Skip to content
BOOSTD
Free SEO audit

SEO

Eight technical SEO issues that lose traffic without an alarm

Published by Zubair Afzal, responsible editor

First published Updated 5 min read

A green dashboard means the server answered

Monitoring tells you when a site is down. It can’t tell you that the site is up, loading normally, and silently failing at its job.

These faults share a shape. Nothing errors, no alert fires, and the page looks right. The only symptom is a number that drifts unnoticed for a quarter, by which time the cause is buried under three other changes.

Eight of them do most of the damage. All eight can be checked in about twenty minutes a month, without a crawler or an audit.

The eight

What breaks without a sound

Roughly in order of how much they cost when they happen.

  • A form that stopped delivering. A plugin update, an expired mail service, a spam filter change. The form submits, shows a thank-you page and sends nothing. Every enquiry is lost until somebody complains.
  • A staging directive shipped to production. The instruction that correctly keeps a test site out of search travels with the code at launch, and pages start disappearing a few weeks later.
  • A canonical pointing at the wrong page, often the staging domain or the homepage. The page is fine and is telling search engines to credit something else.
  • Redirect chains left after a migration. Each move adds a hop. Three hops still work for a person but are slower and less reliable for everything else.
  • Parameter and filter URLs multiplying. Sorting, tracking and filtering produce thousands of near-identical addresses, and the important pages compete with their own variants.
  • Blocked resources. A tightened rule stops stylesheets or scripts being fetched, so the page a crawler assembles isn’t the page a visitor sees.
  • Soft 404s from empty templates. An empty category returns a normal page saying nothing, which is worse than a proper error because it stays in the index.
  • A key template gone slow. One change to a shared component, and the pages that earn the money take twice as long to become usable on phones and ordinary connections.

The routine

Twenty minutes, once a month

Do it on the ten pages that matter, not on everything.

  1. Submit your own enquiry and phone your own number

    Fill in the main form from a phone, on a browser you aren’t logged into, using a name you’ll recognise. Check it arrives in the right inbox, and check the spam folder. Then ring the number on the site and confirm it reaches the right place.

    You get: A confirmed enquiry and a confirmed phone call, dated

  2. Read the page indexing report and view one page source

    Open the page indexing report in Search Console and look for anything newly excluded, especially anything blocked or with a different canonical. Then view the source of your most important page and search it for the robots directive and the canonical: two lines of code on one page.

    You get: A note of anything newly excluded, with the date it changed

  3. Compare the dates before you investigate anything

    If a number moved, put the date it moved next to your list of deployments, plugin updates and content changes. Most of these faults are shipped, not developed, so the date match usually names the cause. Keep a simple log of changes for this reason.

    You get: A cause, or a confirmed absence of one

Why they survive

Nobody owns a website between redesigns

The reason these faults last for months isn’t technical. A website usually has an owner during a build and nobody afterwards. The agency that built it has moved on, the marketing supplier works on content, and the person who could notice doesn’t have the site open.

So the work isn’t difficult, it’s unassigned. Give it twenty minutes a month on a named person’s calendar, with the results written in the same place every time.

The log matters as much as the check. Its value comes six months later, when a number has moved and somebody can see exactly what was true in April.

Questions

Common questions about silent faults

Does uptime monitoring not cover this?

No. Uptime monitoring asks whether the server responded. Every fault here happens on a server responding perfectly, with a page that loads and looks correct.

That’s why they survive for months. Nothing in the normal running of the business produces a signal.

How would I know a form has stopped sending?

Submit one yourself, monthly, from a phone on a browser you aren’t logged into. Use a distinctive name and check the inbox it should arrive in, including the spam folder.

It’s a two-minute test that catches the most expensive item on this list, and the one people skip because the form obviously works.

Do I need a crawler to do these checks?

For the monthly version, no. Search Console, a browser and your own eyes cover it, on the handful of pages that matter to the business.

A crawler becomes worth it above a few hundred pages, or to check every page, not just the important ones. That’s a quarterly job.

What is the single most common one?

A staging directive reaching production. Staging sites are rightly blocked from search, and the block travels with the code if nobody removes it at launch.

It’s one of the fastest to fix and one of the slowest to recover from, because pages have to be crawled again before anything returns.

How fast should I expect recovery after fixing one?

It depends on how quickly the affected pages are recrawled, which depends on the size and prominence of the site. Small sites are often recrawled within days, larger ones unevenly over weeks.

Request indexing for the most important pages, not a thousand of them. Nothing makes recovery faster than a crawl.

References

Sources for technical SEO issues that lose traffic

The official documents and research this page is based on. Platform rules change, so check the source before you act on a detail.

  1. Google Search Central: block search indexing with noindex
  2. Google Search Central: consolidate duplicate URLs with canonicals
  3. Google Search Central: redirects and Google Search
  4. Google Search Central: Search Console page indexing report

Related

Keep reading on technical SEO issues that lose traffic

Start here

Start with the form test this week

If you do one thing from this page, submit your own enquiry form and check it arrives. It takes two minutes, and a broken form costs money from the day it breaks.