Discussion starter: a publisher and community SEO question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.
- News SEO
Article comments: have they helped rankings through fresh long-tail text, or hurt through spam and thin content?
Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.
1 reply
-
adminAdmin
Both effects are possible, and which one you get is decided mostly by moderation and page type rather than by comment volume. Google has described comments as part of a page’s primary content. Good comments can add detail and wording that match real searches; spam links and filler become part of what the page offers too. What comments do not provide is freshness in the sense people hope for: a new “thanks, great post” changes the HTML, but it does not make a three-year-old article a better answer for a query where searchers want recent information.
The plausible upside sits on pages where readers add information the article lacks: substitutions and fixes on recipes, error messages and workarounds on troubleshooting guides, local detail on travel pages. The downside sits where moderation is weak, and on news and opinion pieces, where comments lean towards argument and the article’s search life is short. Public evidence on direct ranking effects is thin in both directions, so the better question for your site is how much of your search traffic comes from queries that only the comment text matches. That you can measure.
What Google has said about comments
In a December 2019 hangout, John Mueller said Google tries to recognise blog comments but treats them as part of the primary content, and that a post with little content but many good comments can still be useful, so comments should not be ignored by default. It is an older statement, so read it as a description of the principle rather than of current weighting. It explains why both camps in this debate have a point: if comments are part of the page, they can add to it or subtract from it.
The rest of the documentation is about spam. Google’s spam policies list comment spam on blogs as an example of user-generated spam. Its guidance on preventing user-generated spam notes that manual approval can cut spam considerably, and covers CAPTCHAs, IP ban lists and anti-spam plugins such as Akismet. It also suggests considering
nofolloworugcon links in untrusted content, andnoindexon posts from new users who have no reputation yet. Left unmoderated, comment spam can also lead to a user-generated spam manual action in Search Console, which is a much bigger problem than any ranking nuance.First, check whether Google sees your comments at all
A lot of the disagreement about comments comes from sites that deliver them differently. Before debating effects, check what Google actually receives.
How comments are delivered What Google is likely to get How to check Server-rendered in the article HTML Comments as part of the page content URL Inspection, “View crawled page”: search the HTML for a recent comment Injected by JavaScript on page load Included if rendering completes and the script runs without user interaction URL Inspection live test: rendered HTML Loaded on scroll May or may not be included; it varies with how loading is triggered Live test rendered HTML; do not assume Loaded on click (“Show comments”) Not included, because Googlebot does not click buttons Rendered HTML will not contain them Third-party iframe widget Comments live on the provider’s URL and may not be associated with your page Rendered HTML of your page; a quoted-phrase search Paginated onto separate URLs Extra thin URLs that mostly repeat the article Crawl stats, server logs, site:searchesVery long server-rendered threads have one more quirk. Google explained in March 2026 that Googlebot fetches up to 2MB of an HTML file and ignores the bytes beyond it. The article normally sits above the comments, so it is safe, but anything placed after a huge thread (late JSON-LD, related links, footer navigation) could fall outside what Google fetches. Only very large threads approach that size, so this is a check for your biggest pages, not a general concern.
Where comments help, and where they hurt
Page type What comments can add Main risk What I would do Recipes, DIY and how-to Substitutions, fixes, equipment variations, first-hand results Low if moderated Keep indexable; curate the best tips into the article Technical troubleshooting and product support Error messages, model numbers, workarounds, confirmation that a fix still works Outdated fixes and link drops Keep; fold confirmed fixes into the article with dates News and opinion Argument and off-topic replies High moderation cost, low search value, brand risk Close after a set period, or load on click Evergreen guides with years of “thanks!” and link drops Very little Spam and noise in the main content Remove spam, close old threads, qualify links Health, finance and legal topics Personal advice from strangers Unvetted advice sitting in the page’s content Strict pre-moderation, or no public comments The last row deserves emphasis. If comments count as page content, a reader telling others to stop a medication, or to put their savings into a speculative scheme, sits inside a page on a topic where Google’s helpful content guidance says its systems give even more weight to content with strong E-E-A-T. That is a reason to moderate harder on those topics regardless of any ranking argument.
