Discussion starter: an indexing 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.
- Content SEO
- Indexing
Unanswered forum threads: noindex until they get a reply, or index immediately to catch long-tail demand?
Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.
1 reply
-
adminAdmin
For most forums I would not use “has a reply” as the indexing switch. I would use a quality gate, and a reply is only one way through it. A detailed question with an exact error message, model number or clear context can deserve indexing on day one. A one-line question from a new account usually does not, and a “same problem here” or “bump” reply should not flip it to indexable either.
The real trade-off is not long-tail traffic versus a tidy index. It is page-level upside (a few unanswered threads that genuinely match searches) against site-level risk (a large, permanent pool of dead-end pages that Google crawls, often declines to index, and may count against the site). Which side wins depends mostly on your answer rate, how long questions wait for a first answer, and how much spam sits in the unanswered pool.
Why “unanswered” is the wrong unit of decision
An unanswered thread can be three very different pages:
- A specific, well-described problem. Someone pastes an error string or describes an edge case nobody else has written about. A searcher with the same problem lands there, learns they are not alone, and may subscribe or answer later. That page has some value even empty.
- A vague or duplicate question. “Help, my app won’t work” or a question that was answered in another thread last month. The searcher lands, finds nothing, and leaves. That is the page type people worry about, and they are right to.
- Spam or low-effort posts. Spam rarely attracts replies, so the unanswered pool usually holds a disproportionate share of it. This is an underrated reason why “noindex until answered” works reasonably well as a crude rule: it doubles as spam containment.
There is also a fourth category that zero-reply rules catch by mistake: threads that are not questions at all. A build log, a guide or an announcement with no replies is not “unanswered”. If your forum software treats reply count as the trigger, check that these content types are exempt.
What Google documents, and what it leaves to you
Google’s QAPage structured data documentation says a question “must have at least one answer” to be eligible for the Q&A rich result. That is a rich result eligibility rule, not an indexing rule, but it shows how Google thinks about the page type: the answer is the value.
Beyond that, Google does not publish a threshold for how much thin user-generated content a site can carry. Google has said its core ranking systems work mainly at the page level but also use some site-wide signals. That leaves a real but unquantified risk: nobody can tell you that 20% unanswered threads is safe and 40% is not. Anyone who gives you a number is guessing.
Context matters too. Forum visibility has been volatile. Aleyda Solis’s analysis of the May 2026 core update found forums and Q&A sites lost UK visibility (Reddit about 24% and Quora about 31% in that dataset). That is third-party data, not a Google statement, but it is a reason not to assume that any forum page will be rewarded simply for existing.
Important distinction:
Leaving a thread indexable is not the same as getting it indexed. Many thin threads end up as “Crawled - currently not indexed” or “Discovered - currently not indexed” regardless of your directive. If that is already happening at scale, “index immediately” is partly a choice you do not get to make.
The mechanics that decide how costly each option is
Noindex until answered sounds reversible, but the reversal is not instant. Google has to recrawl the thread to see that the noindex has gone, and practitioners commonly observe that URLs which have carried noindex for a while get crawled less often. John Mueller also said back in 2017 that a long-term noindex page is eventually treated as if its links are nofollowed. So a thread answered three weeks after posting may wait a further stretch before it is eligible, and that lag is the real cost of the noindex approach.
You can shorten the lag:
- Add the thread to your XML sitemap only when it becomes indexable, with an accurate
lastmod. Google’s sitemap guidance says it useslastmodwhen the value is consistently accurate, so do not let the forum bump dates on every view or edit. - Link newly answered threads from a crawlable “recently answered” or “solved” list, not just “latest activity”.
- Purge the CDN or page cache when the first qualifying answer is posted, so Googlebot does not keep receiving the old noindex HTML.
