The article graveyard almost every older site has
A company blog usually starts with gusto. Twenty articles appear in the first year, ten in the second, three in the third — and then silence. After a few years the site has forty or sixty texts, most of which nobody reads: how-tos written for a product version that is no longer sold, prices from a time when materials cost a third less, “news” that stopped being new long ago.
The owner knows about that graveyard. But they don’t touch it, because they are afraid of two things at once: that deleting something will break it (“what if someone is searching for that very article”) and that fixing every text is a month’s work they don’t have time for.
Both can be solved with a single decision process. You don’t need to fix everything — you need to answer a few data-based questions about each article and, based on the answers, do one of three things: update it, merge it, or redirect it and let it go.
Why bother digging through old articles at all
The first reason is commercial. An old article has something a new text won’t have for a long while: history. The search engine knows it, the page already shows up for some queries, and sometimes other sites link to it. A new article starts from zero and waits weeks or months for visibility. An update builds on what already exists — which is why it tends to be a cheaper path to enquiries than writing yet another text alongside it. The sheer number of articles solves nothing on its own; why more texts don’t mean more work is covered in how many articles a company website needs.
The second reason is trust. A customer who runs into a 2021 price in an article, or a link leading nowhere, doesn’t think “old article”. They think “sloppy company” — and they carry that over to your quote. The same texts are now also read by AI assistants: when a customer asks about your company or your field, the answer they get may be assembled from exactly the out-of-date article you left lying on the site.
The third reason is practical: old and new articles start competing with each other. When your site has three texts about choosing a heat pump from different years, the search engine picks between them on its own — and often shows the one you wouldn’t have picked.
Stocktake first: three questions about every article
Before you delete or rewrite anything, build a simple table: article URL, year published, topic. Then add three columns with questions.
1. Is it pulling its weight? Open the performance report in Search Console, filter to the specific page and set the range to the last twelve months — anything shorter will mislead you on seasonal topics. Look at impressions, clicks and, above all, queries. A thousand impressions on queries unrelated to your business is not a result; fifty impressions on “gas boiler repair price” for a service company is.
2. Does it lead anywhere? Check whether people move on from the article to a service page, the price list or contact details. And watch out for texts with miserable numbers that still play a role in sales — an article your salesperson sends to prospects, or a text customers refer to when enquiring. That kind of article is the hardest and most expensive to delete. How this “quiet” value of content is measured is covered in how to tell whether content is paying off.
3. Does it overlap with another? When two articles show up for the same queries, you have candidates for a merge. You can spot it in Search Console when two different URLs from your site appear under a single query.
For a site with fifty articles this is one longer afternoon’s work. It isn’t a dozens-of-hours audit — it’s sorting into three piles.
The decision process: three piles
| What the data shows | What to do with the article |
|---|---|
| Gets impressions on meaningful queries, but the content is out of date | Update |
| Two or more articles show up for the same queries | Merge into the strongest one |
| No impressions, no clicks, no role in sales | Redirect and let go |
| Weak numbers, but helps in conversations with customers | Keep and update, don’t delete |
Update: what specifically to change
An update doesn’t mean rewriting the article from scratch. It means going through it with the eyes of a customer who finds it today:
- Fix the facts. Prices, years, legislation, product and version names, numbers in examples. Anything that no longer holds should either be corrected or cut.
- Add the questions people are asking now. In Search Console you’ll also see queries the page shows up for but nobody clicks — often because the text doesn’t answer that question. Add the answer.
- Fix the links. Dead links out and in, and above all add links to service pages created since the article was published. What the linking between articles and services should look like is described in internal links: how to connect articles and service pages.
- Rework the title and description if the page shows up but nobody clicks.
- Change the update date — but only if the content really changed. Rewriting the date and leaving the text untouched is a trick the reader sees through at the first invalid price. On top of that, Google judges how useful the content is, not the date label.
Where to start: with articles that show up for meaningful queries but sit just outside the first page of results. That’s where the ratio of work to effect tends to be best — although nobody can honestly promise you a specific movement, and we deliberately stay away from ranking guarantees.
Merge: when articles fight over the same query
The typical situation: “How to choose a boiler in 2021”, “Choosing a boiler 2023” and “What to watch out for when buying a boiler”. Three texts, one topic, one query — and the search engine doesn’t know which to show.
The process is always the same:
- Pick the stronger URL. The one with the bigger numbers in Search Console, or the one other sites link to.
- Move everything of value from the weaker texts into it — paragraphs missing from the stronger article, better examples, answers to questions.
- Permanently redirect the weaker URLs (with a 301 redirect) to the stronger one. That carries over what the old URLs accumulated over the years.
- Fix your internal links so they point straight at the destination article, not through a redirect.
The result is one article that covers the topic properly and can be updated every year — instead of three half-finished ones ageing separately.
Let go: when to delete, and how
The third pile holds texts where all three answers are negative: they don’t pull their weight, they don’t lead anywhere, they don’t overlap with anything. Typically invitations to events that have been and gone, “news” about withdrawn offers, and off-topic articles from the days when the blog was filled with anything at all just to keep something coming.
Even here, though, delete deliberately:
- Redirect the old URL to the closest related page — a newer article on the same topic, a blog category, the page for the service the text was about.
- Don’t redirect everything to the homepage. According to Google’s own documentation, a redirect to a page unrelated to the original content is treated much as if the page didn’t exist — you save nothing by it.
- When there is no related page, let the URL return a 404. That’s fine; it’s cleaner than forcing readers onto a page they weren’t looking for.
What to avoid during the clean-up
Deleting on traffic alone. The most-read article isn’t necessarily the most valuable, and vice versa — a text with thirty readers a year that helps close deals is worth more than a thousand random visits. So always ask about its role in sales too, not just about the graph.
A bulk “update” by date. Rewriting the date to this year across thirty articles is ten minutes’ work and months of damage. Once a customer hits invalid information under a fresh date, they stop trusting even the articles that are current.
Merging without a redirect. Deleting the weaker article and letting its URL fall into the void means throwing away everything it accumulated over the years. A redirect is a few extra minutes of work.
Recycling a URL for a different topic. Rewriting an old article about boilers into a text about air conditioning because “that URL already has history” doesn’t work — history attaches to the topic, not the URL. A new topic belongs on a new URL.
How to handle it without it eating a month
Don’t turn it into a project. Devote one afternoon to the stocktake and sorting into three piles. Then take two or three articles a month from the “update” pile — start with the ones whose visibility is within reach, and the ones where out-of-date information is directly hurting sales. Merging and deleting can wait; invalid prices on commercial pages can’t.
If you aren’t sure what the data says — which articles are pulling their weight, which overlap and which have a quiet role in sales — this is exactly the job for one sitting with an outside reader of that data. The initial analysis goes through Search Console, your site and your real enquiries and returns a specific list: what to update, what to merge, what to let go, and in what order. That’s a cheaper start than a year of wondering whether to begin the clean-up at all.
And so the graveyard doesn’t come back: reviewing old texts belongs in the regular care of your site, not among one-off campaigns. What that ongoing care involves and when to hand it over is covered in content management for your website — and what it looks like when we do it is in our service offering.
An old blog isn’t a burden. It’s work in progress you have already paid for — and cleaning it up is usually the fastest way to finally get enquiries out of it.
