5 proven Claude + Hygraph MCP workflows that actually save you time

Your content has an expiration date: how to spot and fix content decay

Content decay costs more than traffic now. Learn what it is, why AI makes it urgent, and how to spot and fix decaying pages before AI repeats them.
Jing Li

Written by Jing 

Nov 13, 2026
What is content decay? How to spot and fix it (2026)

Has this ever happened to you? You update the pricing page when your product price changes. However, the comparison page, the FAQ, the help article and last spring's blog post are left out, and nobody notices until a customer quotes the old number to your sales team.

Every page you publish starts aging the day it goes live. Prices move, specs change and policies get rewritten, while the pages describing them stay frozen. That used to be a slow, quiet problem, but AI changed the pace. Assistants read your pages faster than your team ever will, and they repeat what they find with no way of telling which version is current.

That leaves you with two risks. Outdated content loses relevance, so it slowly drops out of search results and AI answers. Inaccurate pages are worse, because they keep getting quoted, and your customers walk away with answers you'd never stand behind.

Both are content decay, and in our view the wrong answers do far more damage than the lost traffic. This guide shows you how to spot both and fix them before an AI assistant does the talking for you.

Before we start, ask yourself how confident you are about what people are being told about your product today.

#What is content decay?

Content decay is the gradual decline of a page's performance and reliability over time. It shows up as falling rankings, traffic and AI citations, and less visibly, as facts on the page that are no longer true. Every page decays eventually, some within weeks and some over years.

The first half of that definition gets most of the attention. A page that used to rank on page one slides to page two and takes its traffic with it. That's real and worth fixing, but we split content decay into two problems:

  • Traffic decay: the page loses rankings, clicks and visibility.

  • Accuracy decay: the page still ranks, still gets read and cited, but some of what it says is wrong.

The first costs you visitors. The second costs you your customers' trust, which takes far longer to win back.

#The half-life of a fact

In 2012, scientist Samuel Arbesman published a book called The Half-Life of Facts: Why Everything We Know Has an Expiration Date. His argument was that knowledge decays at a fairly predictable rate, much like radioactive material. Some fields move faster than others, but every fact has a shelf life.

Business content works the same way, only faster. A price might hold for a quarter. A product spec holds until the next release. A return policy or a compliance statement might last a year or two. Your pages, on the other hand, come with no expiration date at all. They stay published, looking exactly as confident as the day they went live, until someone edits or removes them.

The trouble starts because most facts don't live on one page. Take a single price. It sits on the pricing page, in two comparison posts, in an FAQ, in a help center article, on a partner landing page and in a sales one-pager someone uploaded as a PDF. When the price changes, the pricing page gets updated. The other copies drift, one by one, into being wrong.

We'd rank accuracy decay as the more critical kind. A page that loses traffic hurts you quietly, because fewer people see it. A page with a wrong fact keeps working against you, because people still see it, believe it and act on it. Now AI tools repeat it at scale too, often without a link anyone could click to check.

If your team runs several brands, markets or channels, multiply that by every locale and every site. A pricing update that reaches the main site and misses three regional ones is a story we hear often.

#Why content decays (and why it decays faster now)

The causes of content decay have barely changed over the years, but the speed has.

Competitors publish better, fresher pages

Someone covers your topic in more depth, with newer data and a clearer structure. Search engines and AI assistants notice, and your page slides down. Your page stayed the same while the pages around it got better.

Search intent shifts

People searching for a term start wanting something different. A query that once called for a definition now calls for a comparison, or a tutorial now competes with a two-line AI answer. If your page answers last year's question, it loses to the pages answering this year's.

Facts go stale

Prices, specs, screenshots, statistics, product names, team members and links each have their own half-life. Every one that expires makes the page a little less trustworthy, to readers and to machines alike.

Your own pages compete with each other

Over the years, most sites collect several pages covering the same ground. They split rankings between them, and they often disagree with each other. When two of your pages give different answers, search engines have to guess which one is right, and so do AI assistants.

AI answers take the click and prefer fresh sources

This is the part that speeds everything up. When Google shows an AI summary, people click a regular search result on about 8% of visits, compared with 15% without one. Your page can hold its position and still lose nearly half its clicks.

