Content decay detection in Search Console

Content decay is the slow slide of a page that used to perform: clicks drifting down month over month, not in one sharp drop but in a long, quiet decline. It is the traffic you lose without noticing, because no single week looks alarming. You can find it in your own Google Search Console data, one page at a time, before it costs you the ranking.

The catch is that a falling line can mean three different things, and Search Console shows you the same shape for all of them. A page can be truly decaying, or it can be holding its rank while an AI Overview absorbs the click, or it can be so small that the drop is just noise. This guide covers what decay is, why it happens, how to read it honestly in your own data, and how to tell real decay from the two things that look exactly like it.

TL;DR:

• Content decay is a steady decline in a page's clicks over time, not a one-off dip. Read it on page-level data, which is the honest source for volume and change.
• Before you call it decay, rule out two look-alikes: an AI Overview taking the click at a steady rank, and a tiny page whose numbers are just noise.
• Clicks falling at flat impressions is a different disease from impressions falling. The first is a snippet or SERP problem; the second is a ranking or demand problem. The fix differs.
• Search Console only keeps 16 months. To see decay older than that, you have to snapshot the data forward yourself, because Google offers no backfill.
• Search Console stops at the click. It can show a page is losing clicks. Whether the visitors you kept still convert is a different measurement on a different machine.

What is content decay?

Content decay is a gradual, sustained fall in a page's organic performance after it has matured, distinct from the ordinary week-to-week wobble every page has. The content marketing agency Animalz named the pattern with a lifecycle model: a page spikes on publish, settles into a trough, grows as it earns rankings, then enters a slow decline. Decay is that last phase, and it is the one that hides, because it looks like nothing is wrong on any given day.

The word that matters is steady. A single bad week, a seasonal dip, or a Google update wobble is not decay. Decay is a trend: three, four, six months of a line pointing down while the page itself sits untouched. That is why you cannot diagnose it from a snapshot. You need two comparable windows, or better, a rolling one, so a genuine slide separates itself from the noise around it.

Why does content decay happen?

A page decays because the world around it changes while the page stands still. Competitors publish something newer or deeper and take a share of the clicks. The search intent behind the query shifts, so the page that fit last year fits less well now. Freshness signals fade: for queries where recency matters, Google's query-deserves-freshness behavior quietly favors newer pages, and an old page loses ground it never did anything to lose. And increasingly, the SERP itself changes shape: an AI Overview appears above the results and answers the query before the user clicks anything.

Naming the cause matters, because the cause decides the fix. Decay from a stale page wants a refresh. Decay from a competitor wants more depth or a better angle. Decay from an AI Overview wants a citable answer, not a rewrite. Decay from a cooling topic may want nothing at all. If you treat every falling line as "refresh the page," you will spend effort where it can't help and miss the pages where it would.

How do you detect content decay in Search Console?

Compare a page's clicks across two comparable windows and look for a sustained drop, reading the numbers at the page level. Page-level data is the honest source for volume and change, because it counts every click, including the ones tied to queries Google hides. The difference between page-level and query-level data is the whole reason to start here: if you sum up the query view instead, you undercount by the anonymized share and can invent a decline that isn't real.

The method, one page at a time:

• Use comparable windows. Set the same length against the same length: last 28 days versus the 28 before, or this month versus last. Comparing a 28-day window to a 90-day one, or last week to a holiday week, manufactures false decay. A rolling window smooths out the weekly noise better than two fixed blocks.
• Read the trend, not the week. One down week is weather. A line that points down across several windows is climate. Decay is the climate read, so give it enough time to separate from the wobble.
• Rank the losers by absolute clicks lost, not percent. A page that fell from 40 clicks a month to 25 lost more real traffic than one that fell from 4 to 1, even though the second looks worse in percentage terms. Sort by clicks lost so the pages that actually cost you rise to the top. The four Search Console metrics explained covers what each number honestly measures if you want the ground under this first.

Is it decay, or an AI Overview taking the click?

Here is the look-alike that fools most people: a page whose clicks fall while its impressions and its position hold steady is usually not decaying. It is still ranking, still being shown, still where it was. Something is intercepting the click before it reaches you, and in 2026 that something is most often an AI Overview answering the query on the results page. Ahrefs measured the top organic result losing 58% of its clicks when an AI Overview is present. Your rank did not move. The SERP moved.

This distinction changes the fix entirely, which is why it is worth reading carefully:

• Clicks down, impressions and position steady points at the SERP, not the page. The honest move is to make the page the thing the AI Overview cites: a crisp, quotable answer, a stat, a clearly sourced fact. Rewriting the whole page for "freshness" does nothing here, because the page never lost its rank.
• Clicks down, impressions and position also down is the real decay signal. You are being shown less and ranking lower, so the page is genuinely losing ground, and a refresh, a consolidation, or a better angle is on the table.

Reading the two apart is the difference between fixing a problem and rewriting a page that was never broken. It is the same honesty that decides whether a low CTR at a good position is a title you can fix or an AI Overview you cannot, and that separates a real quick win from a phantom one on the intent-mismatch side of the diagnostics.

