I pruned about 40% of the pages off an independent hotel site last year and traffic went up. I have also watched a well-meaning hotelier delete 60 pages in an afternoon and lose a third of their organic bookings overnight. Same activity, opposite outcome. The difference was entirely in the decision-making before anything got deleted.
So this is the post I wish existed when I started doing content pruning for small hotels. Not “delete your bad pages” as a slogan, but the actual prune-vs-merge-vs-keep framework I run, the part where you decide what touches the kill list and what gets left alone. If you do this carelessly you will absolutely tank your traffic. If you do it deliberately you make the whole site stronger.
Why thin content quietly drags down a hotel site
Here is the thing nobody tells independent operators: Google evaluates your site partly as a whole, not just page by page. A pile of weak, near-empty, or redundant pages doesn’t just sit there harmlessly. It dilutes how often Google bothers crawling your good pages, spreads your internal links across junk, and muddies the topical signal you’re trying to send about who you are and where you are.
Boutique and independent hotel sites accumulate this stuff fast. A CMS that spat out a separate page for every room type and every room number. Event recap pages from a 2021 wedding expo. Five different “packages” pages that 90% overlap. Tag and category archives nobody asked for. A blog post about a local festival that moved venues three years ago. None of it is malicious. It just piles up.
Thin content isn’t about word count. A 150-word page that perfectly answers “what time is checkout” is not thin. A 1,200-word page stuffed with filler that no human would ever search for is. The real test is whether the page satisfies a genuine intent better than nothing at all.
The reason this matters more in 2026 than it did five years ago: AI answer engines and LLM-driven search are now reading your whole site to decide what to say about you. AEO is a real and growing search behavior (around 27,100 US searches a month for the term itself), and a site cluttered with contradictory, redundant, half-finished pages gives those models a messier picture to summarize. Clean sites get summarized more cleanly. If you want the deeper version of that argument, I wrote it up in is your hotel invisible to ChatGPT.
Step one: inventory everything before you judge anything
You cannot prune what you cannot see. Before I make a single call, I build one spreadsheet with every indexable URL on the site and four columns of data next to each one:
- Organic clicks (last 12 months, from Search Console)
- Impressions (last 12 months, also Search Console)
- Backlinks to that exact URL (any decent backlink tool)
- Internal links pointing to it
Twelve months matters for hotels specifically because of seasonality. A page that looks dead in September might be your top earner every March for spring break, or every December for holiday packages. Judge a seasonal page on a three-month window and you’ll execute pages that were about to do their job. I’ve nearly done it. Use the full year.
Then I add two human-judgment columns I fill in by actually looking at the page: does this satisfy a real intent? and is there a better page on the site for this topic? Those two questions are the whole game.
The decision framework: keep, improve, merge, redirect, or delete
Every URL gets sorted into exactly one of five buckets. Here’s how I draw the lines.
| Bucket | When I pick it | What I do |
|---|---|---|
| Keep | Earns traffic, links, or bookings; satisfies a clear intent | Leave it alone, maybe refresh later |
| Improve | Right intent, weak execution; has potential | Rewrite and expand; do not delete |
| Merge | Overlaps heavily with a stronger page | Fold content in, 301 the weaker URL |
| Redirect | Obsolete but has equity or a close topical home | 301 to the most relevant live page |
| Delete | No traffic, no links, no intent, no home | Remove, return 410 |
The mistake I see most is treating this as a two-bucket problem: good pages and bad pages, keep or delete. That binary is exactly how people tank their traffic. Most of your weak pages don’t belong in the delete bucket at all. They belong in improve or merge, where their accumulated signals get preserved and concentrated rather than thrown away.
Keep
If a page pulls meaningful organic traffic, has earned a backlink, or directly drives bookings, it stays. Full stop. Don’t get cute. The point of pruning is to protect these pages by clearing the clutter around them, not to put them at risk because you got delete-happy.
Improve
This is the bucket that saves people from themselves. A room page with two sentences and a photo isn’t thin because it’s a room page, it’s thin because it’s underbuilt. Rooms, your restaurant, your spa, your neighborhood guide: these are the pages guests actually want and the pages that feed your book-direct CRO funnel. You don’t delete a weak version of a page you fundamentally need. You make it good.
Merge
Merging is the most underused move and usually the highest-value one. You’ve got “Romance Package,” “Couples Getaway,” and “Anniversary Special” pages that are 80% the same words. Pick the strongest URL, fold the unique bits from the others into it, make one genuinely excellent page, then 301 the losers into the winner. You go from three mediocre pages competing with each other to one strong page inheriting all their combined signals. This is how I cut page counts without losing traffic, because nothing is actually lost, it’s consolidated.
