Thin content: how to decide what to improve, merge or remove
Translated from the Spanish original. Leer en español
A two-hundred-word text can answer a question. A two-thousand-word text can leave it unanswered. Thin content is not fixed by making the word counter bigger.
It is fixed by understanding why a page exists, what the person visiting it needs and what it offers compared with the alternatives. Then you decide whether it needs an improvement, a merger with another piece of content or removal. In that order, rather than starting with the bin open.
What is thin content, and why is it not measured in words?
Thin content is a common SEO expression for content that offers little value to the person using it. The word "thin" suggests something scarce. The dangerous leap is to turn "scarce" into "fewer than a certain number of words".
Google does not have a preferred word count. Its guide to creating helpful, people-first content asks you to assess whether a page helps readers achieve their goal and offers information, analysis or original work. It also asks whether the content leaves someone feeling they need to search again for a better answer.
That is a much more useful check than counting paragraphs. An opening-hours page can do its job with very little text. A guide that promises to explain a procedure but leaves out the requirements, steps or where to complete it can fall short even if it fills several screens.
Nor should "thin content" automatically be treated as a diagnosis of a penalty. Google's spam policies describe specific practices, such as producing large numbers of pages to manipulate rankings without helping users, or publishing affiliate content without added value. A page that could be improved does not, on its own, prove the site is engaging in those practices.
How can you distinguish a short page from one that offers little value?
The difference lies in the task. Before judging the amount of content, you should be able to complete this sentence: "This page exists so that someone can...".
Checking opening hours, comparing two services and learning to configure a tool are different tasks. They do not need the same length or the same evidence.
Here is a fictional example. A workshop's page explains which bicycles it repairs, where it is, when it opens and how to book an appointment. It can be short and serve the visit well. Another page from the same workshop spends a thousand words explaining the importance of looking after a bicycle but never says whether the workshop repairs hydraulic brakes. The second has more text and leaves a business question unanswered.
For an initial review, I would check these points:
- The page delivers on the promise in its title.
- The information lets someone make a decision or complete a task.
- The details are specific to the service, product or question, rather than interchangeable across twenty pages.
- Important facts have a source or a verifiable explanation.
- The page offers more than a condensed version of other pages.
This is not an official Google test or a scoring formula. These are editorial review questions. If a page fails, write down exactly what is missing. "It needs more content" does not explain what needs fixing.
What data should you review before changing a URL?
Reading the page is essential. Looking at what is happening with it helps you avoid making a blind decision.
I would start with its role within the website. It might attract visits from Google, explain a commercial condition, help a customer who is already comparing options or provide support after a purchase. Low organic traffic does not invalidate all those roles.
Next, I would check Search Console. Its Performance report lets you review clicks, impressions, queries and pages, as well as compare periods. Filter for the URL and look at the searches it appears for, rather than stopping at the site-wide total.
If you need to organise that first review, my Search Console guide explains what to check and in what order.
A URL with zero clicks might not be indexed, might have little demand, might be new or might answer a seasonal search. It might also be unhelpful. The number alone cannot tell you which explanation is right.
I would also check whether it receives links, visits from other channels or enquiries. I would review which internal pages use it and whether another URL meets the same need. When comparing periods, I would choose windows that make sense for the business and its seasonality, without setting a universal time frame.
Before changing anything, I would leave a simple note: the URL's role, the problem observed, the data reviewed and the proposed change. It helps you remember why the decision was made when it is time to check it.
When is it worth improving the content?
When the page has a clear role but its answer is incomplete, confusing or hard to use.
Improvement can mean adding missing information. It can also mean removing an introduction that delays the answer, organising the sections, replacing a generic explanation with verifiable details or updating a condition that is no longer correct.
In the fictional workshop example, I would not add a history of the bicycle. I would clarify which repairs it offers, what it needs to know to provide a quote and how to book an appointment. If there are limitations, I would explain them. If a price depends on inspecting the fault, I would not invent a rate to fill a table.
