Skip to content
SEOCheckGet $100 free

Diagnosis

You did the work and nothing moved. Now what?

Flat traffic after months of effort is the most common complaint we see. Usually one of seven things is behind it.

SEOCheck research team ·

Person looking up at large clock gears turning without moving the hands

Start with the boring checks

Most flat results are not mysterious. Pages are not indexed, two pages compete for the same query, or the target query is simply owned by much bigger sites. Fix those before blaming Google.

The order you check things in matters. Start cheap. Cheap checks come first, and expensive theories come last. Index status in Search Console takes a minute. Reading a competitor link profile takes five calls. Rewriting the whole site takes a month. A domain backlink checker belongs in the middle of that order, not at the start.

A page can be indexed and still never rank. That happens. It can sit in the index for a year and never appear for anything. Treat that case as a visibility question first.

The Search Console screens you already use separate the two problems. Performance shows clicks and impressions by query. Pages shows what is indexed and what is excluded. Together they tell a visibility problem from a click problem, and the remedies differ. Know which one you have.

If impressions are flat and the page sits beyond position 30, you have a ranking problem, usually relevance plus links. If impressions are fine but clicks are low, the title and snippet are the issue. Different fix. For the first case, compare against the leaders with the five-call teardown. For the second, rewrite the title before touching links.

The diagnostic order

  • Is the page indexed? Check Search Console, not a site: search.
  • Which query is it supposed to win, in one phrase?
  • Do two of your own pages target that phrase? That is cannibalisation.
  • Does the ranking page match the intent: guide, product list or tool?
  • Run a domain backlink checker on you and the top three results.
  • Did anything change on the page, such as keyword density or headings?
  • Is the timeline realistic for the site's age?

Worked example, clearly hypothetical

Invented numbers to show the reasoning, not a case study.
SiteLinking domainsDRVerdict
You4023Weak but not hopeless
Rival A90041Strong, likely beyond this quarter
Rival B6026Within reach
Rival C5525Within reach
Gardener examining one healthy plant and two wilting plants on a bench

What the example tells you

Two of three rivals sit within reach, so links are not the whole story for those results. You would study their pages next. For rival A, you would pick another query.

Now the awkward case: you bought 40 links and nothing happened. Check the linking domains first. Did they land at all? The index refreshes about monthly, so a link placed three weeks ago may not show yet. Then check the pages. Links on thin pages that nobody visits carry little. Then check relevance.

Sometimes the honest answer is that 40 links is a small number against rivals with hundreds. Some links also get ignored. Google SpamBrain began neutralising link spam in the December 2022 link spam update, so paid links on obvious networks may simply count for nothing.

Later, when the audit says links are weak, read what separates a good link from a filler one before spending anything. Cheap placements on thin pages are the typical way people lose a quarter.

Content changes that can move a page

Many flat results follow a change on the site itself. Look there first. The change may have been small and sensible. A new template, a rewritten intro or a batch of new posts can shift how a page reads to a crawler and to a reader.

Before you blame an update, list what you changed in the last few months. The list is often shorter than you think, and it is easier to read than your memory of the summer.

Automated content, and what to check

Did automated content help or hurt? The honest answer depends on the pages. Pages that answer a real question with specific detail tend to hold up. Pages that repeat the same generic paragraph across many URLs tend to fade, and the fade often starts in the second or third month.

Pull the pages that lost impressions. Read them one after another. If they all sound alike, the problem is the template, not the topic.

AI-assisted pages after a core update

Some site owners now ask how to judge AI-assisted content after a broad update. Read the page as a stranger would. Does it answer the query in its first screen? Does it add a number, a step or a case that a competitor does not have?

A page that passes that test does not need to be rewritten from scratch. A page that fails it needs a better answer, not more words.

Rewriting a title without moving the page

If impressions are fine but clicks are low, the snippet is the problem. Rewrite the title to match the question people type, and leave the body alone for a while. Changing everything at once makes the next reading impossible to interpret.

Internal links after a restructure

Moving a page or changing its folder can cut the internal links that used to reach it. Check whether the old links point to a redirect, and whether the new location still receives links from the rest of the site. A restructure that looks tidy can quietly remove the paths that carried the page.

Thin pages after a batch of new posts

Publishing a batch of short posts in one week can dilute the site. Each post may be accurate and still add little for the query it targets. If several new pages sit near the same phrase, they compete with each other, and the older page often loses ground.

Consolidate where two posts answer the same question. Merge the best material into one page, then redirect the weaker URL to it.

Intent, format and the patience a new site needs

Two pages can target the same phrase and only one of them ranks. Matching intent wins. The winner usually matches the intent behind the query better. Guides, lists and tools answer different needs, and Google tends to show the format people already expect.

New sites carry an extra burden. Their pages have no history, so every signal arrives slowly. Patience is part of the diagnosis, not an excuse for it.

Guide, list or tool?

Open the current top results for the query and note their format. If the leaders are tools, a guide will struggle however good it is. If they are long guides, a short post will struggle in the same way. Match the format first, then compete on detail.

Low-volume terms that flop

Keyword tools often show a term with small volume and call it an easy target. In practice a low-volume term can attract very few searches, and the few people who search may want something different from what you wrote. Check the live results before you build a page around a number.

Does the domain extension matter?

People often wonder whether a taken .com domain should push them towards .so or .io. In most cases the extension does little to a ranking. The page, its links and its fit to the query do most of the work. Pick the name you can remember and move on to the page.

Why a page slips after it ranks

A page that climbs and then falls back within weeks is often just being tested. Search results shift as new pages appear and as users react. Before you act, check whether the page held its position for more than a few days, and whether the query itself changed.

If the drop continues for a full cycle, go back to the checklist above. The cause is usually one of the basic checks rather than a mystery.

Updates worth knowing when dates line up

  1. Penguin 4.0

    Link evaluation became real-time and more granular.

  2. Link spam update

    SpamBrain rolled out against link spam.

  3. Core and spam updates

    A large core update plus new spam policies, announced in March.

  4. Core update

    Another broad core update. Volatility is normal around such dates.

Craftsman clamping a broken ceramic vase on a wooden workbench

Recover after a drop, step by step

The sequence matters because each step gives you a fact to decide the next one with. Skipping ahead is how a small problem becomes a rewrite.

01

Find the date of the change

Line up the drop with your own edits and with the update timeline above. A drop that starts the same week as a deploy needs a different check from one that starts in a quiet month.

02

Separate impressions from clicks

If impressions fell, look at ranking. If impressions held but clicks fell, look at the title and snippet first.

03

Check the pages, then the links

Confirm the page is indexed and answers the query. Only then run the five-call teardown on the leaders.

04

Change one thing

Make a single, written change and note the date. Two changes at once make the result impossible to attribute.

05

Wait a full cycle before judging

For a new site, that is months. For an established one, it is usually several weeks. Read the trend, not the day.

Quick answers

How long should I wait?

For a new site, think months, not days. A page that jumps within days and falls back is usually just testing.

Why do irrelevant sites rank?

Often they hold strong links or match a query's intent better than the relevant ones. Check their profile before assuming a glitch.

Do I check once or keep checking?

Monthly is enough. Our index refreshes at that pace anyway.