Within about 48 hours, the two big community platforms announced opposite moves. On October 27, Discord said that content from its public Discoverable Servers will start appearing on the open web and in search. On October 29, Reddit said it will become a primarily logged-in service. discord.com: Making Content from Discoverable Servers Easier to Discover
I’m sure Discord will look amazing on reports for the first year or so. You’ll get new citations, a new source in the charts and a nice line for the monthly deck. But Reddit is pulling back for a reason, and I’d keep that in mind before building a strategy on Discord.
This is useful if you already have a real community on Discord, especially if you run support there. If you don’t, you can probably ignore most of it.
Server pages go live on November 16, and messages follow in mid-December
Every Discoverable Server gets a public Server Web Page on November 16, 2026. If the admins haven’t set one up, Discord publishes a default page from the server’s profile. These pages can show two kinds of content, and each has its own settings.
Channel Summaries can start on November 16. Discord creates them with AI from public channels, and its own help page says they may not always be accurate. Admins can:
- curate them
- choose the source channels
- turn them off
Channel Messages roll out gradually from mid-December. Discord describes them as “answers to questions your members sent in public channels”, shown in the members’ own words. Only messages sent after the rollout begins are included, and for now they are text only. Public threads and forums count if the parent channel takes part.
Usernames and avatars aren’t shown, and these are excluded:
- private servers
- private channels
- DMs
- teen accounts
Both features are on by default for public channels, so a server that does nothing is in.
Each member has one account-wide setting. Sharing is on by default in most places, so messages can appear until the member turns it off. A European community may contribute far fewer messages than an American one of the same size, because sharing is off by default in:
- the EU/EEA
- the UK
- Canada
- Brazil
- Australia
- South Korea
A server needs 1,000 members and eight weeks of age to qualify
To qualify, a server has to:
- have at least 1,000 members
- be at least eight weeks old
- meet Discord’s activity and safety requirements
Discord says more than 63,000 servers currently qualify.
The age rule should help against rogue servers created just for this. A server created on November 1 reaches eight weeks around December 27, after the rollout starts. Spammers have other routes, which I cover in the spam section, but a throwaway server is harder to use.
Companies already running support on Discord have the clearest case
If you run SaaS, developer tools or a game and already support customers on Discord, you have real questions with real answers. Discord says it wants to surface that content, so the work is mostly organizing what you have.
I wouldn’t start a Discord server for SEO. You would need to build a real community of 1,000 people and keep it active, and the search benefit is still unproven.
What to do if you run a server
Do these before November 16:
- Check that your server is Discoverable, under Server Settings > Access > Discoverable. If it isn’t, none of this applies yet.
- Open Server Settings and select Web Presence under Community, then go through the channels. Summaries and messages have separate lists. Remove any public channel you wouldn’t want quoted out of context.
- Publish your own landing page instead of taking the default, and describe the server in specific terms people would search for.
- Tell your members what is changing and how to opt out. People posted in your server without expecting to end up on Google, so they should hear about it from you.
Do these before mid-December:
- Move support questions into forum channels or a dedicated help channel, and answer them there. Since only new messages count, I’d start this early. I expect clear question and answer threads to work best, though nobody can confirm that yet.
- Staff the channels you leave open. If a channel is included, someone should be moderating it for spam.
- Learn the removal process. Deleting a message removes it from Discord’s pages, but turning a feature off only stops new content, and already published content can stay up. Search engines keep their own copies for a while too.
After November 16, read your summaries regularly. They are AI generated and could misdescribe your product on a public page.
What to do if your brand is only discussed in other servers
Find the servers where your category is discussed, and read what people say. If you join, do it openly, as yourself, under your company’s name. You can’t edit what those summaries say about you, so your realistic options are fixing the real product issue or answering openly in the channel.
Discord is inviting spam
Until now, a message in a server reached the people in that server. From December it can also reach Google, which gives people a reason to spam. The server requirements apply to the server, so a spammer can skip them by posting in someone else’s eligible community.
