My biggest issue about Slack is that it really doesn't live up to its mission: "Where work happens." Work may happen, or it may not happen in Slack. As an Engineer, I find it distracting. Between having to go to meetings and coding, I also have to stay on top of conversations within Public and Private channels. (Sometimes, there are a lot of channels.)
So how can we improve Slack to make it more useful?
One way to quantify "useful" conversations
One metric to measure our solutions against is "meaningful engagement" in a given week. Assuming we have access to great Data Scientists, I'd look at all the conversations Engineers have. Specifically, I'd count a conversation as useful if it has certain words like "feature", "JIRA", "ticket" and more. From there, we can gleam:
# of useful chats in a week
total # of chats in a week
With this metric, let's see how we can create better more meaningful conversations.
Several of ideas
1. Role-Based Notification Icons
The basic idea here is that, instead of the usual red dot you see for notifications, you would see that you had a "E1" or "E2" notification. This means you had 1 Engineer or 2 Engineers who have sent you messages recently. As an Engineer, you might think this is useful because, though you have minimal interaction with Marketing, seeing an "M1" notification might make you pay attention more -- marketing might really need something.
This would be easy to implement. Slack already has an integration with internal HR systems on a person's role. To build this feature, you could simply look at the Slack Profile Role and take the first letter.
2. "Rainbow" AI Notifications
Every chat notification an Engineer sees will be a different color based on how likely the message is to be useful. If red is the most important, we make light grey the least important kind of notification. We can also base this color on how "meaningful" of a conversation the sender usually starts. To be specific, this color would be based off a the "meaningful" engagement metric we discussed above.
To implement this would take a bit more effort. We'd have to have Data Science give us their analysis, we'd have to create a table for mapping and updating how useful each person's conversations are, and then create an API for the FrontendEngineers to consume.
3. AI-driven a cutoffs for future
We can offer companies a feature to turn on chat windows during the most "meaningful" hours. That might be for 4 hours of the day or 6. This time can shift between different times of the day - kind of like a Nest thermostat.
(Imagine if Slack could turn on only when "meaningful" conversations are most likely to happen.)
This way Engineers and other employees are only required to pay attention when conversations are going to be useful.
To implement this, we can again use the "meaningful engagement" metric. From there, we can extrapolate when people within the company -- including Engineers -- have "meaningful" conversations.
My verdict
Of the three options, I think the most promising seems to be "Rainbow" AI Notifications. While it requires a little more engineering effort, I think the payoff could be huge. The Engineer will have certainty with what notifications are most important based on the color. And over time, the AI will get smarter and telling you what is chats are most meaningful.
Comments
Post a Comment