- Click decay is long-form pages whose clicks fell more than 20% versus the previous calendar month, after more than 10 clicks in that prior month.
- Compare complete calendar months, not a rolling 28 days mixed with an in-progress month. The current month versus the previous month is the window.
- Sort first by how many clicks were lost, then by percent change. A page that went from 80 clicks to 40 is a different week than a page that went from 12 to 5.
- Restrict the queue to long-form content paths. Product and app URLs often move with launches and seasonality, so this list is limited to long-form paths.
- Answer engines tend to cite URLs that already look like answers. A page that is losing Google clicks is also at risk of losing citations. The list itself still comes from Search Console, not from ChatGPT.
You already have long-form URLs that earned clicks last month and earned fewer this month. The drop is more than 20%, from a previous month with more than 10 clicks.
Teams often notice decay in a traffic report, then brief a new article on the same topic. The click-decay play is a short queue of those URLs so you diagnose the page that already ranked, before you commission a replacement.
Why decaying pages are worth the work
More than 10 clicks in the prior month means the URL had a real job. A 20% drop from that base is large enough to notice and small enough to still be the same page. Under that floor, a change from 4 clicks to 2 can look like a large percent drop even though the sample is small.
Clicks, not impressions, are the signal. Impressions can bounce with SERP layout. Clicks show searchers choosing your result, then choosing it less often.
Refreshing a decaying URL is usually hours, not a new brief, subject-matter-expert (SME) review, and a wait to get indexed. The purpose is to diagnose the URL that already ranked, instead of publishing a second article because clicks fell.
Keep the list short enough to work through. A handful of material drops is easier to act on than a percent-change sort of every URL.
Answer engine optimization (AEO) often involves the same URLs. Models cite pages that already look like answers for a query. If Google is sending fewer clicks, check the page and its queries before assuming the topic no longer has demand. This list still comes from Google Search Console (GSC), not from an AI citation report. The same URLs are often the ones answer engines can cite once the page states the answer clearly. If you are weighing recovery against net-new AEO work, how to budget for AEO and SEO is the planning frame.
How to use Search Console to build a click-decay queue
Start from two complete calendar months of page-level Search Console data: current versus previous. Skip an in-progress month if it is still empty or partial and you would be comparing a stub to a full month.
Then apply the filters in order:
- Keep long-form content paths.
- Keep pages with more than 10 clicks in the previous month.
- Keep pages whose clicks fell more than 20% versus that previous month.
- Sort by absolute click change (largest losses first), then by percent change.
The columns that matter are page, clicks (current), clicks (previous), change, and percent change.
You don't need érculespot to run this. In Search Console, export page clicks for the last two full months, filter to long-form URLs with more than 10 prior clicks and a drop over 20%, sort by lost clicks, and stop at a short queue. érculespot is one way to see that list assembled.
How to read a click-decay row
The pane title in the example is “Click decay.” That is the diagnostic name. The work is to understand and, where appropriate, recover pages that lost clicks.
Read a row left to right. The page is the URL that had a job last month. Clicks (previous) is the baseline. Clicks (current) is now. Change and percent change show how large the drop is.
In the example, one long-form URL went from 12 clicks to 5: −7 clicks, −58.3%. That clears the floor (more than 10 prior clicks, more than 20% down). It is a small absolute loss, but it still meets the filters because the page had demand and then lost most of it. A larger site will fill this list with bigger absolute drops, such as 80 clicks to 40. The same filters apply.
If the URL is a recap of a conference, a version-specific migration, or a time-boxed announcement, decay may be the topic ending. Otherwise, the ranking page is the place to edit.
Why month-over-month clicks are the right decay signal
Calendar months match how Search Console is stored and how editorial calendars run. A rolling 30-day window mixed with an unfinished month will flag every page in the first week of the month.
The 20% floor excludes small month-to-month movement such as 11 clicks to 10. The 10-click floor excludes tiny bases such as 3 clicks to 2. Together they keep the queue on pages that had a job and then lost a material share of it.
A new URL for the same topic often splits the remaining clicks instead of recovering them. That is a different play (overlapping pages). For click decay, the usual first step is to diagnose the existing page rather than publish a new one.
What to do with a click-decay row
Open the ranking page. Compare the two months in Search Console at the query level. Then:
- Check whether impressions fell, position worsened, or CTR fell. Those three are different first edits.
- Search the queries that used to click. See who occupies positions 1–3 now.
- Update facts, dates, and missing sections if the SERP moved.
- Align the title if CTR fell while position held.
- Don't rewrite the URL because a single month dipped if the topic is seasonal.
If the drop lines up with a broader measurement change, Google killed “num=100”. It's changing your performance metrics. is useful context if the drop lines up with that measurement change. Choose the row based on the URL, its queries, and the two months you are comparing.
Ship the row, then look at the next month's queue. Success looks like the page falling out because the drop shrank under 20%, or because you can explain the remaining drop (season, SERP feature, retired topic).
Click-decay work is a filter on pages you already have
This list is drawn from pages you already published. The URL had clicks, and it has fewer now. A short list of material drops is a practical amount of diagnostic work.
Making the answer explicit on those same pages can help with answer engines as well as Google. You do not need a separate AEO dashboard to start. Adapting your content system for AI search is the broader operating model. This list is one working set inside that system.
Frequently asked questions
What does click decay mean in search engine optimization?
Here it means a long-form URL whose Search Console clicks fell more than 20% versus the previous calendar month, after more than 10 clicks in that prior month. It is not a lifetime traffic chart, and it is not an impressions-only dip.
Why require more than 10 clicks in the previous month?
Because percent change from a tiny base is hard to interpret. 12 clicks to 5 is a real loss of a small amount of traffic. 3 clicks to 1 is too small a sample to prioritize. The floor keeps the queue on pages that had demand.
Why compare calendar months instead of a rolling 30 days?
Because an in-progress month will look like decay on every URL. Complete months are comparable. They also match how you store GSC history.
Why restrict the click-decay queue to long-form content paths?
Homepage, pricing, and app URLs move with launches, ads, and seasonality. Long-form paths are where a month-over-month click drop is more often something you can address with an editorial update.
Does click decay still matter with AI Overviews and answer engines?
Yes, with a limit. The list is still GSC, not an AI citation report. Fewer Google clicks can mean a SERP feature stole the click, or it can mean the page is no longer the answer. Check the queries before you retire the URL. You can start this work in Search Console. A separate AEO tool is not required to review a page that just lost a fifth of its clicks.
Related reading
- Content updates: Refresh or rewrite? — How to decide whether a decaying URL needs a pass or a replacement.
- Did ChatGPT click traffic suddenly decline? — A reminder that not every click drop is a content problem; some are channel and attribution.
- Google Search Console: Getting Started — How to export the month-over-month page report this play is built from.