Clicks down or impressions down: two different diseases

Once you have ruled out the AI Overview, the direction of the drop still tells you which disease you have. A fall in clicks and a fall in impressions are not the same problem, and the daily-store view in your own data lets you see which one is moving.

• Impressions holding, clicks falling is a click-rate problem. You are still being shown for the same searches, but fewer people are choosing your result. That points at the title and description, or at a SERP feature (an AI Overview, a snippet, a competitor's richer result) eating attention above you.
• Impressions falling is a ranking or demand problem. Either you slipped down the results for queries you used to win, or the queries themselves cooled off and the demand simply shrank. The freshness mechanism runs both ways: a topic that spiked and faded takes your impressions with it, and no rewrite brings back demand that left.

Same falling clicks line, two different root causes, two different fixes. Read impressions alongside clicks and the page tells you which one you are looking at.

Phantom decay: mind the click floor

The smallest pages lie to you. A page that went from three clicks last month to one this month is down 67%, which sounds like a crisis and is almost certainly nothing. At that volume a single curious visitor is the entire signal, so the percentage swings wildly on pure chance. If you rank decay by percentage without a floor, these tiny pages flood the top of your list and bury the pages that actually matter.

The fix is a floor: ignore decay on any page that had fewer than a handful of clicks in the prior window, because below that line the number is noise, not a trend. This is exactly why sorting by absolute clicks lost beats sorting by percent. Real decay is a page that earned meaningful traffic and is measurably losing it, not a page that was always in the low single digits and jittered by one.

You can only see the decay you kept

Search Console holds your data for 16 months, then deletes the oldest day every day as the window rolls forward. There is no archive and no backfill: once a day ages past the window, Google does not give it back. That is a hard limit on how far back you can look for decay.

The practical consequence is that a slow, multi-year slide is invisible from inside the tool. If a page has been quietly declining for two years, Search Console can only show you the most recent sixteen months of it, and by then the early, cheap-to-fix part of the decline is already gone. The only defense is to snapshot the data forward, keeping your own daily history so the window never closes on you. If you want the full picture of the wall and the freshness lag that sits alongside it, the 16-month retention breakdown walks through both.

What should you do when a page is decaying?

Match the move to the cause you diagnosed, not to the fact that a line went down. You have four honest options: refresh the page, consolidate it with a competing one of your own, re-target it, or accept the decline. Which one fits depends on why the page is losing, which is the whole reason the diagnosis came first. How to revive a decaying page walks the full decision, which move to pick from your own data, and how to run each.

• Refresh when the page has genuinely fallen behind: facts are stale, the intent shifted, or a competitor went deeper. Update it against what the query wants now, not just the year in the title.
• Consolidate when you are competing with yourself. If two of your own pages drifted onto the same query, they split the signal and both decay. Merge them into the stronger URL and redirect the weaker one.
• Re-target or accept. Sometimes the demand simply left and no fix brings it back, and sometimes a small, slow decline on a minor page is not worth an afternoon. Spend the effort on the pages losing real, absolute traffic.

Then the limit that keeps this honest: Search Console stops at the click. It can show that a page is losing clicks, and it can show you which of the three shapes the loss is. It cannot show you whether the visitors you still get convert, or whether the traffic you lost was ever worth anything. That happens after the click, in your analytics or your product, on a different machine. Treat a decaying page as a lead to act on, not a proven loss to mourn.

Detecting decay without watching every page by hand

Done by hand, this is relentless work: pull every page's clicks across matched windows, separate the trend from the noise, check impressions and position to rule out an AI Overview, apply a click floor so the tiny pages don't shout, and remember the whole time that the 16-month wall is quietly erasing your oldest history. It is exactly the kind of weekly discipline that slips first when you are busy.

That is the work QueryScope is built to hold. It reads your real Search Console data in the terminal, compares each page against its own recent baseline from a daily history it keeps for you, and separates a decaying page from one that is merely losing the click to a SERP feature, so the pages actually sliding surface instead of hiding inside a healthy-looking total. If you want the one-line definition behind any term first, the Search Console glossary covers clicks, impressions, and the 16-month wall, each with the caveat that comes with it.

Sources

• Google Search Central, A deep dive into Search Console performance data filtering and limits (rare queries are dropped from the query table while their clicks stay counted in the page and site totals, so page-level data is the honest source for volume and change).
• Google Search Console Help, Performance report data retention (Search Console retains performance data for 16 months, with no historical backfill).
• Ahrefs, AI Overviews reduce clicks by 58% (December 2025 update: the top organic result loses 58% of its clicks when an AI Overview is present, so a page can hold its rank and still lose the click).
• Search Engine Land, Query Deserves Freshness (QDF): what it is and how it works (Google weights recency more heavily for queries with spiking or time-sensitive demand, so freshness can lift a newer page and let an older one slip).
• Animalz, Why organic search traffic declines, and what to do about it (the content lifecycle and the decline phase that names the decay pattern).