# Thoughts around simulating topics?

**URL:** https://meta.discourse.org/t/thoughts-around-simulating-topics/377632
**Category:** Community Building
**Created:** [August 6, 2025, 10:07pm UTC](https://meta.discourse.org/t/thoughts-around-simulating-topics/377632 "2025-08-06T22:07:30Z")
**Posts on this page:** 14
**Page:** 1

<div class="post-metadata">

### Author: ![noahl](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/noahl/32/552343_2.png) [@noahl](https://meta.discourse.org/u/noahl)
#### Post date: [August 6, 2025, 10:07pm UTC](https://meta.discourse.org/t/thoughts-around-simulating-topics/377632/1 "2025-08-06T22:07:30Z")

</div>

Was approached by our Growth team _(SEO and GEO (visibility in generative content - ChatGPT, Perplexity etc.))_ who see the value, and already see forum content being surfaced in GEO conversations.

They have a wide range of content that they’re focusing on at the moment, and proposed ‘simulating’ topics within our forum, but in a way that externally, there at appears no different to any other conversation.

For example, creating 5/6 accounts, and posting questions from these accounts, where we can then reply/encourage the community to reply from the normal ‘regular faces’ within the community.

To me, at its core, this is no different to organic conversations and answers - except we are planting these conversations for the purpose of better SEO/GEO.

Is this something others are doing? Is there anything I should think about before we go forward with this effort?

---

<div class="post-metadata">

### Author: ![Canapin](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/canapin/32/119591_2.png) [@Canapin](https://meta.discourse.org/u/Canapin)
#### Post date: [August 6, 2025, 10:40pm UTC](https://meta.discourse.org/t/thoughts-around-simulating-topics/377632/2 "2025-08-06T22:40:54Z")

</div>

I remember a few topics/posts about that, most being pre-AI-era:

> [@Ways to easily seed content as posted from various users (startup)](https://meta.discourse.org/t/ways-to-easily-seed-content-as-posted-from-various-users-startup/78608):
>
> Hey Guys, First post on here.. I’ve been using Discourse in one or two other places now and i’m just about to deploy it for my own site. Among the various ways that I will see the various categories with content is to post it myself. For this I plan on creating a bunch of ghost accounts to help get the ball rolling. I read that’s what they did at Reddit and seems to be a common approach. That said, it still seems controversial. What is the general attitude towards that here? Is it generally …

> [@Seeding a new forum by creating users and content](https://meta.discourse.org/t/seeding-a-new-forum-by-creating-users-and-content/49345):
>
> Hi there Discourse looks amazing. Wow! A big change from my old days using phpbb. In the past, admittedly a long time ago now, I’ve had greater success starting new forums by initially seeding some content. A little social proof does wonders for engagement levels! Would love to know whether it is possible to seed a new discourse forum. What I’d like to do is to create say 200 user accounts and roughly 1000 topics and replies before launching. Is it possible to do this without having either 2…

> [@Publishing posts from fake users](https://meta.discourse.org/t/publishing-posts-from-fake-users/314547):
>
> Hello everyone, please tell me. After creating forums, as a rule, at first there are no users, so the creators often resort to writing posts and replies to them themselves. Is there a plugin that helps organize this? For example, publish posts from different users and so on. I hope you brought up what I mean.

> [@AI sockpuppet accounts to jumpstart my community?](https://meta.discourse.org/t/ai-sockpuppet-accounts-to-jumpstart-my-community/292396):
>
> I am considering to add a discourse instance to a blog with a certain audience. However, since at the beginning it will be a mosly empty space with little ammount of activity in threads and posts, it would be handy to have a bunch of ai powered user accounts that generate some activity in order for people to stay engaged in the community. Does such thing exist on discourse?

I guess the main thing in your suggestion is that no bots will answer; they will just start topics, right?

I don’t know from a business perspective. Maybe it’s beneficial in the long run. _Perhaps_ it can help a community grow.

My guts tell me that I don’t like the idea of answering a bot without knowing it’s a bot, but I don’t have any particular argument to expose (though some are written in my posted links’ answers).

