How to know which content works (and what to do with the rest)
Which metrics reveal whether a piece of content works, how to decide between updating, merging or deleting, and why deleting pages sometimes raises traffic.
Digital Marketing, Medical Marketing & AI Consultant · btodigital
Updating an article that already has impressions pays better than writing a new one. It is the conclusion I find hardest to get accepted, because publishing something new feels productive and revising old work does not.
But the numbers are clear: a page at position 11 that moves to 6 can multiply its clicks. A new page takes months just to reach position 11.
Which metrics tell you whether content works?
Not visits. These four, and they have to be read together:
| Metric | Where to see it | What it reveals |
|---|---|---|
| Impressions | Search Console | Whether anyone cares about the topic |
| Average position | Search Console | How close you are to receiving clicks |
| CTR | Search Console | Whether the title and description convince |
| Conversions from that page | Analytics | Whether the content serves the business |
Combined, the four give a fairly precise diagnosis:
- Many impressions, poor position: the topic is wanted and the content does not compete. Rewrite.
- Good position, low CTR: the content is fine and the title is not. Change title and description.
- Good CTR, zero conversions: you are attracting the wrong people, or the call to action is missing.
- Nothing at all: nobody searches for that. Merge or delete.
Update, merge or delete?
| Situation | What to do |
|---|---|
| Impressions and position between 8 and 20 | Update. The best investment available |
| Two articles competing on the same topic | Merge into one and redirect |
| Old data but a current topic | Update the figures and the date |
| No impressions in 12 months and off-topic | Delete and redirect to the nearest thing |
| Well ranked and no conversions | Review who it attracts and what you offer them |
Merging is more common than it sounds. Two of your own articles on the same topic weaken each other and neither wins.
Can deleting content raise traffic?
Yes, and it is counterintuitive. If you have a hundred pages and thirty receive no impressions in a year, those thirty dilute the signal from the ones that matter.
Careful with the scissors, though:
- Verify before deleting. A page with no search traffic may be receiving links or visits from an email.
- Always redirect, never leave a 404 where content used to be.
- Do not delete in bulk. Review one by one; it is tedious and it prevents accidents.
How do you prioritise without losing weeks?
My monthly routine, which takes about an hour:
- Open Search Console, last three months, sorted by impressions.
- Flag everything between positions 8 and 20. That is the working list.
- From that list, prioritise what relates to what you sell.
- Pick two or three and work them properly.
- Note today’s position so you can compare in a month.
Step five is the one almost nobody does, and without it you cannot know whether anything worked.
What does “updating” an article mean today?
A few years ago it meant changing the date and adding paragraphs. Today, with AI-generated answers taking part of the clicks, updating means something else:
- Turn headings into the questions people actually ask.
- Put the direct answer in the first lines of each section, not the last.
- Add a table if there is anything to compare.
- Add a frequently asked questions block.
- Include at least one figure of your own that does not exist elsewhere.
- Remove what aged badly, especially figures and version names.
That last point matters more than it seems. A technology article naming a discontinued version loses credibility instantly.
Frequently asked questions
How do I know whether a blog article is working? Cross four data points from Search Console and your analytics: impressions, average position, CTR and conversions from that page. Together they tell you whether the problem is the topic, the title, the content or the audience.
Is it better to write new content or update old content? Update, almost always, when you have pages between positions 8 and 20. They already have history and signals; moving them up five places is faster and cheaper than starting from zero.
Does deleting articles improve SEO? It can, if they are pages with no impressions and off your topic. Verify first that they receive no links or traffic from elsewhere, and always redirect rather than leaving a 404.
How often should I review content? A monthly hour in Search Console is enough to pick two or three pieces. A full audit, once a year.
Is it enough to just change the update date? No, and it backfires. If the content did not change, it is misleading and improves nothing.
How long before an update shows an effect? Between two and eight weeks, depending on how often your site is crawled. Note the position on the day you publish so you can compare.
How do AI-generated answers affect this? They change what updating means. It now matters more that each section answers a complete question in a few lines, with tables and original data, because that is what gets cited.
