Content Decay: Why Your Best Pages Stop Ranking

A page that used to book two jobs a week now books zero. Your traffic graph looks fine. Nobody on the team notices for five months, and by then the page is on the second page of Google and the competitor who overtook it has a year of head start.

That is content decay, and on a 40-to-100-page service site it is almost always invisible until it is expensive. Here is how to see it early, work out what actually caused it, and change only the things that move it.

What is content decay and why doesn't it show up in your traffic total?

Content decay is a measurable, sustained drop in a single URL's rankings and impressions over months — typically 20% to 60% of a page's search visibility over two or three quarters — while sitewide traffic stays flat because other pages absorb the loss.

Total sessions is a sum. Sums hide substitution. Your /emergency-drain-cleaning page can lose 300 impressions a month while a blog post about water heater noises picks up 320, and your dashboard shows a green line. One of those pages sells a $600 job. The other sells nothing.

This is why decay is a URL-level problem, not a traffic-total problem. The unit of analysis is the page, and the pages that matter are the ones tied to revenue: your service pages, your city pages, and the three or four blog posts that actually feed your free audit form or your booking page.

The tell is not a cliff. Cliffs are usually technical — a noindex tag, a broken redirect, a botched migration. Decay is a slope: minus two positions this quarter, minus three the next, and the clicks fall off a cliff only when you cross out of the top five.

How do you detect decay on a site with low traffic?

Compare average position and impressions per URL across two non-overlapping 90-day windows in Google Search Console — not sessions, and not week-over-week. On a site doing a few thousand impressions a month, session counts are too noisy to read; position and impressions are the only signals with enough sample to trust.

The 15-minute version, which you can run yourself once a quarter:

• Open Search Console → Performance → Search results. Set the date range to the last 90 days, tick Average position and Impressions, and switch on Compare against the previous 90 days.
• Click the Pages tab. Sort by impression change, ascending. The worst offenders float to the top.
• Flag any URL that lost impressions and moved down in average position. Losing impressions while position holds is usually a demand drop (seasonality), not decay.
• Ignore anything under ~100 impressions in the prior window. Below that, the average-position number is noise.
• For every flagged URL, click into it and switch to the Queries tab with the same comparison. You want the query that used to drive it and what happened to that specific query.

Ninety days is not arbitrary. Google's own core-update guidance says that after a change, 'some changes can take effect in a few days, but it could take several months for our systems to learn and confirm that the site as a whole is now producing helpful, reliable, people-first content' (Google Search Central). A 30-day window will show you weather. A 90-day window shows you climate.

If you have never set the tool up properly, start with our walkthrough on Google Search Console for service businesses — the decay report is worthless if your property is verified at the wrong host or your date ranges cross a migration.

Why did the page stop ranking — stale, out-competed, or eaten by an AI Overview?

There are four causes worth separating, and they have different fixes: stale content (your facts and prices aged out), competitive displacement (someone published something better), intent drift (the SERP now wants a different kind of page), and SERP-feature loss (you still rank, but something above you takes the click).

Attribute the cause before you touch the page. Rewriting a page that lost clicks to an AI Overview is wasted money — the ranking was never the problem.

That last row is the one nobody publishes, and it is the one that has grown fastest. It deserves its own section.

How do you tell a SERP-feature loss from a real ranking drop?

The signature is unmistakable once you know it: average position is flat or improving, impressions are flat, and clicks are down. A ranking loss drags position and impressions down with the clicks. A SERP-feature loss does not touch them at all — you are still on the page, people just stop clicking you.

The most common culprit now is an AI Overview sitting above your result. Ahrefs compared 300,000 keywords using Google Search Console data and found that the presence of an AI Overview correlates with a 58% lower clickthrough rate for the #1 organic result (Ahrefs, December 2025 data). The penalty is worst at the top: position 1 loses 58% of its clicks, position 3 loses 46%, and position 10 loses 19%.

Independent behavioral data backs it up. Pew Research Center tracked the actual browsing of 900 US adults across 68,879 Google searches and found that when an AI summary appeared, users clicked a search result on just 8% of visits — versus 15% when no AI summary was present (Pew Research Center, March 2025 data).

Two things before you panic. First, this hits informational pages far harder than money pages: AI Overviews trigger on 21.4% of informational keywords but only 4.3% of commercial-intent and 2.1% of transactional keywords, and just 7.9% of local searches (Ahrefs, 146M SERPs, September 2025). Your 'emergency plumber in Tulsa' page is mostly out of the blast radius. Your 'why is my water heater leaking' post is not.

