New Ways to Measure Engagement at BanklessDAO

TLDR:

Instead of relying solely on Discord’s definition of engagement, we can better measure engagement using additional metrics including:

  • Number of First Quest Tag(s) achieved
  • Amount of Tips Received
  • Level of Churn

Discord measures engagement using message activity, voice activity, including the number of readers and chatters in various channels. We spend so much time in Discord, this became a convenient way to consider engagement at BanklessDAO as documented here, here, and here

The metrics Discord provides are aggregates of our entire Discord,

but sometimes we want more granular data. DAO Dash allows us to captures this. Here’s a chart of the most messages sent by members in the last 30-days.

Or if we wanted to see activity over the past week, DAO Dash helps us filter for that:

Now this might raise the question,

How many weekly active members are there within the DAO by roles?

One role I’ve been looking at recently are the First Quest related roles.

We could ask, out of all the Bankless DAO members who have attained their “First Quest Complete” role tags, who sent the most messages in the past 7-days?

We can see differences in the list between these two charts showing us there could be meaningful insights to filtering by roles.

Because members can have multiple roles, it can be tricky to attribute differences in engagement levels to any specific role.

What if roles were actually indicators of engagment themselves?

We’ve shared data with the First Quest team to see how new members were progressing through the various quests, but we hadn’t considered the attainment of First Quest tag(s) themselves to be indicators of engagement.

Consider the First Quest funnel.

As mentioned in this post, although it should take new comers 15-20 minutes to work their way through the quests, we see a significant drop off after First Quest Welcome. By the time a new member completes First Quest (attain their ‘First Quest Complete tag’, there’s been roughly an 80% drop-off).

first_quest_funnel

In the past 30-days, only 110 out of 693 new members attained their First Quest Complete tag (only 15% follow through).

The data indicates that completing all First Quests requires effort given how many people drop-off or fail to follow through, so we can argue that having the First Quest tags are a sign of engagement.

Another metric worth considering are Tips Received.

Earning tips indicates a member is around long enough to add value. They’re engaged enough to contribute ideas, effort, knowledge, skills and abilities to receive tips.

There are certain benefits to using Tips Received, it’s a indication of an active behavior on the part of the recipient to have contributed in some way. Economically, tips go towards helping guest pass holders become L1 and there’s no better sign of engagement than someone earning their BANK.

Now that we’ve considered progressively more behavioral indicators of engagement, from messages to completing First Quest(s) to earning tips, we might combine metrics.

Do people who go further down the First Quest funnel earn more tips?

Data from the past 7-days suggests that the further down the First Quest funnel a member goes, the more tips they’re likely to earn. We can’t make any conclusive statements (yet) and we’ll have more confidence as we get more data:

One possible explanation is that member who make the effort to go further down the First Quest funnel (i.e., complete more Quests) will, over time, display the same level of effort that earns tips through various contributions.

This might useful data for the First Quest team where First Quest isn’t just an onboarding tool, but could potentially be predictive of a new joinee’s tips earning potential (although we can’t make any conclusive statements at the moment) cc: @angyts @Behold @ThePsychGuy

Finally, our community member @JokerDelMar has created a table looking at Bankless DAO member churn, which is the opposite of engagement, another metric we’ll monitor moving forward:

TLDR:

Instead of relying solely on Discord’s definition of engagement, we can measure engagement using a mix of metrics including:

  • Number of First Quest Tag(s) achieved
  • Amount of Tips Received
  • Level of Churn

The DAO Dash team will stabilize infrastructure to allow continued data collection so we can further explore the evolving nature of Engagement at Bankless DAO.

7 Likes

Wow, what a cool post! So awesome to see bDAO generating so much activity that we can measure it and draw findings from it.

Something I’m interested in: how many bots do we have trying to come through First Quest? That could pump the numbers. I noticed that our Community Call attendance dropped by about 50% today, since it was the first day non-guest pass members (bots, likely) were barred from joining. I’d be interested to see some community call engagement metrics in the future, as (to me) that is the best way to guess how many active members we have.

4 Likes

Thanks for sharing this @paulapivat.

Question regarding the suggested measurement metrics - can these be captured at the guild level too to have a weighted average of participation from new onboards i.e. completed First questers?

Part of my slides will focus on this regard, as the granular view from guilds give deeper insight to DAO engagement together with the overall DAO activity - especially for tips and churns.

What do you think?

1 Like

All great questions @Joey_magiA , I believe it’s possible to capture these metrics at the guild level. We can see when someone joined the DAO and also the time they activated a Guild tag.

You can explore the data tables discord_user, discord_user_roles and discord_roles for these.

At the moment, the data is patchy and we’ll work in Season 3 to stabilize these data sources.

If your slides could also shed light on Guild-level tips and churn, that would be pretty cool - happy to chat more :+1:

1 Like

Ooh, we’re also fascinated by bots! We’ve had to deal with bots in the Analytics guild…sometimes it’s now clear whose a bot or not.

@McKethanor - any part of the Discord data you’re pulling for us could shed light on “bot or not” ?

2 Likes

This is a great step in the right direction!

1 Like

Great insights! You are squeezing loads of useful metrics out of this data. Keep up the good work.

1 Like