Watch for this:
Put the robots meta tag in the server-rendered HTML and change it on the server. Google’s JavaScript SEO documentation notes that a noindex in the original HTML can stop rendering, so removing it with JavaScript once replies load may not work. Also avoid robots.txt for this: a blocked thread can never show Google that its noindex has been lifted.
The index-immediately side has its own mechanics. On a small forum, a few thousand extra thin URLs cost little crawl. On a large one, Google’s crawl budget guide applies once you are in the millions of pages or see many URLs in “Discovered - currently not indexed”. There, a permanent tail of empty threads competes with answered ones for crawl attention.
A gate that works better than reply count
New thread posted↓New account, link-heavy, or flagged by spam checks? If yes: noindex, hold for moderation↓Question substantive on its own (specific detail, error text, clear context, not a duplicate)? If yes: index now↓Otherwise noindex, and wait for a qualifying answer (not from the original poster, above a minimum length, not flagged)↓Qualifying answer arrives: remove noindex server-side, add to sitemap with accurate lastmod, link from “recently answered”↓Still unanswered after 60 to 90 days with no organic entrances? Merge into a canonical thread, keep noindexed, or removeThe timeout matters. Without it, the noindexed pool grows forever and you never make a decision about it. With it, moderators periodically clear out duplicates and abandoned questions, which also improves the forum for members.
Which way to lean for your forum
Condition Lean towards indexing immediately Lean towards noindex until answered Answer rate High (most threads get a real answer) Low, leaving a large permanent unanswered pool Time to first answer Hours, so Google rarely sees the empty version anyway Days or weeks, so the empty version is what gets crawled Question quality Technical, specific, often with error strings or model numbers Short, vague, frequently duplicated Spam exposure Strong moderation, established members Open registration, spam lands in unanswered threads Index health Most answered threads are indexed and earning clicks Many threads already “Crawled - currently not indexed” Scale Thousands of threads Hundreds of thousands or more Hypothetical example:
A software support forum has 300,000 threads. In one category (installation errors), 85% of questions get answered, the median wait is four hours, and most posts include error text. In another (general chat and “which tool should I buy”), 40% are answered, the median wait is five days and duplicates are common. The sensible policy differs by category: index installation questions immediately, and gate the second category behind a qualifying answer with a 60-day cleanup.
How to test it without fooling yourself
Start with evidence you already have. Segment organic landing pages by thread state (answered, unanswered) and look at what happens after the visit: registrations, follows, clicks into related threads, or an immediate exit. Also check how many unanswered threads received an answer from someone who arrived via organic search. If search is bringing in answerers, that is a real argument for earlier indexing.
Then run a URL-level split, ideally by a hash of the thread ID so both groups contain the same mix of categories:
- Hypothesis: indexing substantive unanswered threads immediately earns more organic entrances per thread over 90 days without reducing answer quality.
- Test group: passes the content gate, indexed on posting. Control: same gate, noindexed until a qualifying answer.
- Primary metric: organic clicks per thread over 90 days from posting. Secondary: time from answer to first impression, share of threads answered, share indexed, engaged sessions from organic entrances.
- Confounders: core updates during the window, seasonal support spikes (product releases), and moderation changes.
Be clear about the limit: a URL split measures page-level effects. If a large pool of thin pages affects the whole site, that effect lands on both groups equally and the test cannot see it. To assess site-level risk you need a before-and-after comparison on the whole site, alongside a check that answered-thread performance did not fall once more unanswered threads were indexed.
What would change the decision: if unanswered test threads earn meaningful entrances that lead to engagement or answers, widen the gate. If they earn impressions but almost no engaged visits, or mostly land in “Crawled - currently not indexed”, tighten it.
A practical position
If this were my forum, I would index substantive questions from trusted members immediately, noindex everything else until a qualifying answer arrives, flip the directive server-side with a sitemap update when it does, and run a 60 to 90 day cleanup on anything still empty. Then I would let the category-level data on answer rates and organic engagement decide whether to loosen or tighten the gate, rather than debating one rule for the whole forum.
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.