Best Practices & Trends / webinar

How to scale support with service desk and KB

Learn how Atlassian's own build engineering team uses a combination of ticketing in JIRA Service Desk, and knowledge management with Confluence Questions, HipChat, and the Knowledge Base Blueprint in Confluence to scale the support they provide to a rapidly growing number of developers.

Q & A

Q: Does anyone ever call you on the phone? How do you track phone conversations?

A: We don't have phones, which is pretty much the norm in Atlassian. We do still get some private messages on HipChat though. In these cases we usually just flat-out request that they re-ask their question through one of our approved channels.

Q: Do you attempt to link similar questions asked within Confluence Questions?

A: Yes, sometimes we do this, especially if there's been some useful conversation in it (e.g. why some answers are bad options). But we usually prefer to link to the relevant knowledge base article or even create a new knowledge base article and link to that. This is because knowledge base articles (just Confluence pages) are easier to maintain. For example, I can't edit an answer that a colleague of mine has written, so it's easier for the team to keep information up to date in the knowledge base.

Q: Is there a smooth transition for HipChat help requests into Service Desk?

A: We don't have anything automated for this as people are usually happy to follow the link to our Service Desk and fill out the 3 (or so) fields themselves. It would certainly be possible to use a HipChat bot to automate this though – cool idea!

Q: Are you able to keep items persistent in HipChat (such as current issue being addressed via JIRA ticket)?

A: We have the HipChat topic up the top of the window that we sometimes change to notify people joining the room about high-impact issues, but that's not quite as visible as what it sounds like you're after. We also have a company-wide Outage Notifications room that is only used for serious outages. Since it's low-traffic, a lot of people stay in that room or check the backlog of that room when they notice a service is down or seriously degraded.

About the author

Michael Knight

Senior Support Engineer, Atlassian

I'm a Senior Support Engineer on Atlassian's build engineering team. After years of supporting Atlassian OnDemand users, I'm now helping developers run their builds and improving the build team's support processes. Find me on Twitter! @miknight

Wanna learn more?

Solve more problems faster

Try it free

Create and collaborate

Try it free

Want to be an IT Insider? Sign up to get the latest tips and tricks.