The most effective content program I have watched a team run began by deleting about a third of their blog.
Not refreshing it. Deleting it, with redirects where anything pointed at it, and no replacement.
I lead with that because the refresh conversation has become a way of looking busy without deciding anything. A page is underperforming, so somebody updates the year in the title, adds two paragraphs, changes the published date and hits republish. Nothing about the page’s actual problem was ever identified, and in three months it will be on the list again.
The instinct to work on what exists rather than publishing more is correct. HubSpot’s own reporting has long put roughly 76% of monthly blog views and 92% of blog-generated leads on existing posts rather than new ones, and Backlinko has claimed traffic gains of around 106% from properly updating old material. The priority is right.
The execution is where it falls apart, and it falls apart at the diagnosis.
Three problems that look identical in the report
Open Search Console, sort by clicks lost over twelve months, and you get one list. That list contains at least three different situations that need opposite responses.
The page got worse relative to the competition. Somebody published a better answer. Yours has not changed and the bar moved. This is real decay and it is the only one an edit fixes.
The query changed intent. People typing that phrase now want something different from what they wanted in 2023. A page about “best CRM software” written when buyers wanted feature tables is being read by people who now want pricing and migration effort. The page is fine and it is answering a question nobody is asking any more.
The results page changed shape. An AI Overview appeared above you. A video carousel arrived. Your position is unchanged and your clicks halved. Nothing on your page caused this and nothing you do to the page reverses it.
That third case is now common enough that it should be your first hypothesis rather than your last. I went through how to tell a real demotion from the other kinds of drop separately, and the diagnostic is the same one here.
Ten minutes of looking settles which of the three you have. Search the query yourself and read what is above you. Compare your page against the current top three. Check whether impressions held while clicks fell, which points at the results page rather than at you.
Nearly nobody does this, which is why so much refresh work produces nothing.
The refreshes that make it worse
Rewriting a page that is ranking fine because it appeared on a list. If it holds position four and converts, leave it alone. A page in good standing is an asset and a substantial rewrite is a bet against your own past judgement.
Adding length. A concise page that won because it answered fast does not improve by growing to two thousand words. Half the refreshes I see bury the answer that was the reason the page worked.
Changing the date to look fresh. Google’s guidance is explicit that publishers should not change dates simply to appear current when the content has not substantially changed. Beyond the guidance, it is a lie your readers can detect, and a page dated last week that still recommends a tool discontinued in 2024 is worse than an honest old page.
Change the date when you have changed the substance. Not before.
Refreshing on a calendar. The standard advice is to review everything every three to six months. I would push back on that, because a calendar cadence guarantees you will visit pages that have no problem, and having arrived you will edit something to justify the trip. That is precisely how cosmetic refreshes get manufactured.
Refresh on a trigger instead. A fact in the page expired. A competitor overtook you. The query changed. Something in your business changed. No trigger, no visit.
The part nobody wants to do
Consolidation and deletion, which are not refreshes and are frequently the actual answer.
If you have three pages half-covering the same topic, they are competing with each other for the same query and splitting whatever authority you have across three weaker candidates. Google picks one, usually not the one you would have picked. Merging them into a single strong page beats refreshing all three, and it removes a problem you would otherwise fight forever.
If you have pages with no traffic, no links and no commercial purpose, that inventory is doing nothing except making your site look larger than the amount of useful material in it. Removing them is a real improvement.
Two cautions, because this is the step where people cause damage. Check for inbound links before deleting anything, and redirect to the closest genuinely relevant page rather than the homepage or a 404. And do not delete something merely because it has low traffic if it converts the traffic it gets, since a page with forty visits a month and three enquiries is one of your best assets.
What a refresh worth doing consists of
Once you have established the page is genuinely behind, five things in this order.
Add something that was not there. New first-hand data, a number from your own work, a current price, a screenshot of the thing as it looks now. If the update contains nothing that did not exist before, it is not an update.
Correct what is now false. Discontinued tools, changed rules, superseded figures. This is the part that protects your credibility and it is the least interesting to do.
Move the answer up. Most old posts warm up for three paragraphs before saying anything. Restructuring so the answer sits near the top serves both the reader who is in a hurry and the systems that quote pages.
Fix the title to match how people phrase it now. Query phrasing drifts. A title written for 2023’s version of the search is a small permanent tax.
Add the internal links. A refreshed page is a good moment to point it at your commercial pages and to point your strongest pages at it. I covered where those links should come from separately, and it costs nothing to do while the file is already open.
Then republish, with the date changed, because now you have earned it.
Where I would actually spend the quarter
If you are choosing between eight new posts and this work, do this work.
But do it in the order that hurts. Delete and consolidate first, because that is the decision nobody makes and it improves the site immediately. Then diagnose the decay list properly and only edit the pages where an edit is the right instrument. Then, with whatever capacity is left, publish something new and brief it properly, which I set out in the content brief template.
A refresh program that produces fewer, better pages is worth running. One that produces the same pages with new dates on them is a reporting exercise, and everybody involved knows it.
My notes on content marketing cover where this sits alongside the rest. If you’d like a second opinion on your own decay list, drop me a line on email, WhatsApp or LinkedIn and we can have a quick chat. I’m contracted full time so this isn’t a pitch.