State of the Discord Bot Community 2019

When making Discord bots it’s important to know what others want from them and how they are used. Unfortunately there’s no easy way to get this information. Because of this I created a survey asking some of the most asked questions from bot developers. Here are the results:

*Percentages are rounded to the nearest whole number

About the Respondents

235 people completed the survey, which was shared on the Bots on Discord website and across some Discord servers and bots. Of the people who completed the survey,

• 74% were server owners or administrators.
• 51% were bot developers

When asked about how much they use Discord, 46% said they always have Discord open, and an additional 35% said they spend a few hours on Discord every day.

Bot Usage

Server owners were asked how many bots they have in their servers on average. The chart on the left shows the responses for non-devs, but the chart for all responses is very similar. We can see that almost half say their servers use a large number of bots. Few people use less than three bots on average.

Regarding custom bots, 63% said they have custom bots for their server, but 23% of those who did said they would use a public bot if available.

Additionally, 88% of server owners said their server relies on having bots. 29% of those said their server is completely dependent of having bots.

Adding bots

When asked how they handle permissions, 35% said they change some permissions from what is requested when adding a bot. An additional 26% said they create their own role and decide what permissions a bot gets themselves. The remaining 39% said they do not modify the permissions of bots when adding them.

Regarding the controversial Administrator permission: 39% of respondents said they are fine with bots having it, and an additional 49% said they are fine with it if the bot really needs it. This suggests that people are generally trusting of bots.

Categories and Satisfaction

Excluding responses from bot developers, the most used categories are music and fun / memes. The least used are NSFW, utility, and verification.

Looking at how satisfied people are with the bots they use, fun / meme bots are the most liked, while NSFW bots are the least liked.

For charts including bot developers see the links at the end. Results are similar when including all responses.

Finding Bots and Preferences

When asked if they prefer multi-purpose or single-purpose bots, 70% said they prefer multi-purpose. It seems that users would rather only have to deal with one bot instead of many, each having it’s own syntax ans systems.

We were also curious to see how many people cared what a bot is written in. The results show that around 1/3 of non-devs do care. Among developers around half care.

We’ve also noticed a number of users that get annoyed by bots sending them help in DMs, so we asked where people prefer to be sent help messages. 51% said they prefer to get general help responses (for all commands) in DMs, but only an additional 20% said they always prefer DMs.

Bot lists

When asked where they find bots people gave the following responses:

1. 66.8% — discordbots.org / top.gg
2. 45.6% — Personal recommendations
3. 36.3% — discord.bots.gg
4. 27.9% — bots.ondiscord.xyz
5. 24.8% — Non-listing site such as Reddit, YouTube, or other social media
6. 21.7% — Other bot lists (bots for discord was the most written-in)
7. 21.7% — Official Discord recommendations (streamkit / bots tab)
8. 16.4% — carbonitex.net

We also wanted to see if people would use an official Discord bot list, and 77% of respondents said they would use it over third-party lists.

Q: Are there any things you think bot lists should add or change to improve?

The first of four open-ended questions we asked was intended to see what people liked and didn’t like about bot lists. Here are some of the common answers. Of course it’s impossible to cover all the answers here, so if you’re interested you can read them all in the spreadsheet at the end.

• Many respondents complained about the commercialization of bot lists. They complain that the voting system promotes bots that are only in it for the money, making it harder to find small bots or bots that don’t want to force users to vote on lists. They also complain the owners of these bot lists don’t seem to care about bots or being able to find good bots, just making more money from ads
• Many responses wrote that they think bot lists should stop encouraging so much competition and instead just focus on helping people find good, unique, bots
• Some responses asked for more specific categories to search by and the ability to sort by server count
• Some responses asked for more customization of bot pages
• Some responses suggested adding a way to review bots or give feedback to owners.

Features

We asked how important some features were to people. Here are the results:

Ability to disable commands

On a scale of importance from 1 to 5, most respondents indicated that the ability to disable commands was important to them.

However, when it comes to NSFW commands, 20% of respondents said they would not use a bot if it had NSFW commands, even if they could be disabled.

Paid features

When asked if they pay for premium features in any bots, 77% said yes. However, 46% said a bot having paid features negatively affects their opinion of it.

Vote-locking

Vote-locking is very controversial so we had to ask a question about it. When asked if a bot locking features behind votes turns them away from using it, 63% said yes.

Configurable prefix

When asked about the importance of a bot having a configurable prefix, the responses were split. 46% said it is very important to them, and 41% said it isn’t.

Q: What features do you look for in bots?
Q: What features do you want which are not offered by existing bots?

Two open-ended questions were asked to see what features people want from bots, and which on those features they could not find a bot for. Here is a summary of what people said they are looking for in bots.

• Highly-polished and simple-to-use features
• Originality
• Commands that don’t disrupt the chat (long responses, accidental triggering of commands)
• Moderation features
• Commands to increase server activity (usually fun commands were specified)
• Customizability
• Music

The responses for what people wanted from but couldn’t find in bots can’t be summarized because there were so many different responses. If you are a bot developer looking for ideas I would suggest reading the responses yourself to see what people want.

Q: What features do you think Discord could add to make bots better?

In the final open response question I wanted to know what people though Discord could add that would make bots better. Here are some of the things people said that I picked out:

• Add UI elements to the client to interact with bots (menus to manage bots, buttons in messages, a bots tab that is actually useful, commands in client, etc.)
• An official place to find bots; ability to add bots without going to third-party sites
• The ability to be used in group DMs
• Better support / more transparency
• Custom statuses
• Communicate the dangers of bots (most people never think that a bot could do malicious things, intentionally or unintentionally)
• Making it easier to make bots (better API, ability to disable events, more support, search API, etc.)
• Bot stats (similar to server stats)

Languages

Discord is used by people around the world, so we asked a few questions about what languages people like to use bots it.

Overall there was not enough data to give a good answer on which languages are most popular. You may look at the data yourself and see which languages were picked, but here are the top 5 preferred languages:

1. English
2. French
3. German
4. Spanish
5. Arabic

From non-native English speakers only 19% said they prefer to use bots in their native language instead of English. Among non-native English speakers 47% said multi-language support is nice but not necessary, and 24% said it doesn’t matter to them. Only 13% said it’s a must-have. Again, these results are heavily skewed due to the way this data was collected, so don’t take it as fact.

Conclusion

Thank you to everyone who responded to the survey and who shared it with others. I hope it provides value to the bot developers out there. Next year I hope we can get more servers and services on-board to help increase the number of responses and diversity of respondents.

If you would like to see the responses yourself you can get them here:
https://docs.google.com/spreadsheets/d/1Mg1tCqF3pDCgLWFGyR02YW4wCaQUL5E4Lmn7wsxkc8A/edit?usp=sharing
This spreadsheet also includes the charts used in this article.