Google's helpful-content guide asks whether there is original information, analysis and value compared with other pages. That contribution might be a well-made comparison, a documented test or an explanation that lets someone act. It does not require decorating every article with an experience nobody has had.
My approach would be to fix first what prevents the task from being completed. The resulting length is a consequence of the improvement, not the objective of the assignment.
When should you merge two pages, and when should you remove one?
Merging makes sense when two pages essentially meet the same need and keeping them separate offers no useful distinction. Before deciding, I would compare their queries, their content and the role of each. Two similar titles are not enough to justify a merger.
If they are merged, I would keep the valuable material from both in a destination that meets the original need. Then I would review the internal links and the move from the URL that is no longer used.
Google allows redirecting consolidated content to its new shared page, but advises against sending many URLs to an irrelevant destination such as the home page. Not everything removed needs to end up on the home page as though it were a lost-property drawer.
A canonical does not move the visitor either. As I explain in what a canonical is, it identifies a preferred version among duplicate or very similar pages. It is not a substitute for every decision to improve, merge or remove content.
Removal would be the last option, not the first. Google's documentation on changes following a core update recommends considering deletion as a last resort when content cannot be recovered in a useful form.
If a page no longer has a role or an equivalent replacement, removal may be appropriate. In that case, it must return the right status, rather than an empty page with a successful response. My article on when to fix a 404 error and when to leave it alone explains that distinction.
A table to guide the decision
| Verified situation | Decision I would consider | What to check afterwards |
|---|---|---|
| The page has a role but lacks necessary information | Improve the answer and its presentation | The task can be completed and the facts are correct |
| Two pages cover the same need without a useful distinction | Consider merging them into a complete destination | Content retained, relevant redirect and internal links |
| The content is no longer useful and has no equivalent replacement | Consider removal | URL status and no remaining internal links to it |
| The page is short but does its job well | Keep it | It remains accurate, accessible and easy to find |
If you cannot see all the columns, scroll the table sideways.
The table guides the review. It does not authorise bulk changes or replace an analysis of each URL.
How can you check whether the decision worked?
First, I would check the implementation. Is the full content visible on mobile? Do the links lead to the expected destination? Is the page still accessible and indexable when it should be? Does a removed or redirected URL respond as planned?
Next, I would compare the result with the objective recorded earlier. If a condition for buying was missing, I would check that it is now clear. If two pages were merged, I would check that no useful information was lost. If the objective was better customer acquisition, I would track relevant queries and enquiries, rather than only the total number of visits.
Google warns that changes can take time to have an effect and do not guarantee improved rankings. Publishing a correction and looking at the graph the next day does not demonstrate the result.
My closing checklist would be:
- The URL has a role I can explain.
- The content meets that role with verified information.
- I have reviewed other pages before deciding on a merger.
- Links and URL statuses match the decision.
- I have saved a previous reference point and a date for reviewing the data.
When there are many URLs and interdependent decisions, the work needs priorities. That diagnosis and prioritisation are part of an SEO audit. It is not a competition to delete more pages than anyone else.
Frequently asked questions
Does low traffic mean thin content?
No. Traffic describes a result, not how useful a page is. Before judging it, review its role, demand, indexing and the queries it appears for.
Does Google require a minimum length?
Its helpful-content guide does not set a preferred word count. The question is whether the answer lets someone complete their task, not whether it exceeds a number.
Does using AI make a text thin content?
Not simply because AI was used. You need to review the result, its accuracy and the value it offers. As I explain in Google and AI-created content, automation does not remove editorial responsibility.
Should you delete everything that does not rank?
No. It might need improvement, a technical review or simply serve another purpose. Deleting without understanding the reason can remove a page that was helping users.
To finish our coffee
Thin content is not an instruction to write more or delete in a hurry. It is an invitation to justify each page: what it solves, what it contributes and why it deserves to exist separately.
The right decision is not the one that leaves fewer warnings in a report. It is the one that leaves a more useful website, with each URL doing a clear job.
