Content Decay Finder: Free, Runs in Your Browser

Privacy: nothing you paste or upload leaves your device. The file is read and processed by JavaScript in this browser tab. Nothing is sent to a server or saved.

Search Console will show you which pages lost clicks. It won’t tell you why, and the why decides the fix. A page that slipped five places needs a rewrite. A page whose searches dried up doesn’t, and rewriting it wastes a week. This content decay tool reads the comparison export, sorts each loss into a cause and builds the refresh list in that order. It’s built on the period comparison from 7 Search Console reports most marketers never open, so read that if you haven’t used Compare before.

How to Use the Content Decay Finder

Six steps. The export is the only part that takes care, because a sloppy date range produces a confident list of the wrong pages.

Getting the export right

How to Read the Results

The summary gives the before and after click totals, how many pages are decaying and how many were skipped. Below it, five cards split the lost clicks by cause, and you can click one to filter the table. Then come the priority list and the full table.

A page counts as decaying when it had at least the minimum clicks before and lost at least 15% of them. Anything smaller is treated as noise and labelled Stable. The table is sorted by clicks change, biggest loss first, and you can sort on every column header.

What each cause means and what to do

How the finder separates demand from CTR

When the position held, the finder asks one question: if the page had kept its old click-through rate, how many clicks would its new impressions have brought? The gap between the old clicks and that number is the loss from fewer searches. Whatever’s left, down to the actual clicks, is the loss from fewer people clicking. The bigger part names the cause, and the reason column tells you when the other part was big too.

The refresh priority list then sorts by what you can act on. Pages that dropped out, lost rankings or lost CTR come first, ordered by clicks lost, and demand drops come next. Seasonal pages get a priority number in the table but stay off the list. Nothing you do this month changes them.

Worked example: the sample data

Click Load sample data and you’ll get a made-up shop, Acme Kitchen, on example.com: 20 pages, last 3 months against the previous 3, plus a year over year export for 6 of them. Clicks went from 9,355 to 5,724. Thirteen pages lost 15% or more, 4,366 clicks between them.

Sort by clicks lost alone and the homemade ice cream recipes post tops the list, down 1,330 clicks. Its impressions fell 91% while its position barely moved, 3.6 to 3.9. But the seasonal export shows 140 clicks for the same window last year against 130 now. Seasonal. The sourdough starter post is the same story, 3% above last year. Between them, those two pages hold 1,750 of the 4,366 lost clicks (40% of the drop), and neither needs a single edit.

So the priority list starts somewhere else. The best chef knives post held position (3.0 to 3.2) and gained 4% impressions, yet its CTR fell from 2.54% to 1.11%. That’s a CTR loss, so you’d search the query and see what’s sitting above it now. Second is the cast iron seasoning guide, which slid from 4.1 to 8.9 and lost 23% of its impressions. It’s a real ranking loss, the biggest in the sample, and the page that most needs a proper rewrite. Third, the Dutch oven product page went from 9,800 impressions to none. Check that URL before anything else.

Two rows deserve a second look. The kitchen scale guide is marked Ranking loss (7.5 to 14.2), but its impressions grew 150%, so the finder warns that new low-ranking impressions may be dragging the average down. That pattern is covered in why average position drops while clicks don’t. Then there’s the air fryer chart, a demand drop that isn’t seasonal: impressions down 61% and clicks 62% below last year too. Fewer people want that chart. Fold it into something bigger, or leave it.

Before You Trust a Drop

Most bad decay lists come from bad comparisons, not bad pages. Four checks catch nearly all of them.

• Never subtract one date range from another. Search Console totals aren’t additive, so a period worked out by subtraction can be badly wrong. In a client report described in the impression dilution post, subtracting the second half of a month from the whole month gave 225 impressions for the first half. Pulling the first half directly gave 609. So export every period you compare.
• Compare like with like. Same number of days, same filters, same search type. Google’s own help suggests weekly or monthly granularity when you compare ranges, which cancels out the day-of-the-week effect. And if your business has seasons, the year over year export is the only fair comparison.
• Leave out the newest days. Google labels the newest data as preliminary: it’s still being collected and might change. Google’s announcement of the 24 hours view says Search Console shows data points before it has finished collecting them. End custom ranges 2 or 3 days back.
• Know what the export leaves out. The interface export stops at 1,000 rows, and page rows won’t add up to the property total because Search Console counts the chart by property and the table by page. If you pull through the API instead, top-rows pulls can drop pages and queries that are still ranking: why the Search Console API misses keywords shows how to check.

One more trap. When dozens of pages drop out in the same week, look for a release, a migration or a template change first. Use the six real causes of a ranking drop as the checklist for that.

Limits of This Finder

• It works at page level. A page can lose its main query and gain three small ones, and the page totals hide that. So open the page in Search Console and check the Queries tab before you rewrite it.
• Average position is an average. It’s weighted by impressions, so new deep impressions can pull it down while every real ranking holds. The finder flags the obvious cases, where impressions rose 15% or more alongside a slip. Subtler ones get through.
• CTR loss is a symptom. The finder can’t see the results page, so it can’t tell you whether an AI Overview, a snippet or a competitor’s better title took the click. You’ll have to search and look.
• Seasonal needs last year’s data. Without the seasonal export, a seasonal dip is marked Demand drop with a note. The check only covers pages that appear in both files.
• The thresholds are judgment calls. 15% for clicks, impressions and position is a starting line, since Google doesn’t publish one. Move them and watch how your list changes.
• The 1,000 row cap. On a big site the export is the top 1,000 pages, so smaller pages are missing and one that slipped below the cut can look like it dropped out.
• Different date labels. The headers above come from a real English export. Other interface languages and custom date labels haven’t all been tested. If the finder can’t tell which range is newer, it says so and falls back to column order. Swap periods fixes it.

If the finder says CTR loss on pages that sit in the top 3, the CTR gap estimator shows how far below a typical click-through rate they are. And winning SERP features when position one isn’t the top covers what to do about the blocks above you.

Content Decay Finder FAQ