Content Decay: How to Find Pages Slowly Losing Traffic in Search Console

Short answer: Content decay is the slow loss of search traffic from a page that used to perform, usually because competitors improved, facts aged or search intent changed. Find it in Search Console by comparing the last three months with the previous three months and with last year, then sorting pages by clicks lost. Then refresh, merge or leave each page alone, depending on what it is worth to your business.

One of our blog posts, a guide to what SEO is, went from 4,341 impressions to 3 between two 28-day periods. No alert fired, nothing broke, no error appeared. The site carried on, the blog carried on, and a page Google had been showing thousands of times simply stopped being shown. We saw it only because we went looking at page-level data, and a later study of our own Search Console data showed that one post was about 79% of the blog's whole impression loss.

That is what makes page-level decline dangerous. A building that collapses gets everyone's attention. A leak behind the wall gets none, until the damage is expensive. Most pages lose their place in a way nobody is watching for, and most reports are built to show totals, which is exactly the view that hides it.

Decay or drop? They need different responses

Our "what is SEO" case was so abrupt that it is probably not classic decay at all. We have three similar posts on that topic, so Google may have chosen a different one, though we have not confirmed that. The distinction matters, because it tells you what kind of problem you have.

If you are looking at a cliff, use our 20-minute traffic drop check first. The rest of this article is about the slope.

Why good pages fade

• Others caught up. A guide that was the best in 2023 may now be sixth, without having got any worse.
• The facts aged. Screenshots of old interfaces, dead tools, last year's statistics. Anything written about Universal Analytics, for example, stopped being useful when GA4 replaced it.
• Intent moved. Searchers who once wanted a long guide may now want a tool or a short answer.
• The answer moved. Definitions in particular can now be answered on the results page. Google counts an AI Overview as a single position, and your link may or may not be in it.
• Your own pages compete. New posts on the same topic can pull attention from older ones.
• Demand fell. Sometimes fewer people search for it. Nothing about your page can fix this.

Building your decay list in 20 minutes

1. Open the Performance report. Click the date filter, choose Compare, and set the last 3 months against the previous period.
2. Open the Pages tab and turn on the difference columns, then sort by clicks difference, largest loss first.
3. Copy down pages that lost a meaningful amount. On a small site we use at least 10 clicks, because single-click changes are noise.
4. Now change the comparison to the same three months last year. This removes seasonal patterns. An ice cream brand falling in October is not decaying.
5. Keep the pages that fell in both views.
6. For each page, open the Queries tab and find which searches lost the most.
7. Run URL Inspection on the worst few to confirm they are indexed and that Google agrees with your canonical URL.

Search Console keeps 16 months of data, so look now and keep an export. Our own "what is SEO" loss could only be understood because we had earlier numbers to compare with.

Then cross-check with GA4. Open Pages and screens, or the landing page report, and see whether the falling page still brings in real key events. A page losing 200 visits and no enquiries is a very different priority from one losing 20 visits and half your leads.

The four decisions

Every page on the list gets one of four verdicts. Decide using two questions: does this page matter to the business, and can it still win?

"Leave it" is a legitimate result. Our post on Reddit alternatives collects a lot of impressions and would never produce a client, so saving it would be a poor use of an afternoon. Do not spend your refresh budget protecting traffic that was never going to pay.

What a real refresh looks like

A refresh is not changing the date and adding two sentences. Google's helpful content guidance specifically asks whether you are changing dates to make pages seem fresh when the content has not substantially changed. Readers notice as well. Do this instead.

1. Read the current top three results. Find out what the searcher now expects.
2. Read your own page as a stranger. Mark anything out of date, vague or padded.
3. Answer in the opening lines. Older posts often warm up for several paragraphs.
4. Replace old facts with sourced ones. Link to the original source, and add the date where it matters.
5. Add what only you know. Your data, your cases, your mistakes. This is what a rival cannot copy.
6. Keep the URL. Changing it throws away what the page has earned. If a change is truly needed, use a 301 redirect.
7. Link to it from newer pages. Older posts tend to drop out of your internal links as the blog grows.
8. Record the date. Add a chart annotation so you can judge the result in six weeks.

Catching it earlier next time

The trouble with our 4,341 impressions was that nothing was watching that page for us. Three habits would have caught it.

• A weekly look at the biggest losers. It is one step in our 15-minute weekly report.
• A quarterly decay review. Run the steps above once every three months.
• An alert for the pages that matter. GA4 can run custom insights, up to 50 per property, with email notices. Its anomaly detection compares each value with a prediction, training on 90 days of history for daily checks, so an unusual fall can reach your inbox. Set one for your top landing pages.

The page that fell without a sound

We are still working out exactly what happened to our post, and we have a job to do: choose one of our three "what is SEO" articles, merge the others into it, and watch the result. It would be neater to report the ending. The honest version is that no tool warned us, and the reason we found it at all is that somebody went looking at page level.

ONE Dashboard shows your pages and queries with period comparisons in plain English, using read-only access, which makes quiet losses easier to spot. If you prefer to ask, connect an AI assistant and try: "Which pages lost the most clicks in the last 28 days compared with the previous 28?" Then check one figure in the real report.

Your pages will not tell you when they start to fade: connect Search Console to ONE Dashboard free and see which ones already have.