We took on this project because as the community grows and we increase the number of things we do, it increasingly becomes more challenging to learn about the community as a newcomer and for current members to keep up with the changes. Our goal with the Knowledge project is to develop a sustainable model to improve awareness among members of how the community works and what the community has to offer. We began by building out a Support Driven handbook our knowledge base — as explained in the previous post.
Once we wrapped up the initial rewrite of the user manual, the question became “what’s next?” We knew that we wanted to build something sustainable that would help new members of the community. The how was a bit more up in the air.
Enter The Bots
Support Driven is heavily driven by Slack, but as anyone who has spent time in a free Slack account knows, messages disappear quickly (generally after about a week). We built the manual to store knowledge. The question became: How do we deliver it? How will the community know this manual exists?
We also wanted to make sure new members of the community felt welcomed. We wanted to make sure they knew how to interact and how useful Support Driven is. Jumping into an active Slack community can be overwhelming and we wanted to smooth the way for new members.
The question, though, was how? After much discussion we decided to approach the problem (welcome new members and sharing all that useful information) from a couple of different directions. The first, bots.
Because so much engagement happens in Slack, it made sense to look into Slack as the distribution method. We could use a bot of some kind to send messages to brand new members and help them get the lay of the land. But how?
Well initially we dove straight into the deep end and were researching bots to automate the whole process, but then we realized that was the wrong method. Why? Right now adding on a Slack bot was a theory, we were investing time and effort into a project that we thought would help, but we weren’t sure. We needed to prove if this would genuinely be helpful before we put the effort into automating.
As an MVP (minimum viable product or in this case process) rather than having a bot deliver a welcome, I did it myself. Whenever a new member joined the Slack I’d pop up and send them a welcome with a bit of information. It looked something like this (the message varied over time as we narrowed down more what members wanted).
From that MVP we learned that this really would be useful. New members receiving those messages, anecdotally at least, seemed to be diving more into the chat and asking questions (or sharing answers!). They also seemed to respond well to the message and said that it felt really helpful.
Manually sending messages was time consuming, though. And it wasn’t uncommon for someone to be welcomed 24 (or more!) hours after they joined, because one person can’t be online 24 hours a day. This meant new members might end up feeling lost for a while and I definitely got more response from people when they got a prompt welcome versus a delayed one. That wasn’t ideal.
So with a general feeling that the messages were going to be helpful, but it wasn’t sustainable to do them manually, we start investigates bots.
We tested out quite a few different bot options, from Janitr to the new onboarding bot made by the folks at Donut. For a variety of reasons none were just right for us, but then Guru stepped in with an idea. Their bot could welcome new members by automatically DMing them when they join. It looks something like this (though without the posted by me part!).
Now whenever someone joins they’ll get a little welcome and learn more about SD right away.
We want to meet community members where they are — whether that’s Slack or email. The bots help when a new member joins the Slack, but for someone’s never used Slack, or hasn’t made their first login yet, we want them to feel welcome and know what the community is all about.
So next up…an email onboarding campaign. Stay tuned for part 4 and learn how we built it and what we built!
Want to volunteer to help maintain our knowledge or share knowledge? Reach out to @dpotter or @andreabadgley in the Slack for more details. Join the #workgrounp-km channel!