Content pruning

What content pruning is

Content pruning is deciding which pages to delete, deleting them, and making sure every address that disappears points somewhere sensible afterwards. It is subtraction. A site that has been publishing for five or six years is usually carrying a few hundred pages that earn nothing and make the twenty that matter harder to find.

It is not merging. Where several pages on one subject each hold something worth keeping, they get combined instead, and that is content consolidation. It is not updating either. A page that is out of date but still about something you sell gets rewritten under a content refresh plan. Pruning is for the pages where neither of those is worth anybody's afternoon.

When you need it

Three situations, and the first is much the most common. You have an archive: years of posts written to a schedule, about subjects that mattered that month, by people who have since left. Nobody has read most of them since the week they went up, and nobody wants to be the person who deletes them.

The second is a rebuild. Carrying dead pages onto a new site costs money twice, once to move them and once to keep them, which is why pruning normally happens just before a content migration plan is written.

The third is a site where the good pages are genuinely hard to find because they sit among eighty near-identical ones. You do not need pruning if your site is under about thirty pages. At that size nothing is hidden, and a content audit will tell you more for the same money.

What gets deleted

A page is only proposed for deletion when it fails on every count, not one. The test is applied page by page and the reason is recorded beside each one.

• No visitors and no purpose. Nobody has arrived on it in a year, no other page links to it for a reason, and it answers no question a buyer actually asks.
• Expired. Events that happened, offers that closed, products withdrawn, positions filled. These are the easiest deletions and the ones most sites have most of.
• Embarrassing. Old prices, departed staff written about in the present tense, claims the business would not make now. These go because a buyer might read them, not for anything to do with visits.
• Generated by the platform. Empty tag pages, author archives for somebody who wrote one post, date archives, duplicate print versions. There are usually more of these than real pages, and no reader has ever wanted one.

Nothing is deleted if it still brings visitors, if another site links to it, if anyone in the business can name a time it helped a sale, or if it has to exist for a legal or professional reason. Those four exemptions are checked before anything goes.

What you get

The delete list. One row per page, with the reason, the evidence, and an empty column for your objection. You go through it and veto whatever you like. I would much rather argue about twenty pages than have you approve two hundred without looking.

The redirect map. Every deleted address with the page it will point to instead, or marked as deliberately gone where nothing sensible exists. Handing a developer this file is the whole of the technical work.

The record. A saved copy of every deleted page, and a short note of what was removed and when. Deletions are reversible for about a year if that record exists, and permanent if it does not.

A worked example

An illustration. An industrial supplier in Hamilton has 612 addresses on its site. Ninety of them are real: product pages, technical guides, company pages. The rest are 214 blog posts going back to 2013, 180 empty tag pages, 98 date archives and 30 pages for trade shows that have been and gone.

The tag pages and date archives go without discussion. Of the 214 posts, 26 still bring visitors or are linked from the technical guides and are kept; 9 cover subjects that still matter and are handed to the refresh plan; 11 overlap on one subject and go to consolidation; the remaining 168 are deleted and redirected to the relevant product category. The trade show pages are deleted outright. The site ends at about 130 pages, the owner can describe what every one of them is for, and the two hours a month somebody spent tidying the archive stops being spent.

How it runs

1. A half-hour call. How old the site is, what has been published, and who is nervous about deleting things. No charge.
2. Access. A full list of addresses, read-only access to your analytics, and whatever the platform can tell me about which page links to which.
3. The pass. Five to ten working days for a site of a few hundred addresses. Every page is looked at; none of this is decided by rule alone.
4. Your veto, then the map. You review the list and strike out anything you want kept. The redirect map is built from whatever survives, and either handed to your developer or run by me.

What it costs

Quoted in CAD as a fixed price after the half-hour call and agreed in writing before anything starts. It depends almost entirely on how many addresses there are, since the judgment is made page by page. Where pruning follows a content audit much of the reading is already done and the price reflects that; the audit's own price is published on the pricing page.

What happens next

Two things usually follow straight away. The pages set aside as worth keeping go into a refresh plan, and the overlapping groups go to consolidation. Longer term, the way to avoid ever doing this again is a content lifecycle plan, which puts a review date on each page the day it is published so that nothing quietly rots for six years. None of it is assumed. Pruning is finished when the site is smaller and every page left on it has a reason to be there.