Reddit has already been through this. Companies have been planting posts there to influence AI answers, with brand mentions dressed up as ordinary recommendations, and its moderators have had to fight it.
Discord may make this harder to spot from the outside. A reader on Reddit can open an account’s history and judge it. On a Server Web Page there is no name or avatar to check. Moderators can still see who posted, but they are often volunteers, and I doubt most signed up to police content for search engines.
Discord says messages must meet content requirements and its Trust & Safety team can remove summaries that break its guidelines. It hasn’t explained how it will detect planted questions and answers, so I’m treating that as unproven.
If someone offers you “Discord mentions” as a service, say no:
- Moderators can see the accounts.
- You can get banned.
- You may not be able to remove what was posted.
I’d also be suspicious of early Discord citations in any report, including your own.
Why this may not last
The deal with open community content has always been data in exchange for traffic. Reddit has a reason to want out of it. In its Q3 2026 shareholder letter, about 59% of daily active users weren’t logged in (76.8 million of 130.3 million), while logged-in and app users together account for over 80% of monetization. Reddit now says most unauthorized scraping happens through logged-out pages, and it expects AI platforms and search engines to keep paying for the data.
Reddit isn’t disappearing from search. The homepage stays open, and people arriving from a search result or shared link should still see at least the page they landed on. Tests start outside the US, and Reddit hasn’t decided what else to restrict.
My view is that the same pressure will reach Discord. Search engines and AI tools get the answers, and few people click through to join a server. Discord’s stated goal is reach for communities, so its incentives may differ from Reddit’s for now. This is my prediction, and I can’t prove it. In practice, build on what your community gets out of the work, and don’t let the plan depend on Discord keeping these pages open.
When you report on this, keep citations separate from outcomes. Track visits, server joins and signups as their own numbers, and don’t let a growing citation count stand in for them.
Server pages have known URLs, and community messages don’t yet
Server pages use discord.com/servers/<slug>-<server-id>, for example the Dead by Daylight page. Announcements already have message-level URLs in the format discord.com/announcements/<channel-id>/<message-id>. discord.com: the Dead by Daylight page
As of November 1, Discord’s robots.txt lists seven feeds and sitemap indexes. The server sitemap contains only plain server pages, and none of the sitemaps contain community message URLs yet. Discord says messages will appear on Server Web Pages and in search, but it hasn’t said whether they will live on the existing server URLs or get new ones.
Open questions
We don’t know how long any of this will take to show up, or whether it will matter. I’m monitoring Discord’s robots.txt, its sitemaps and a set of server pages every day, and I’ll update this post as answers come in.
- Does each answer get its own URL, or does it sit on the server page? That decides whether one answer can compete for a specific query, or one page has to cover a whole server.
- Does an answer appear with the question it replies to? Chat is interleaved, so someone or something has to pair them, and an answer without its question means little.
- Can admins influence which messages get picked, for example by marking a forum thread as solved, or is it entirely up to Discord?
- What structured data can these pages carry with no author name, and will they qualify for Google’s forum features?
- What happens when an answer goes out of date? For a SaaS company, last year’s support answer can be wrong about today’s product, and so far deleting it is the only documented fix.
- Who can correct a wrong AI summary when the server’s admins don’t care about your brand?
- Will Discord allow AI crawlers on these pages, or block them later and license the data the way Reddit does? Its announcement doesn’t mention them, and its
robots.txtnames no AI bots today. - Do opt-in defaults skew the content toward some countries, and what does that do to answers in languages like Hebrew or German?
- Does a server risk its Discoverable status when planted answers appear in it, or does the penalty only reach the accounts that posted them?
- Will admins get data on web visits and joins from these pages, or will we be measuring this from the outside?
I’ll publish a first update after November 16 and a data piece in January, once there’s enough to measure.
