Consolidating overlapping event pages when two shows chase the same term

The diagnosis is done and the pair is real: two exhibitor list pages, one on each of two sibling show sites, competing on the same category search. Somebody in the meeting suggests deleting the weaker one, because it gets almost no traffic and nobody has updated it since 2024.

Consolidating overlapping event pages is a different job from deleting one of them, and the difference is worth about a year of accumulated links. A deleted URL takes its referring domains with it. A merged one hands them over.

Why a merge beats a deletion

Two things survive a page beyond its content. External links pointing at it, and whatever history the search engine has accumulated about it. Deleting the URL discards both and returns a 404 to anyone following a link from an old newsletter or a partner site.

Google's documentation on consolidating duplicate URLs, current in 2026, lists redirects at the top of the methods that influence which URL gets indexed, describing a redirect as "A strong signal that the target of the redirect should become canonical". That is the strongest instrument available, above the canonical tag and well above a sitemap entry, which the same page calls "A weak signal".

Joshua Hardwick's Ahrefs guide, updated in January 2024, reaches the same conclusion from the other direction and is unusually direct about the alternative, calling noindex "a terrible way to fix cannibalization" because it removes the page from indexing without transferring anything. His recommendation for pages fulfilling the same intent is consolidation with a redirect.

So the sequence is fixed by the tooling. Merge the content, then redirect. Doing it the other way round loses the material.

Picking the winner without a committee

Three criteria, applied in order, and written into the ticket so the losing brand can read the reasoning.

• Referring domains. Links from other sites are the one asset you cannot rebuild by writing something better. Count distinct linking domains and ignore the raw link total, because thirty links from one directory represent one relationship.
• Twelve month impressions. Use the full year, because event pages are seasonal and a three month window will hand the decision to whichever show ran most recently.
• Editorial fit. Where the first two disagree or come out close, keep the page on the brand whose audience the content actually serves.

Seniority of the brand, size of the show and who shouted loudest in the meeting are not on the list. They will be raised anyway, which is the reason for writing the three criteria down before you pull the data instead of after.

Working the merge on two exhibitor list pages

Take the pair. Page A on the materials brand runs at 400 impressions and 12 clicks a month, a click through rate of 3.0 per cent, with 31 referring domains. Page B on the logistics brand runs at 90 impressions and 3 clicks, a rate of 3.3 per cent, with 6 referring domains.

A wins on both of the first two criteria and the decision takes a minute. The interesting arithmetic is what to expect afterwards.

The naive forecast adds them: 490 impressions. That figure will be wrong and it is worth knowing why before you promise it to anybody. Pull the query breakdown for both pages and count how much of B's demand A was already serving. Suppose 58 of B's 90 monthly impressions come from queries where A also appeared in results. Only 32 impressions represent demand A was missing entirely.

So the honest forecast has two parts. Distinct demand of 400 plus 32, which is 432 impressions. Plus whatever improvement in position the merged page earns on the 58 shared queries, where it now has 35 distinct referring domains behind it instead of 31, because 31 and 6 overlap on two domains that linked to both pages.

That second part is the one worth having and the one nobody can size in advance. Write the forecast as 432 impressions of distinct demand plus an unquantified position effect, and you will be right. Write 490 and you will be explaining yourself in three months.

What actually has to move

The redirect is the last step and the smallest one.

Read B and find what it has that A does not. On an exhibitor list pair that is usually a hall by hall breakdown, a couple of paragraphs of category framing written for a different audience, and a set of exhibitor entries that never made it onto the other brand's list. Move all of it into A, in A's structure, with A's headings. A merge that keeps only the winner's content is a deletion with extra steps.

Then find every internal link pointing at B. On a portfolio site that is usually more than anybody expects, because the page sits in a navigation block, in a related links module and in the body of a dozen articles. Suppose there are 47 of them. Update all 47 to point at A directly.

Leaving the 301 to catch them works and is worse for two reasons. Every one of those links now costs a round trip, and in two years somebody will change the redirect and quietly break 47 links whose destination nobody remembers. A redirect is insurance for external links you cannot edit. Internal links you can edit should point at the truth.

Check what the redirect actually returns before you close the ticket. Content systems default to a 302 more often than anybody expects, and a temporary redirect asks the search engine to keep the old URL indexed, which is close to the opposite of what you wanted. Check the chain too: if B already redirected from an older URL, you now have two hops, and the tidy version is to repoint the original at A so every historical URL reaches the survivor in one step.

Keep the redirect permanently. Removing it after six months because the traffic settled undoes the whole exercise.

How long should you wait before judging the merge?

Longer than the person who approved it wants to wait.

Give it two months minimum before reading anything into the numbers, and expect a dip in the first few weeks while the redirect is processed and the merged page is recrawled. A drop in week two is the normal shape of this work and it panics people who were promised an improvement.

Measure at the query level. Page level reporting will show A gaining and B disappearing, which tells you the redirect worked and nothing about whether the group is better off. The comparison that matters is summed clicks for the contested term across both properties, before and after, which is the same measurement that established the pair was costing you traffic in the first place.

Record the date of the merge in the same place you record the numbers. Six months later somebody will ask why the materials brand's traffic stepped up in April, and the answer should take ten seconds to find.

When should neither page win?

Two cases, and both are common enough to check for before you start.

The first is where both pages carry meaningful referring domains and serve genuinely different readers. Merging then trades one brand's audience for a marginal search gain, and the better move is to differentiate the two pages harder: change the angle, change the title, and let them rank for adjacent things. Where the pages are close copies of one another instead of independent takes, a canonical is the lighter instrument and a merge is more surgery than the situation needs.

The second is the past edition page. An exhibitor list from the 2024 edition overlaps heavily with the 2026 one and should usually be kept, dated and left alone, because it answers a question the current page cannot. Folding old editions into the live page is an archiving decision with its own rules and it goes wrong in a way that costs you long tail traffic nobody was watching.

Where this stops

A merge across two domains is permanent in a way that a merge inside one site is not. The redirect hands a URL on one brand's domain to another brand's, and unwinding it later means a second redirect and a second dip. Where the two brands might be sold separately, and in an exhibitions group that is a live possibility, the merge creates a dependency that a future buyer inherits. Check with whoever handles portfolio strategy before you redirect across a domain boundary.

The evidence is also weaker than the process looks. You get one observation, no control, and a search environment that changed underneath you during the two months you were waiting. What you can say afterwards is that summed clicks for the term went up or down. What you cannot say is that the merge caused it.

There is one failure mode with no clean fix. If the losing page's referring domains were pointing at its specific content, a hall by hall breakdown, say, and the merged page reorganises that content away, the links now point at something that no longer answers what they linked to. The page keeps the signal and loses the relevance, and neither number in your report will show it.

Pull referring domains and twelve months of impressions for the single pair you are most confident about, put both rows in a document with the three criteria across the top, and circulate it to both brand owners before anybody touches a redirect. The argument is about ownership and it is cheaper to have it in writing, which is also the point at which whoever runs the content operation finds out how many other pairs are waiting.