On the “fresh text” theory specifically: new comments change the page, and a page that changes may be recrawled more often, but recrawling is not ranking. Freshness matters for queries where searchers expect recent information, and a new comment does not make the article itself more current. I would also stop comments from changing the visible “updated” date or the sitemap
lastmod. Those should describe changes to the article, and alastmodthat moves with every comment is one that search engines have little reason to trust.Measure how much traffic comment text brings
This is measurable with Search Console data and a crawler, and it is a better basis for the decision than anyone’s anecdote.
- Pick articles with a meaningful volume of comments, and pull page and query data for them from the Search Console API over the longest window available.
- Crawl those pages and split each into two text blocks, the article (title, headings and body) and the comment section, using the template’s CSS selectors.
- Flag a query as comment-only when its meaningful words appear in the comments but not in the article.
- Sum clicks and impressions for comment-only queries, per page and across the set.
Treat the result as an estimate, because Google can rank a page for a query through synonyms and links as well as on-page words. Three patterns are useful. A small share of clicks from comment-only queries means comments are doing little for search, and the decision should rest on community value and moderation cost. A meaningful share with healthy click-through means readers are adding information people search for, and the best of it probably belongs in the article itself. Many impressions with very low click-through means comment text is matching queries the page does not answer, which is noise rather than value.
Hypothetical example:
A baking site finds that its 300 most-commented recipes get a noticeable slice of clicks from comment-only queries such as “can I use oat flour in [recipe]” and “why did my [recipe] sink”. That is worth keeping, but the stronger move is to answer the recurring questions in a short, tested section of the recipe itself. The value then sits in the editorial content, and it no longer relies on whether a comment widget renders or stays spam-free.
A controlled test, if the stakes justify one
If comments carry a large moderation cost or a large share of page weight, a randomised test is possible without taking comments away from readers.
- Treatment: switch a random half of 200 or so comment-heavy evergreen articles to click-to-load comments, so readers can still open them but they leave the HTML Google receives. The other half stays unchanged. Readers and Googlebot get the same page, so this is not cloaking.
- Primary metric: clicks and impressions on the comment-only queries identified above, then total clicks, each compared against the control group and the pre-period.
- Secondary metrics: comment rate (click-to-load is likely to reduce it, which is a real cost), return visits, newsletter sign-ups and page speed.
- Timing: the change only counts once Google has recrawled and reprocessed the treated pages. Watch last-crawl dates in server logs, start the clock when most treated pages have been refetched, and run long enough for sparse long-tail queries to accumulate at group level. Low-traffic pages need longer windows than popular ones.
- Confounders: seasonal content such as holiday recipes, core updates, and any template change shipped at the same time.
If the treated group loses comment-only clicks beyond normal variation, comments carry search value and moderation is worth funding. If nothing moves, decide on community grounds. If the treated group improves, the comments were diluting those pages, and moving more of them behind a click, or pruning, is justified.
Housekeeping that removes most of the downside
- Check the rendered HTML to confirm commenter links carry
rel="ugc"orrel="nofollow". Do not assume the theme or plugin adds them. - Hold first comments from new commenters for approval, and run an anti-spam filter.
- Close comments on older posts where they now attract only spam.
- Look for crawlable comment URLs in the Crawl stats report and server logs: comment pagination, reply links with parameters (on WordPress,
?replytocom=is the one to check) and per-post comment feeds. - Measure the comment widget’s effect on Core Web Vitals in field data, since third-party widgets can add heavy scripts and layout shifts.
The position I would take
I would keep comments where readers add information, moderate them as if they were part of the article (because Google has described them that way), and pull the best of them into the article body so the value does not rest on a widget. On news, opinion and sensitive topics I would default to closing comments or loading them on click. And I would make the call on the share of clicks from comment-only queries, not on the hope that fresh text lifts rankings, because that share shows what comments are actually doing for your search traffic.
Need help with this on your own site?
Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.
Add a reply Cancel reply
Replies are open to approved members. Log in or apply to join.