I’m 100% sure I have answered at least one bot at some point on Reddit in the last years or months, and I don’t like this idea. It reminds me of my dog who once tried to interact with a lifeless terracotta pig at a flea market, thinking it was a living creature (that was both sad and funny).

Now, if a topic is created by a bot, and is labelled as such, and a legitimate, constructive discussion starts between participants about the proposed topic, why not?

I hope I understood your post well, and that I’m not off-topic 🙂

---

<div class="post-metadata">

### Author: ![Bas](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/bas/32/294929_2.png) [@Bas](https://meta.discourse.org/u/Bas)
#### Post date: [August 7, 2025, 6:32am UTC](https://meta.discourse.org/t/thoughts-around-simulating-topics/377632/3 "2025-08-07T06:32:19Z")

</div>

> [@noahl](#):
>
> Is there anything I should think about before we go forward with this effort?

Communities are built on trust. Breaking that trust risks immediately losing all your high-value contributors.

For me personally: If I found out that I spent hours answering some bot, I would never volunteer my time to a community again. It breaks the social contract under which a community operates.

* * *

This is key:

> [@Canapin](#):
>
> Now, if a topic is created by a bot, and **is labelled as such** , and a legitimate, constructive discussion starts between participants about the proposed topic, why not?

If you’re honest and transparent about it, then it’s probably a perfectly fine experiment to run.

---

<div class="post-metadata">

### Author: ![Ed\_S](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/ed_s/32/134015_2.png) [@Ed\_S](https://meta.discourse.org/u/Ed_S)
#### Post date: [August 7, 2025, 8:09am UTC](https://meta.discourse.org/t/thoughts-around-simulating-topics/377632/4 "2025-08-07T08:09:15Z")

</div>

> [@Bas](#):
>
> Communities are built on trust. Breaking that trust risks immediately losing all your high-value contributors.

quoted for truth. If you’re pursuing growth at all costs, and you don’t care about your reputation, intend to have no relationship with your users, that’s another case.

---

<div class="post-metadata">

### Author: ![Lee\_Ars](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/lee_ars/32/1597_2.png) [@Lee\_Ars](https://meta.discourse.org/u/Lee_Ars)
#### Post date: [August 7, 2025, 10:34am UTC](https://meta.discourse.org/t/thoughts-around-simulating-topics/377632/5 "2025-08-07T10:34:11Z")

</div>

> [@noahl](#):
>
> To me, at its core, this is no different to organic conversations and answers - except we are planting these conversations for the purpose of better SEO/GEO.

This is the opposite of organic conversations and answers.

Doing this without disclosure to the community is grossly disrespectful both to your users and to the spirit of the community you’re trying to create.

Do what your boss says, but frame what you’re doing properly and don’t try to put a fig leaf over it. This is not “organic.” **This is faking content as a SEO play, full stop.**

---

<div class="post-metadata">

### Author: ![NateDhaliwal](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/natedhaliwal/32/313494_2.png) [@NateDhaliwal](https://meta.discourse.org/u/NateDhaliwal)
#### Post date: [August 7, 2025, 11:39am UTC](https://meta.discourse.org/t/thoughts-around-simulating-topics/377632/6 "2025-08-07T11:39:27Z")

</div>

Interesting concept, to get the engagement going “artificially”. My view on this is that if it’s solely to boost SEO rankings, then hey, that’s a bit like cheating right? But at the same time it could also spark a discussion (which I believe is also the point).

Like others I think have mentioned, stating that it is a bot is important. Some users may not care much for SEO either.

I would say this - use these conversations also for discussion, not just SEO. In fact, I would say that thoughtful, enriching discussions are more important than SEO - it’s about the community.

If SEO improvement is a side-effect then, well, that’s an added bonus. But my opinion would be not to make SEO the center, instead the community and enriching discussions.

Not sure if this is all completely relevant, please feel free to correct me.

---

<div class="post-metadata">

### Author: ![Bas](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/bas/32/294929_2.png) [@Bas](https://meta.discourse.org/u/Bas)
#### Post date: [August 7, 2025, 1:21pm UTC](https://meta.discourse.org/t/thoughts-around-simulating-topics/377632/7 "2025-08-07T13:21:44Z")

</div>

I ran a little experiment with Discourse HelperBot and I think this was pretty cool.

