Playbook, online course

The sales page is fine and the course will not load

A course business usually runs two things that fail independently: the site that sells the course, and whatever serves it once somebody has paid. The assistant lives on the first and has never been able to see the second, which means it can be answering fluently from a working page about a system that is completely down.

Why this is not the general answer

The handling pattern for service outages holds across every trade. What follows is the part that does not.

  • The two halves are separate systems, so the widget answering confidently is not evidence of anything at all, and a learner takes a working sales page as proof that the problem must be at their end.
  • A live session is a fixed appointment rather than an hour of work, so an outage during cohort week costs an event that cannot be repeated, and the recording promise becomes the only remedy anybody has.
  • For a self paced learner the cost is the one evening they had set aside that week, which is a small loss that feels large, and a reply written for a business customer reads as dismissive of it.
  • Small course businesses rarely run a status page, so unlike almost every other trade the honest answer here often has no destination, and the only truthful move is to say so and collect the detail.

How it arrives

  • the course site will not load is it down for everyone
  • the live session is in an hour and i cannot get in
  • when will the recording be posted if the session broke
  • everything worked last night and nothing works this morning
  • will you extend my access for the days it was down
  • how will you tell us when it is fixed

What has to be indexed for this to work

Material behind this answer
Where you announce problems, whatever that place isA status page, the community, a notice on the login screen or an email. Name the one place learners should look, and if there is genuinely nowhere, that is the gap this pair exposes and it costs an afternoon to close.
The recording commitment for live sessionsWhether sessions are recorded, how many days until a recording is posted, and what happens when a session could not run at all. This is the promise people count against, so publish the number of days rather than a word like shortly.
Your position on extending access after a disruptionWhether time lost is added back, and to what. Learners on a fixed access period ask this immediately, and an unpublished position means somebody has to invent one under pressure and then apply it inconsistently.
Planned maintenance and how it is announcedWhen you take things down deliberately and how much warning goes out. A scheduled window experienced as an outage produces an angry message that was entirely preventable by one line in an email.

The reply

A reply worth copying
I have no view of the course platform at all, so I cannot tell you whether it is down. I answer from the pages on this site, and those keep working whether the platform is up or not, which is worth knowing before you take anything I say as reassurance. Where problems are announced is on our support page, and that is the place to check [1]. If a live session is affected, the recording position is published there too, including how many days it takes for one to be posted [2]. If you leave your name, the email you enrolled with, your cohort if you are in one and the time this started, it goes straight to the team.

It disqualifies itself in the first sentence, because every reassuring sentence after that would be an accident. Explaining why the site still works stops the learner concluding the fault is theirs. And it collects four things a person needs, with the cohort included, because that is what makes one of these urgent rather than annoying.

Where it stops

The trigger. The visitor asks whether the platform is down, when it will be back, or says a live session is about to start and they cannot get in.

The handover, worded
I cannot see the platform or say when something will be fixed. Leave your name, the email you enrolled with, your cohort if you are in one and when this started, and it goes to the team now rather than into a queue.

It stops answering before it guesses, says who will pick it up, and asks for the one thing that makes a reply possible. Nothing about it reads as a dead end.

Never say this here

Out of bounds

  • Never say the platform is up, working normally or unaffected, because nothing here observes it.
  • Never estimate when something will be back, or what caused it.
  • Never promise a recording, an extension of access or a repeat session, since all three are decisions for a person.
  • Never tell a learner the problem is at their end, as there is no way to tell that apart from an outage from here.

Questions

We have no status page. What should it say?
That it cannot see the platform, and then take the details. It is a poor answer and it is the only honest one. The fix is not a better sentence, it is a page or a notice somewhere that a learner can check without writing to anybody.
Why not index our status page so it can answer properly?
Because indexed material is read on a schedule, so during an incident it would be quoting a page that changed several minutes ago, with a reference attached that makes it look authoritative. Handing somebody the page is faster and cannot be wrong.
Is it useful at all while the platform is down?
Yes, because it keeps answering everything that is not about the outage, and it collects a clean report from the people who are affected: the cohort, the time, the email. That is a list you can email an apology to, which is the only thing that actually helps afterwards.

Keep reading

Try it on your own material

Upload a document or point it at your site, paste one line of HTML, then ask it something only your business could answer.