Enrolment is the least discussed part of training tooling and one of the most consequential, because a course assigned to the wrong set of people produces confidently wrong completion numbers.
Channel-based assignment fixes a specific failure: lists going stale.
Why lists go stale immediately
If you assign training by selecting individuals, the assignment is a snapshot. It was correct on the day you made it.
Then someone joins the team and isn't on it. Someone moves departments and is still on it. Someone leaves and their incomplete assignment sits there forever, dragging the completion number down for reasons nobody remembers. Within a quarter the list describes a team that no longer exists.
The reporting consequence is the dangerous part: completion reads 100% while a real gap exists, because the person who should have been trained was never counted. That looks like success.
Why a channel stays accurate
A channel is maintained continuously, for reasons unrelated to training. People are added when they join the team because they need to see the conversation, and removed when they leave for the same reason.
That maintenance is done by the people closest to the team, as a side effect of normal work. It's far more reliable than any list a training admin curates, because nobody has to remember to do it — the incentive is already there.
So /assign #support targets whoever is actually in #support now, not whoever was in it when the course was created.
Assigning to future joiners
The stronger version handles onboarding: assign to everyone who joins from a date forward.
This is what turns onboarding from something people remember to do into something that happens. A new hire gets the sequence whether or not anyone noticed they started, whether or not it was a busy week, whether or not the person who normally handles it was on leave.
It also means the process survives the departure of whoever set it up, which is a common and quiet failure.
When a channel is the wrong unit
Channels are a proxy, and proxies fail in specific ways worth checking:
- Channels with observers. If managers, stakeholders, or a bot sit in
#supportwithout being support staff, they'll be assigned training they don't need. - People who need it but aren't in the channel. Someone covering a function occasionally may need the training without being a member.
- Overly broad channels. Assigning to
#generaltrains everyone, which is right for a company-wide policy and wrong for anything role-specific. - Archived or renamed channels. Worth confirming an assignment still points somewhere live.
Before assigning, look at the member list and ask whether it matches who needs the training. Usually it does. When it doesn't, assign to individuals for that case rather than distorting the channel.
A practical pattern
Company-wide policies — code of conduct, security awareness, expenses — assign to #general plus all future joiners.
Role-specific training assigns to the relevant team channel, which keeps it accurate as the team changes.
One-off or remedial training assigns to individuals, because it genuinely is a snapshot and should be.
Checking coverage
Whatever the unit, verify the denominator. /progress returns live completion by team, so it's quick to check whether the people you expected to be assigned actually were — before an auditor does it for you.
Add Poplearn to Slack and try assigning a course to a channel. Free under 20 seats, installs in about a minute.