> **[New Discourse Features: Five Unasked Questions](https://meta.discourse.org/discourse-ai/ai-bot/shared-ai-conversations/ey8W6ultnvD5TJu-zxGl5A)**
>
> Bas: Draft five original questions about Discourse that haven’t been asked in the forum before. Base them on recently added documentation

If you use this – **properly labeled** – in public, that could actually spark some interesting discussion. Last year we went through a similar exercise internally at Discourse. Where we played a game of “trip up the bot”. Any time someone managed to make the bot answer with a lie.. we’d fix the documentation.

And yes, I do think having something like this could help with SEO/GEO. Have the AI generate a ton of content, which you verify to be true. Have it be a part of the Google index, but put it in a default-muted category maybe?  
It can certainly be helpful for other users who are using search phrases that don’t appear in the documentation for example.

But don’t break trust, never break the trust 🤗

---

<div class="post-metadata">

### Author: ![chapoi](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/chapoi/32/537252_2.png) [@chapoi](https://meta.discourse.org/u/chapoi)
#### Post date: [August 7, 2025, 2:11pm UTC](https://meta.discourse.org/t/thoughts-around-simulating-topics/377632/8 "2025-08-07T14:11:06Z")

</div>

> [@Bas](#):
>
> actually spark some interesting discussion

Exactly, I wouldn’t want to undermine that kind of intention either. I personally wouldn’t mind a forum where an AI is trying to start conversations at all; think “What’s everyone’s opinion on XYZ”; but as others have said, it just has to be clear it’s AI.

I definitely think there can be value in it, especially for communities that are just starting out and having trouble filling up their content, but I’m convinced quality above quantity is the way to go.

---

<div class="post-metadata">

### Author: ![noahl](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/noahl/32/552343_2.png) [@noahl](https://meta.discourse.org/u/noahl)
#### Post date: [August 7, 2025, 2:33pm UTC](https://meta.discourse.org/t/thoughts-around-simulating-topics/377632/9 "2025-08-07T14:33:10Z")

</div>

Appreciate all the thoughts and perspectives shared here 🙌

Just to clarify one detail that might help add context - to make sure I explained it fully:

This wouldn’t involve bots, AI, or company-branded accounts. These would be real people from our team, using their personal accounts, asking real questions _(which happens already from time to time)_ - the same way anyone might if they were facing an issue and turned to the community for help.

The only difference is that the questions would be intentionally seeded to cover topics we know are valuable _(from an SEO/GEO perspective)_ but aren’t yet well represented in the community.

Not sure if this shapes the conversation, but definitely hear all the points about trust and transparency, and I really appreciate the discussion 🙏

---

<div class="post-metadata">

### Author: ![Moin](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/moin/32/554653_2.png) [@Moin](https://meta.discourse.org/u/Moin)
#### Post date: [August 7, 2025, 2:47pm UTC](https://meta.discourse.org/t/thoughts-around-simulating-topics/377632/10 "2025-08-07T14:47:06Z")

</div>

This topic reminded me of this blog post

> **[8 Tips to Improve Your Community’s SEO](https://blog.discourse.org/2021/10/forum-seo/)**
>
> Helping people find your community takes time and patience. It also takes a little knowledge about search engines, how they work, and how people locate your community through them. In this article, we’re going to demystify the basics of SEO and how...

> [@8 Tips to Improve Your Community’s SEO | Blog](https://meta.discourse.org/t/8-tips-to-improve-your-community-s-seo-blog/206438/1):
>
> ### 1. Start truly helpful discussions
> 
> With the quality and relevance of your site’s discussion content as top priority in the eyes of search engines, the best way to improve your site’s SEO is to start, moderate, and curate valuable discussions. This can range from [user-to-user support](https://blog.discourse.org/2021/07/support-community-customer-experience/), asking and answering questions, or sharing thoughts on a relevant topic to stimulate conversation.
> 
> If a conversation is a match for a user’s search intent, it’ll increase the click-through rate from the search engine results page (SERP) and the chances someone will link to your forum in a natural way.

