Why dated content gets skipped by AI engines (and how to refresh it).

AI engines skip dated content because they treat freshness as a signal of whether the information can still be trusted. A page that references last year's rules, quotes prices that have moved, or carries a stale publish date reads as a risk. When an engine like Google's AI Overviews or ChatGPT builds an answer, it wants sources it can stand behind. A page that looks abandoned rarely makes the cut.

This matters more now than it did two years ago. Search used to reward pages that existed. Now the question is whether a page is current enough to be quoted inside a generated answer. If your best content was written in 2021 and never touched again, it may still rank on page one and still get skipped by the engine writing the summary above the results. Those are two different jobs, and dated content fails the second one.

Why do AI engines care about freshness at all?

AI engines care about freshness because they are answering a question in the present tense, and stale information makes them wrong. When a model generates a response, it is making a claim on your behalf. If that claim is out of date, the engine looks unreliable to the person who asked. So the systems that feed these engines lean toward sources that show recent activity, recent dates, and recent facts.

Think about what the engine is protecting. Its whole value is being right the first time. A user who asks about immigration filing deadlines does not want a rule that changed eighteen months ago. A homeowner asking about permit requirements does not want last year's code. The engine has no way to verify every fact from scratch, so it uses proxies. Publish date is one. Last-updated date is another. Whether the surrounding facts match what other current sources say is a third.

Freshness is not the only signal. A well-sourced page from three years ago can still get cited if the topic has not moved. But on any subject where facts change — law, tax, pricing, medical guidance, local regulations — age becomes a strike against you. The engine assumes the world moved on and your page did not.

How do engines decide a page is dated?

Engines decide a page is dated by reading a stack of signals, and no single one carries the whole verdict. The most visible is the date shown on the page and in the structured data behind it. But the deeper signals matter more: whether the facts on the page still match current reality, whether anything has been edited recently, and whether the page still gets linked to and visited.

Here is what gets flagged as old:

• A publish date with no update date. A page stamped 2020 and never touched since tells the engine nothing has been checked in years.
• Facts that contradict current sources. If your page says the standard mileage rate is one number and every current source says another, the engine trusts the others.
• Broken or dead references. Links to pages that no longer exist, tools that have been renamed, statistics with no year attached.
• Named years in the body. "The best strategy for 2022" tells the engine exactly how old the thinking is.
• No recent engagement. Pages that stopped getting traffic and links look abandoned, and abandonment reads as staleness.

A lot of owners assume changing the visible date fixes this. It does not. Bumping a date without changing the content is a trick engines caught onto years ago. If the date says 2024 but the facts say 2021, the mismatch makes the page look worse, not better. Freshness has to be real to count.

What actually happens to your traffic when content goes stale?

When content goes stale, you lose the citation before you lose the ranking, and the citation is where the attention has moved. A page can hold its position in the blue links and still stop appearing in the AI summary that sits above them. Fewer people scroll past that summary. So even a page that "ranks fine" quietly bleeds visits, because the answer got taken from someone else's page and yours never got named.

The pattern shows up in the numbers. A page holds its keyword rankings but its click-through rate drops. The impressions stay flat or rise while clicks fall. That gap is the engine answering the question for the user without sending them to you. Your content did the work of informing the model, and a competitor's fresher page got the credit and the visit.

For a professional services business, this is expensive in a specific way. Your content is often your proof of expertise. A family law firm's guide to custody arrangements is not just traffic — it is the thing that convinces a nervous parent you know what you are doing. If that guide is dated and gets skipped, you lose the visit and the trust-building moment in one go. The cost is not abstract. It is the client who read a fresher, weaker page from a firm across town and called them instead.

Speed compounds this. A dated page that also loads slowly gets skipped twice — once for being old, once for being slow. We wrote about this in why your slow site is a sales problem, not an IT problem, and the same logic applies here. Every technical weakness stacks against you when an engine is deciding whether to trust your page enough to quote it.

How do you refresh content so engines pick it up again?

You refresh content by changing the substance, not the timestamp, and then making the change visible to the engine. A real refresh updates the facts, replaces dead references, adds what has changed since you last touched the page, and cuts what no longer applies. The date should move because the page moved, not the other way around.

Here is the order we work in:

1. Find the pages that matter. Not every page needs a refresh. Start with the ones that earn business — service pages, high-intent guides, anything that used to convert. Pull the pages where impressions are steady but clicks have fallen. That gap is your refresh list.
2. Check the facts line by line. Read the page as if you were the client. Every price, date, rule, statistic, and named year gets verified against current reality. Fix what moved. Delete what died.
3. Add what changed. If the law, the process, or the standard shifted since you published, write the new version in. This is the part engines reward most, because it is the part that makes the page current instead of just re-dated.
4. Cut the filler. Old content often has padding that was written to hit a word count. Cut it. A tighter, more accurate page beats a longer, vaguer one.
5. Update the structured data. Make sure the last-updated date in your schema reflects the real edit. This is the signal the engine reads directly.
6. Rebuild internal links. Point fresh pages at each other. A refreshed page linked to from other current pages looks alive.

Do this on a schedule, not once. Pick your top twenty pages and set a review cadence — quarterly for fast-moving topics, twice a year for stable ones. A page you refresh every six months holds its freshness signal far longer than a page you rewrite once and forget.

This is the work we do inside Search Foundations. We treat content freshness as a maintenance job, not a launch event, because that is what the engines now reward. When we rebuilt the content base for McShanes Solicitors, the refresh work was not one push — it was a rolling review that kept their strongest pages current while competitors let theirs go stale.

What technical work makes a refresh stick?

The technical work that makes a refresh stick is the same work that makes any page trustworthy — clean structured data, fast load times, and a crawlable, well-linked page. A refreshed page with broken schema or a three-second load undoes half the benefit. The engine has to be able to read the freshness signal, and it has to want to trust the page it is reading.

Start with the last-updated date in your schema markup. If your content management system does not surface this, that is the first fix. The engine reads that field directly, and if it is missing or wrong, your real refresh goes uncredited. Then check that the page loads fast enough to keep the crawler and the reader engaged. A slow page gets crawled less often, which means your refresh takes longer to register. The three numbers behind this are worth understanding — we covered them in Core Web Vitals: the three numbers that decide if Google bothers.

Make sure the page is actually reachable. Orphaned pages — ones with no internal links pointing at them — get crawled rarely and refreshed slowly in the engine's index. Link your refreshed pages from your navigation, your related content, and your homepage where it fits. The more a page is referenced from inside your own site, the faster a change registers. Foundations first. A refresh on a broken foundation is a repaint on a leaking roof.

Where this breaks down

Refreshing content will not save a page that was thin to begin with. If the original was 300 words of filler, updating the date changes nothing — there is no substance to refresh. Some pages should be rewritten, merged, or deleted, not refreshed. And no refresh cadence outruns a topic you have no real authority on. Freshness makes good content current. It cannot make weak content credible.