The purpose of this post is to begin a dialogue about data that would be useful for Bankless DAO’s growth.
Incidentally, this is also part of a larger effort to begin collecting and storing data that would allow the Analytics Guild to report on / display data useful for the DAO. For this post, we’ll start with data we currently have access to (minimal cleaning and handling required).
The data comes from our own Discord server and can shed light on member growth, activation and engagement.
Unclear what Discord is noting as “benchmarks” presumably all servers, even those outside of DAOs/crypto
We’re working on getting membership roles/levels data (for a separate post)
Data is presented daily June 1st - August 16 (only data from last 120 days are available for non-Partner / non-Verified servers)
As you read through this post, you might note down what specific questions you’d like answered so we, the Analytics Guild, can begin collecting and storing the kinds of data that could answer those questions. Feel free to drop those questions in the comments.
GROWTH & ACTIVATION
These numbers provide an at-a-glance view of growth and engagement.
We can see number of new members from the previous day, and new communicators. Discord makes a distinction between “visitors” and “communicators” (see Engagement section below). A notable metric is New Member Retention showing % of new members that still came back to the server a week after joining.
How many new members are joining?
(where are they coming from?)
This shows the total number of people joining over the last 3 months and where they joined the Bankless DAO server. Early on (May) there were a lot of invites and overtime, there’s been more Vanity URLs (e.g., https:// discord.gg/bankless) as the way to help newcomers join.
Server Leaves over time
1 month appears to be a threshold. Members who are around for at least a month are less likely to leave. Certainly the case early August.
How many new members successfully activate on their first day?
How many new members retain the next week?
Retention numbers, on whole are higher than the benchmark of 20%. This is tied to “activation” (see above).
Most popular referrers
We have dramatically more “unknown” referrers. (n = 1,172) compared to the next highest referrer of www.google.com. It’s unclear how Discord came to label these popular referrers as “unknown”.
- Visitors = a user who has clicked into the server and viewed a channel
- Communicators = a visitor who talked (voice or text)
- Total Messages sent
- Total Voice minutes
If you wanted to see active members on a weekly basis, one metric could be the ratio between visitors : communicators, here are the current weekly stats:
We can see that from 2059 members who clicked into the server and viewed a channel, 280 (13.5%) talked or texted. cc: @frogmonkee
How many members visited and communicated?
We have more visitors than communicators; more people who click in and viewed silently than people who send at least 1 text message or has spoken in a voice channel.
This might be a consideration for First Quest or Onboard (although I think Onboard is more focused on DeFi). cc: @RedVan
This chart re-iterates that for the number of “visitors”, we have fewer “communicators” (below 30% benchmark), and communicators are averaging less than 10 messages per person.
Zooming out a bit for a weekly view, we see a drop in Message Activity from mid-to-late June:
We’re averaging 500 - 600 speaking minutes; the drops are generally during the weekend.
The following questions are broken down into Categories of Users including:
This will require additional data crunching, but here are the top-3 per category:
Which text channels do people use the most?
(messages from bots are excluded)
- #announcements (2084)
- #general (1236)
- #first-quests-welcome (612)
- #join (296)
- #general (120)
- #cc-chatter (86)
- #cc-chatter (1029)
- #general (777)
- #bounty-board-alerts (661)
Which voice channels do people use the most?
- #cross-collab-voice (56)
- #dev-workroom (45)
- #Treasury-hall (34)
- #cross-collab-voice (42)
- #dev-workroom (40)
- #Treasury-hall (28)
Inactive Users or How many users could be pruned?
- Inactive for 7 days (1,903)
- Indactive for 30 days (1,131)