Redirect
Some pages are genuinely obsolete: the 2023 event that won’t recur, the discontinued package, the old “meet our chef” page for a chef who left. If that URL has any backlinks or had real historical traffic, it has equity worth keeping. Redirect it with a 301 to the closest relevant live page. A retired package page goes to your current offers page. An old neighborhood post goes to your updated area guide.
The cardinal sin of redirects is sending everything to the homepage because it’s easy. Google treats an irrelevant redirect as a soft 404 and the equity evaporates. A redirect is a promise that the destination is the natural next home for that content. Keep the promise or you’ve wasted the move.
Delete
Reserve deletion for pages with no traffic, no links, no intent, and nowhere sensible to redirect. Auto-generated tag archives. A blank draft that got published by accident. True orphan junk. When you delete, return a 410 Gone rather than a 301 to nowhere, so search engines understand the page is intentionally gone, not broken.
How to execute without tanking traffic
Sequencing is where careful people still get burned. A few rules I never break.
Don’t do it all at once. I prune in batches and watch each batch settle before the next. Nuke 200 pages in one shot and if something goes wrong you have no idea which move caused it. Batches give you a clean signal and an easy rollback.
Fix internal links before they 404. When you merge or redirect a page, hunt down every internal link pointing at the old URL and repoint it at the new one. Relying on the 301 to do that job works, but it’s sloppy and bleeds a little link equity each hop. Point links straight at the final destination.
Map redirects deliberately, one to one. Every redirected URL needs a specific, hand-chosen destination. No bulk “everything in /old-packages/ goes to /packages/” unless you’ve genuinely confirmed each one fits. This is tedious. It’s also the difference between a clean prune and a traffic crater.
Crawl the site after each batch. Run a crawler and confirm no broken internal links, no redirect chains (A to B to C), and no accidental redirect loops. This is exactly the kind of cleanup that protects the rankings you have for your own brand name, which I dig into in why your hotel ranks below OTAs for your name.
Track results at the site level, not just the pruned-page level. The whole point is that strengthening 30 pages lifts the other 270. If you only watch the URLs you deleted, you’ll panic when their traffic hits zero, which is the entire plan. Watch total organic sessions and total bookings instead.
A realistic example of how this plays out
Picture a 22-room boutique property with 180 indexed pages. Inventory reveals the shape of it: maybe 25 pages earn essentially all the organic traffic, around 40 are solid supporting pages, and the rest are some combination of redundant, obsolete, or auto-generated filler. That distribution is wildly common for independent hotels. The long tail of junk is usually the majority of the URL count and a rounding error of the traffic.
The framework sorts it. The 25 earners and 40 supporters: keep, with a few flagged to improve. A dozen overlapping package and room pages: merge down to four strong ones. Old events and discontinued offers with a backlink or two: redirect to current equivalents. Tag archives and orphaned drafts: delete with 410s. The page count drops by a third or more, but every signal that mattered got preserved or concentrated rather than discarded. (To be clear, that’s an illustrative pattern, not a promise about your numbers, every site is its own animal.)
The reason this connects to your bottom line: a cleaner, more authoritative site competes better for your own name and your “best boutique hotel in [your city]” terms, which is exactly the high-intent traffic the OTAs are siphoning off. I’m not going to pretend pruning lets you escape the OTAs, nobody can, and anyone who says otherwise is selling something. But a stronger organic and AI-search footprint is a real lever for reducing OTA dependence and winning back more direct bookings at a healthier mix. The commission math (those 15-25% cuts) is brutal enough that even a modest direct-booking shift pays for the work; I broke that down in the book-direct math post.
The short version
Pruning thin content is not deletion. It’s a sorting exercise where deletion is the last and rarest option. Inventory everything on a full-year window, sort each URL into keep, improve, merge, redirect, or delete, and bias hard toward improve and merge so signals get preserved instead of thrown away. Execute in batches, map redirects one to one, fix internal links, crawl after each pass, and measure at the site level. Do it that way and you concentrate quality instead of destroying it. Rush it and you’ll join the people who deleted their way into a traffic hole.
If you want a second set of eyes before you start swinging, this is exactly the kind of audit we run for independent hotels, finding what’s safe to cut, what’s secretly load-bearing, and what just needs to be made good. Tell me about your site over on the book a call page, or read how content and authority fit the bigger picture on our content and reputation service page. I would much rather help you sort the list than help you recover from a bad prune.