AI assistants also lean toward newer content. When Ahrefs analyzed about 17 million URLs cited by AI assistants, the cited pages were on average about a quarter newer than the pages ranking in regular search. Citations don't last long either. Scrunch tracked 3.5 million citation events and found that half of a typical source's AI citations disappear within four to five weeks.

Put those together and reviewing your content every six months starts to look slow. That rhythm was built for search engines that recrawled now and then, and for readers who clicked through to check. AI assistants refresh their sources in weeks and it’s hard for teams to keep up with that by hand, and we don't think they should have to.

#Types of content decay

Decay doesn't look the same in every report. We group it into five types, each with its own fingerprint.

Ranking decay

Your positions slide over weeks or months as competitors improve or intent shifts. Impressions and clicks drop together.

Click decay

Rankings hold, but clicks fall. AI Overviews, featured snippets and other search features answer the question before anyone reaches your page.

Intent decay

The page still ranks for some terms but loses the ones that matter, because searchers now want a different format or a different answer.

Citation decay

AI assistants stop citing your page, or start citing a competitor's newer one instead. This often happens while your organic rankings look fine, which makes it easy to miss.

Accuracy decay

The page still ranks, still gets cited and still converts, but something on it is no longer true. This is the type we worry about most, because none of your traffic dashboards will flag it. The numbers look healthy right up until a customer, a support ticket or an AI answer exposes the problem.

#What content decay costs you

Lost traffic and leads

This is the cost everyone sees. Decaying pages lose visitors, and the first pages to decay are often your best performers, simply because they've been live the longest. If those pages feed your pipeline, the pipeline shrinks with them.

Lost visibility in AI answers

More of your buyers' research now happens inside ChatGPT, Perplexity and Google's AI Mode. If your pages drop out of those answers, you drop off the shortlist, and you may never see it in your analytics.

Wrong answers about your brand

Here's the cost nobody puts in a report. AI tools repeat what they find, including your current pricing page and the three-year-old blog post with the old price. They have no way of knowing which one you meant.

Air Canada found out what that costs. In 2024, a customer took the airline to a tribunal after its website chatbot told him he could apply for a bereavement fare refund after booking. The airline's own policy page said the opposite. Air Canada argued that the chatbot was responsible for its own words. The tribunal disagreed, called the chatbot part of Air Canada's website, and ordered the airline to pay. The customer, it said, had no reason to know which of the two versions was correct.

That case was about a chatbot, but the lesson applies to every page you publish. When two versions of a fact exist, your customers can't tell which one is right, and you're the one held to the wrong version.

That's why we see decayed content as noise, and harmful noise at that. It carries your name, sounds like you and misleads the people you're trying to win over.

#How to spot content decay

1. Signs in your data

Three patterns are worth watching. Impressions slide slowly over several months. Click-through rate falls while the average position holds. Rankings drop for specific queries that used to bring in traffic. Any one of these deserves a closer look, and two together usually mean the page is decaying.

2. How to measure it

Compare two equal periods in Google Search Console, page by page. For example, take the last three months against the same three months a year earlier, and sort by the biggest drop in clicks. Set a threshold that fits your site (many teams flag anything down 20% or more) and start with those pages. Before you call it decay, rule out seasonality, tracking changes and technical issues like broken redirects or pages dropping out of the index.

3. Check what AI says about you

Your analytics won't show citation decay or accuracy decay, so ask the AI tools directly. Open ChatGPT, Perplexity and Google's AI Mode and ask the questions your buyers ask. What does your product cost? What's included in each plan? What's your refund or data policy? How do you compare with a named competitor? Note which of your pages get cited and whether the answers are right. Every wrong answer points you to a page that needs fixing.

4. Audit your time-sensitive facts

List the facts with a short half-life, such as prices and plans, product specs and limits, availability, legal and policy copy, statistics, dates and the people you name. Then write down every place each fact appears. The list is almost always longer than anyone expects, and it's the best map you'll have of where accuracy decay will show up next.

5. How often should you check?

Set the rhythm by fact type. Prices and plan details should be checked every time they change, ideally automatically. Product specs need a review with each release. Policies and legal copy follow your legal team's cycle. Evergreen guides can run on a quarterly traffic review. A single "audit everything twice a year" plan checks some pages far too often and others far too late.

