Every site that has been publishing for a few years carries them: a landing page built for a service the business still sells, live for eighteen months, and never once the reason anybody visited. The instinct is to delete it. The instinct is right about a third of the time.
A page with no traffic has either lost a competition or never been entered into one. Those look identical in an analytics report and need opposite responses.
What counts as a fair run for a new landing page?
Long enough to be indexed, linked and crawled repeatedly — which in practice means a few months, not a few weeks. A page published in March and reviewed in April has not failed; it has barely been assessed.
The clock that matters starts when the page is genuinely reachable, not when it was published. A page that sat unlinked for a year and was added to the navigation last month is one month old for this purpose, whatever the byline date says.
Was the page beaten, or never entered?
Search Console answers this in one screen. Filter to the URL and read the impressions, because impressions and clicks tell you different halves of the story.
| What you see | What it means | The move |
|---|---|---|
| No impressions at all | Not indexed, or indexed and never a candidate | Check index status and internal links first |
| Impressions, average position past 30 | Indexed but not competitive on this query | Merge into a stronger page |
| Impressions, position 8–15, few clicks | Ranking, but the snippet is not winning the click | Rewrite title and description |
| Impressions on the wrong queries | The page answers a question you did not intend | Rewrite to match, or accept the new topic |
| Impressions climbing, clicks flat | Still maturing | Leave it alone for another quarter |
The first row is the one people misread most often. Zero impressions is not evidence that the page is bad. It is evidence that the page was never in the auction, and the usual cause is structural rather than editorial.
Zero impressions is a distribution problem; poor position with impressions is a quality problem. Deleting a page for the first reason destroys work that was never given the chance to fail.
How often is the real problem internal links?
Often enough that it should be the first thing checked. A page nobody links to is an orphan page: reachable by URL, effectively invisible to the site's own structure, and receiving none of the page authority that flows through internal links.
Crawl your site and list every URL with zero internal links pointing at it. If the page you were about to delete is on that list, you have not tested it yet. Add three or four contextual links from pages that already rank on related queries, then wait a quarter. That is a genuinely cheap experiment, and the result is unambiguous either way.
When is merging better than deleting?
Whenever the page contains something the target page lacks. Deleting throws away the paragraphs, the examples and any links the page picked up. Merging keeps them and consolidates the signals into one URL rather than splitting them across two.
The mechanics are unglamorous and matter more than the decision: move the genuinely useful sections into the stronger page, redirect the old URL to it, and update the internal links that pointed at the old URL so they go straight to the target rather than through a hop. A redirect left in place while every internal link still points at the old address is a job half done.
Straight deletion is right in one case: the page is thin, duplicative, has no links, and there is nothing on it worth carrying anywhere. Then let it return a 404 and move on. A 404 for a page nobody wanted is not a problem.
Auditing on organic traffic alone. A page can be dead in search and still be the destination for an email campaign, a paid ad, or a link printed on a business card. Check every channel before deleting, because the sales team will notice before your report does.
Which pages should stay even though they will never rank?
Several kinds, and they are worth naming so they stop reappearing in every audit:
- Pages that serve paid or email traffic. Their job is conversion, not ranking, and on-page SEO metrics are the wrong yardstick.
- Pages that answer a question sales gets constantly. Being a link a salesperson can paste is a real job.
- Pages that exist for credibility. Nobody searches for your accreditation page; some people check it before signing.
- Pages that support a cluster. A supporting article that sends signals inward can earn its place without earning traffic.
What these have in common is a purpose stated somewhere other than the traffic report. If a page has one, record it — a line in your content inventory saves it from the next audit and from the one after that.
How do you keep this from becoming an annual clean-up?
By writing the review date and the success condition when the page is published. "This page should be earning impressions on plumbing-repair queries by December" is a testable claim, and in December it either happened or it did not.
Without that, every audit restarts the same argument from scratch, with nobody able to say what the page was supposed to do. The pages that survive audits are then whichever ones somebody remembers commissioning, which is not a strategy.
Take the ten pages you are least sure about, pull their impressions, and split them into two piles: never entered, and entered and lost. Only the second pile is a content decision — the first is a linking one, and you can start it this afternoon.
Sitting on a hundred pages that never went anywhere?
We will audit them into four piles — improve, merge, redirect, leave alone — and tell you which ones are worth the hour.
Get in Touch →Does this apply to your site?
Reading about it is one thing. Point the scan at your own site and see whether this applies to you, and what it is worth fixing.
Free and unlimited. No account, no card, and you get every finding rather than a teaser.