Writing  ·  Playbook  ·  June 2026

When to merge, when to kill: a content portfolio decision tree.

For the team staring at 8,000 URLs and no clear framework. The questions that actually determine whether a page gets consolidated, culled, or kept — and in what order to ask them.

Content audits tend to produce one of two outcomes: a spreadsheet that nobody acts on, or a consolidation programme that deletes the wrong things. The first happens because the audit doesn't produce prioritised decisions, only data. The second happens because the decision framework is wrong, usually traffic-based, which treats low-traffic pages as low-value pages, which is often backwards in B2B.

Here is the decision tree I use. It won't produce a perfectly optimised site in one pass. Nothing does. What it will do is give you a consistent, defensible framework for making decisions at scale, and a prioritisation logic that ensures you work on the highest-leverage pages first.

Before you start: the questions the data can't answer.

Content audit frameworks that start with traffic are building on shaky ground. Traffic is a lagging indicator of past decisions and past search demand: it tells you what has happened, not what should happen. Before you pull any data, answer these three questions for your specific situation:

  1. What does the site need to do in the next 12 months? A site that needs to capture new category demand has different consolidation priorities than a site that needs to convert more of its existing traffic. Know the goal before you build the framework.
  2. Which pages are commercially important regardless of traffic? In B2B SaaS, a product page for a feature that is actively being sold to enterprise accounts is commercially critical even if it has 200 visits a month. Never delete or consolidate commercially critical pages based on traffic alone.
  3. Which pages support the sales process directly? Many B2B teams don't know this. Ask sales. There are almost always three to five pages that get shared in deal cycles that don't appear important in the analytics.

With those answers documented, the data analysis can begin. Without them, you're making decisions that look rational in a spreadsheet and turn out to be expensive in practice.

The decision tree.

For each page or cluster of pages, work through these questions in order. Stop at the first decisive answer.

Question 1: Is this page commercially critical or sales-supporting?

If yes: Keep. Improve if needed but do not consolidate or remove regardless of traffic. These pages are not a content problem; they're a conversion optimisation problem.

If no: Continue to Question 2.

Question 2: Does this page currently rank for any query with meaningful search volume?

If yes, and the ranking is relevant: Keep. A page that ranks is doing work. Improving it is almost always better than removing it. The question is whether the traffic it earns converts appropriately. If not, that's a post-click problem, not a content problem.

If yes, but the ranking is for irrelevant queries that don't reflect your current positioning: This is a signal to update or consolidate. A page ranking for terms you no longer want to be associated with can be redirected to a page that better reflects your current positioning, once a better candidate exists.

If no: Continue to Question 3.

Question 3: Is there a better, live page on the same topic?

If yes: Merge. Redirect the weaker page to the stronger one. Be specific with the redirect: the destination should be the page most likely to satisfy the same intent, not just the homepage.

If no, but there could be: Kill and redirect to the closest relevant page. Don't create a new consolidation target just to give the redirect somewhere to go. That defers the problem rather than solving it.

If no, and the topic is genuinely worth covering: Kill the thin version and plan a proper replacement. A thin page on a topic you want to own is often worse than no page, because it occupies the ranking position with content that doesn't deserve it and suppresses the chance to produce something that does.

Question 4: Does this page have any backlinks or external citations?

If yes, meaningful links: Keep or merge carefully. Backlinks are worth preserving. If you're consolidating, use a 301 redirect and the link equity transfers. But check the linking context: sometimes the page was linked because it covered a specific angle that the consolidation target doesn't cover. Make sure the destination page earns the link.

If no: Kill or consolidate without concern for link preservation. A page with no external signals and no ranking is a pure overhead cost.

Question 5: Is the content accurate, current, and non-embarrassing?

If no to any of these: Fix before keeping. An inaccurate page that ranks is worse than no page. It's creating a bad first impression for a percentage of buyers who will never come back. Update or kill.

/ What to do with the merge pile

Merging pages is not just redirecting URLs. The destination page needs to be genuinely better than the pages it absorbs: covering the topic more completely, with better structure and more specific claims. A redirect to a thin page is not a merge; it's a burial.

The AI search dimension.

Content audits have always been primarily about search performance. In 2026, that means accounting for AI search as well as traditional organic. A page that ranks in position 3 for a traditional query but never appears as a citation in ChatGPT or Perplexity responses for the same query is doing partial work. The question is whether the gap is a content quality issue or a source-worthiness issue, and the answer affects whether the right intervention is rewriting the page or building off-site presence.

When running a content audit today, I add a lightweight AI citation check to the process: for the top 20–30 most important pages by traffic or commercial value, test whether those pages' topics produce citations to the site in the main answer engines. If high-traffic pages are not being cited, that's a signal to look at the source-worthiness of the content rather than just its on-page quality.

The content portfolio decisions that follow from this (particularly the choice between updating an existing page vs. creating a new, more citable version) are also covered in the GEO roadmap I'd build if starting today.

NOTES
  1. The decision tree is sequential. Skipping questions (particularly Question 1) is the most common mistake in content consolidation projects and the source of most "we deleted something we shouldn't have" incidents.
  2. For sites over 10,000 URLs, segment the audit by page template type before applying the tree. Applying the same logic to blog posts and product pages simultaneously creates confusion. Work through each template class separately.

/ Frequently asked

How do I handle pages that are commercially important but ranking for the wrong queries?

Treat this as a positioning problem, not a content audit problem. The page should stay; the content should be updated to better match the intended buyer intent and search queries. Run this update as a separate programme from the consolidation audit. Conflating the two creates scope creep and unclear success metrics.

Should I consolidate content before or after a site redesign?

Before, where possible. Redesigns tend to reset the content inventory to "whatever exists now," which means you carry technical debt forward. Consolidating before a redesign gives you a cleaner starting architecture and means the new site structure can be designed around the pages you're actually keeping.

How long should a content consolidation project take?

For a site with 2,000–8,000 URLs: the audit and decision-making phase typically takes three to four weeks with one dedicated person. Implementation (redirects, content merges, new page production) takes longer and depends on development and content resource. Prioritise the top 10% of impact decisions first: those are usually 30–40 URL decisions that account for the majority of the performance opportunity.

tc
/ Written by

Thomas Cox

Twelve years in B2B SEO, most recently at VP level. Now independent — helping companies stay discoverable as buyer search moves into ChatGPT, Perplexity, and Google AI Overviews. Remote · UK.

About Thomas Work with me
Keep reading

More from the notebook.

June 2026 · 11 min

The B2B IA mistakes that quietly kill organic.

READ →
June 2026 · 14 min

The B2B GEO roadmap I'd build if I started today.

READ →
June 2026 · 9 min

Source-worthiness, explained for SEO teams.

READ →

Want this in your inbox?

Roughly one essay a month. No drip sequences, no upsells.

Subscribe