiple cities, someone has probably told you to build a page for each one. That advice isn't wrong. The execution usually is.
The failure pattern is a template where the only thing that changes is the city name. Looking for a plumber in Katy? Our expert Katy plumbers… Twenty of those, one per suburb. It feels productive.
Google has gotten very good at spotting find-and-replace at scale, and its spam policies name the behaviour directly. The usual outcome isn't a penalty - it's quieter. A few pages get indexed, most sit at Crawled – currently not indexed, and the ones that rank compete with each other for the same queries, so none ever builds enough authority to win.
In Search Console, open the Performance report and filter by page. If you have twelve city pages and impressions are scattered thinly across all of them with no clear winner, they're cannibalising each other. Cross-check against Indexing → Pages to see how many were never indexed at all.
The fix is the boring one. A location page earns its place when it contains something that could only have been written about that place: a job you did there, a permit quirk, a response time from your actual depot, a photo of your van somewhere recognisable. If you can swap the city name and the page still reads fine, it isn't a location page - it's a template with a variable in it. Six real ones beat twenty generated ones, and it's perfectly acceptable to cover several areas on a single strong page.
5The leads are arriving. You're just not receiving them.
The first four cost you rankings. This one costs you the leads you already earned, and it's the one we find most often.
Your form works. The visitor fills it in, hits submit, sees the thank-you message, and gets on with their day expecting to hear from you. The submission then has to travel from your site to your inbox, and that leg is where it dies.
Most forms send notification email straight from the web server, and modern inboxes are deeply suspicious of mail like that. Without email authentication on the sending domain, notifications land in spam, in the Promotions tab, or nowhere. Nothing errors. The success message fires either way - it's triggered by the form, not by the delivery.
The variants are just as mundane and just as expensive. The notification address is an inbox nobody has opened since a staff change. Someone tightened the spam filter and it's binning real enquiries with the bots. You switched email providers last year and the form was never reconfigured.
What makes it costly is the symmetry of the misunderstanding. The customer thinks you ignored them. You think nobody enquired. Both are wrong, and neither has any way to find out.
Submit your own form right now, from your phone, using a personal email address rather than the office one. Check the destination inbox and its spam folder. If your form plugin stores submissions, compare that log against the emails you actually received - the gap between those two numbers is your answer, in leads.
Then set a monthly reminder to repeat it. Forms don't only break when you change your site; they break when your email provider changes something and nobody tells you.
The ten-minute version
If you do nothing else from this article, do these five things this week:
Search Console → Indexing → Pages. Read the reasons pages aren't indexed. Anything in Crawled – currently not indexed or Duplicate, Google chose different canonical is a page you're paying to maintain that cannot generate a lead.
URL Inspect your three most valuable pages. Confirm Google's selected canonical matches the URL you expect.
View Page Source on your main location or service page and search for your phone number. If it's not there, it's not in the first thing Google reads.
Filter Performance by page and check whether your city pages are splitting impressions between them.
Submit your own contact form and confirm the email lands somewhere a human will see it.
None of that needs a paid tool, and none of it means touching your website.
That's the pattern worth keeping. Every problem here was invisible from the front end and visible in Google's own free reporting. These things survive for months not because they're hard to fix - most are half a day of work - but because nobody thinks to go and look at the other version of the site.
Not sure which of these is affecting you?
We'll run the check and tell you what we find - free, no obligation, and the findings are yours either way.
W
Waystar Studio
We build websites and run local SEO for small businesses across the US - one team handling design, build, search and ads, month to month with no contracts.
More about us →