Second, the trend flipped. Seer Interactive found that organic CTR on AI Overview queries rebounded from a floor of 1.3% in December 2025 to 2.4% in February 2026 — an 85% increase (Seer Interactive, April 2026). Anyone still selling you a rewrite because 'CTR is in freefall' is working from last year's data. We cover what actually changed in how AI Overviews rewrote the SEO playbook.

Diagnostic shortcut: open an incognito window, search the query the page used to win, and look. If an AI Overview is there and you are not cited in it, that is your answer, and no amount of rewriting the intro paragraph fixes it.

What should you actually change in a refresh?

Five things carry almost all of the lift, and you can do all five in about 90 minutes per page: the answer paragraph, the dated facts, the price ranges, the internal links, and the query coverage.

• The answer paragraph. Rewrite the first 40-60 words under the H1 so they answer the query completely, on their own, with a number in them. This is the passage a snippet or an AI summary lifts.
• Dated facts. Any figure older than about 18 months is a liability. Re-open the original source, confirm the number is still what you said it was, and update it or cut it.
• Price ranges. If the page quotes '$150-$300 for a drain clean' and your market now runs $220-$450, the page is actively lying to buyers and they bounce. Nothing kills a service page faster.
• Internal links. Add two or three links from your newer, stronger pages into the decaying one, on anchors that describe the target. New links from pages Google already crawls often is the cheapest re-crawl trigger you have.
• Query coverage. In the URL's Queries tab, find the questions it now gets impressions for but never answers on the page. Add an H2 for each real one. This is where most of the recovered impressions come from.

If the page in question is a money page, its title and description do a second job — they set the click rate at whatever position you hold. Our guide to title tags, meta descriptions, and organic CTR covers that half of the problem.

One rule while you work: never invent a fact to fill a hole. If you cannot source a number, cut the sentence. We would rather ship a shorter honest page than a padded one, and so should you.

What refresh tactics change nothing at all?

Changing the publish date without changing the content does nothing, and Google's documentation is explicit that the date must describe the page: 'The dates must describe the publication or update date of the page, not the stories or events described therein' (Google Search Central).

The date is a reporting field, not a ranking lever. Bumping datePublished on an untouched page is the SEO equivalent of repainting a car with a dead engine — and if your visible date and your structured data disagree, you have created a real problem where there was none. Google's guidance is to 'ensure that the date (and optional time and timezone) match between the equivalent user-visible and structured values.'

• Bumping the date stamp on unchanged content — zero effect, and a trust risk with readers who notice.
• Swapping the year in the title ('2025' → '2026') while the body still references old facts — a promise the page does not keep.
• Adding 500 words of preamble to hit a word count — dilutes the answer, does not add coverage.
• Regenerating the whole page with AI and shipping it unreviewed — Ahrefs found the correlation between a page's share of AI-generated content and its ranking position is 0.011, effectively zero, across 600,000 pages, so this neither helps nor hurts on its own. It just wastes the one chance you had to actually fix the page.
• Firing the page into a link-building blast — links do not fix intent drift or a lost SERP click.

Google's own core-update advice says the quiet part out loud — avoid doing quick-fix changes, like removing some page element because you heard it was bad for SEO (Google Search Central). Diagnose, then act.

Should you refresh the page or rewrite it?

Refresh when the page still targets the right query and still holds a position between roughly 4 and 20; rewrite when the SERP has changed shape, the page has fallen past position 30, or more than half the body is wrong. Anything past position 50 with no links pointing at it is usually a consolidation candidate, not a refresh candidate.

Verdict: refresh in place, at the same URL, in the overwhelming majority of cases. A rewrite that keeps the URL keeps every link, every internal link, and the page's history. Publishing the update at a new URL and redirecting the old one buys you nothing and costs you a re-crawl cycle — do it only if the topic genuinely changed.

Deciding *which* pages to keep, merge, or kill across the whole site is a different exercise with different rules. That is a content audit for a service-business website, and it should come before a decay sweep if you have never done one.

How often should a money page be refreshed?

Review your revenue pages every 90 days and your informational posts every 6 to 12 months — but review is not the same as edit. Most quarters, the correct action on a healthy page is nothing.

A workable cadence for a service business with 40-100 pages:

One caution: a real ranking drop takes months to show and months to recover, and a lot of what looks like decay in a 30-day view is just normal SERP movement. Before you rebuild a page, check whether it is decay or just SERP volatility bouncing your rankings, and set expectations with our honest take on how long SEO takes to work.

If your traffic looks flat but your phone stopped ringing, the decay is already six months old and it is sitting in your Search Console data right now. We will pull the URL-level comparison, attribute the cause on each page that lost ground, and tell you which ones are worth the 90 minutes — see what that looks like on our SEO service, or Get my free audit.