> [@8 Tips to Improve Your Community’s SEO | Blog](https://meta.discourse.org/t/8-tips-to-improve-your-community-s-seo-blog/206438/1):
>
> ### 7. Research keywords for important discussions you’re seeding
> 
> Seeding discussion is an important part of starting any community. Nobody wants to join a ghost town, right? One tactic to employ when seeding this content is researching relevant keywords to target. You can use a paid tool like [Ahrefs](https://ahrefs.com/?ref=blog.discourse.org), [SEMrush](https://semrush.com/?ref=blog.discourse.org), or [Moz](https://moz.com/?ref=blog.discourse.org) to dig into the keywords other sites are targeting, or you could use a free tool like [Google Trends](https://trends.google.com/trends/?ref=blog.discourse.org) to find out what people may be searching for. The point here is to try to target topics those interested in your niche will be looking for. Don’t forget to write the content in a way that starts a discussion while also providing valuable information to any visitors.

---

<div class="post-metadata">

### Author: ![Lilly](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/lilly/32/575047_2.png) [@Lilly](https://meta.discourse.org/u/Lilly)
#### Post date: [August 7, 2025, 2:48pm UTC](https://meta.discourse.org/t/thoughts-around-simulating-topics/377632/11 "2025-08-07T14:48:40Z")

</div>

> [@noahl](#):
>
> This wouldn’t involve bots, AI, or company-branded accounts. These would be real people from our team,

I guess it is a sign of the times when it seems most (including myself) assumed you were talking about bots. 😆

---

<div class="post-metadata">

### Author: ![chapoi](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/chapoi/32/537252_2.png) [@chapoi](https://meta.discourse.org/u/chapoi)
#### Post date: [August 7, 2025, 3:33pm UTC](https://meta.discourse.org/t/thoughts-around-simulating-topics/377632/12 "2025-08-07T15:33:43Z")

</div>

I think the word “simulating” tripped us all up. Mods posting topics to get things going just feels… normal to me? xD

---

<div class="post-metadata">

### Author: ![Bas](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/bas/32/294929_2.png) [@Bas](https://meta.discourse.org/u/Bas)
#### Post date: [August 8, 2025, 7:12am UTC](https://meta.discourse.org/t/thoughts-around-simulating-topics/377632/13 "2025-08-08T07:12:41Z")

</div>

Yeah, that doesn’t seem all too bad 🙂

Many, many, communities have started that way. Usually people stop after the community takes off. But doing it for SEO reasons makes total sense and is wise community management.

Don’t overdo it though. In stage i’d probably limit it to 2-3 topics per week at most, probably less. Still nets you 100 topics in a year, good win for SEO.

---

<div class="post-metadata">

### Author: ![Kartoon](https://avatars.discourse-cdn.com/v4/letter/k/e36b37/32.png) [@Kartoon](https://meta.discourse.org/u/Kartoon)
#### Post date: [August 10, 2025, 12:52am UTC](https://meta.discourse.org/t/thoughts-around-simulating-topics/377632/14 "2025-08-10T00:52:15Z")

</div>

> [@Canapin](#):
>
> I’m 100% sure I have answered at least one bot at some point on Reddit in the last years or months, and I don’t like this idea. It reminds me of my dog who once tried to interact with a lifeless terracotta pig at a flea market, thinking it was a living creature (that was both sad and funny).

I think the main difference to “is it upsetting or not” is are you answering a QUESTION or participating in a DISCUSSION?

If a bot asks how to set up a garden or something and you spend a lot of time giving a detailed answer, that would be annoying because it doesn’t care, it isn’t helping, and it’s (possibly) pointless (it could end up helping someone who wasn’t the intended target at a later point, as is often the case with reddit overwhelming google results). Now if the bot is just starting a discussion about something (“What are some gardening techniques you’ve picked up”) where you’re not actually responding to the bot and the original post isn’t really the focus of the topic the posts are, you (hopefully) have a discussion amongst people in the comments, _prompted_ by the bot; the bot is not really important and actually mostly irrelevant.

In the second case it’s not any different from an admin/mod starting some random starter topics just to get the site going. And if it’s relevant to the board I don’t know how many people would really care how the discussion was started even if you hid the bot interaction and it was discovered later.