#How to fix content decay

Update, consolidate, redirect or prune

Situation What to do
The page still matches what people search for, but it's outdated or thinner than competing pages Update it
Several of your pages cover the same topic and split rankings Consolidate them into one page and redirect the rest
The topic is gone, but the URL still has links or traffic Redirect it to the closest relevant page
The page has no traffic, no links and no purpose Prune it (unpublish or noindex)

Our default is to update. A page with history, links and rankings is worth more than a brand-new one, and most decay can be reversed with a real update.

Content decay vs. content pruning

The two often get mixed up. Content pruning is one way to deal with content decay: you remove or merge pages that no longer earn their place. Pruning has its uses, but we’d reach for it last. Deleting a decaying page means giving up on it, and many of those pages only need their facts corrected.

Fix the fact at its source

When a fact is wrong, resist fixing it page by page. Use your fact audit to find every place it appears, decide which version is true, and update it in one place that every other page pulls from. If your setup doesn't allow that, fixing the copies by hand is still better than leaving them. Treat it as a sign that your content structure needs work.

Don't just change the date

Bumping the "last updated" date without changing anything is tempting, and it doesn't work. Google's John Mueller has said that changing dates alone won't improve rankings, and readers notice when a "2026" guide still quotes numbers from three years ago. It also does real harm, because a fresh date on a stale page tells AI assistants your outdated facts are current. Update the content first, then the date.

#How to prevent content decay

Store facts once and reuse them everywhere

The most effective prevention is structural. When your prices, product limits and policy text live as structured content, stored once and referenced by every page that needs them, a change happens in one place and reaches all of them. A fact you paste into ten pages will eventually be wrong on some of them, while a fact you reference stays the same everywhere.

The same structure is what makes your content usable for AI. Content that's structured, current and governed gives AI tools context they can rely on.

Give every fact an owner and a review date

Most teams assign owners to pages, but rarely to the facts on them. Assign each time-sensitive fact to the team that knows it best, such as pricing to product marketing, policies to legal and specs to product, and give it a review date. When the date comes up, the owner confirms it or changes it.

Plan updates by content type

Match your review rhythm to how fast each kind of content ages. News and announcements barely need upkeep, as long as they're clearly dated. Tutorials need a check with every major release of the tools they cover. Comparison pages and listicles need a full refresh at least once a year. Pillar guides deserve a quarterly look at their traffic and AI citations.

#Let a workflow catch content decay

Everything above works. It also takes a lot of people's time, and it still runs on a schedule while AI tools run around the clock. Content decay needs a workflow that runs continuously, and we think every team publishing at scale will end up with one. A good one spots what changed, proposes a fix and leaves the final call to someone on your team.

MAD Design Group already runs one with Claude and the Hygraph MCP. Their workflow keeps every translated page in step with its source, so an edit to an English page doesn’t leave an outdated version live in another language. Each entry has a status field. When someone edits a page that’s already translated, they set it back to pending, a scheduled run retranslates it, and a second agent reviews the result before it goes live. When they improve their translation rules, they ask the AI in plain language to reset every entry translated more than two months ago, and it finds and updates those entries itself, logging each change so it can be undone.

The next step is content that heals itself, where the system notices that a fact has changed, finds every page that repeats it and prepares the updates for a person to approve. That's where content management is heading. Content keeps getting more demanding, and keeping it accurate by hand doesn't scale.

#Your content has an expiration date. What are you going to do about it?

Every page you publish will go out of date. You can't stop that, and you don't need to. What you can control is whether you find out before your customers do, or before an AI assistant repeats the old version to someone deciding whether to buy from you.

Start with your fact audit, ask the AI tools what they say about you this week, and fix wrong facts at their source. Then move the checking into a workflow, so your team spends its time on judgment calls instead of hunting down old prices.

Want to see how structured content keeps your facts consistent across every page, channel and AI tool? Take a look at Hygraph, the agentic content platform and headless CMS.

Frequently Asked Questions

Blog Author

Jing Li

Jing Li

Jing is the Organic Growth Lead at Hygraph. Besides telling compelling stories, Jing enjoys dining out and catching occasional waves on the ocean.


Share with others

Sign up for our newsletter!

Be the first to know about releases and industry news and insights.