Delivering legendary service is something the Atlassian Support team is constantly striving for. At Atlassian Summit 2010, Andrew Rallings, our Support Director, shared the three ways we try to achieve this goal and some of the innovative tools we have created to better serve our customers.
In the following video, check out Andrew Rallings’s keynote (approximately 10 minutes in length) about Atlassian’s Legendary Support. (Starts at 48:59).
Firstly, how do you know if you are delivering Legendary Support?
The only true way of knowing if you are delivering legendary support is by asking your customers. Each month we send out a Customer Satisfaction Survey to a randomly selected group of customers who have raised support issues that month. We roughly get a 40% response rate which is around 4 times the industry average (thank you to all our customers for such an awesome response rate, you guys rock!).
Measuring our support quality is one thing but how do we go about achieving it? We identified three ways to achieve legendary support…
1) Build products with no issues
When it comes to support, the best service you can get is the one you never need. Development, Product Management, Quality Assurance, Performance Engineering and Support work closely together to deliver not only a great product, but a high quality product with as few bugs as possible. As a result we have a number of large improvements scheduled to be delivered over the next 3-6 months such as the Universal Plugin Manager, improved user management and LDAP integration, and a dramatically improved Rich Text Editor for Confluence. Also, in the last year alone, we have fixed approximately 3,500 bugs across all of our products. There is still a long way to go, in particular we want to make installing and upgrading even easier, but we are heading in the right direction.
2) Provide great tools and online help
When you do have issues, we want to provide you the best online help and tools available that can help you find the answers to your issues quickly any time of the day.
Some of the things we have done to do just that include:
- Revamped The Knowledge Base: by making known issues easily findable, we’re empowering customers with the knowledge they need to prevent known issues from occurring in the first place (see our Confluence Knowledge Base as an example);
- Released Hercules the Atlassian Support Bot: our friendly Hercules Support bot scans logs and support issue descriptions to display, in seconds, related known issues from within our knowledge base and/or public JIRA instance (Hercules was actually created by Atlassian’s support engineers);
- Released Our Support Tools Plugin: available for Confluence (and soon JIRA) our support tools plugin allows you to scan your instance to detect known issues before you even knew you had a problem. Over the next 6 months we will release similar plugins for the rest of our products.
3) Deliver best-in-class service
When you do need to file a support issue, we want to provide you with the best service we possibly can. In your customer satisfaction surveys you said that you wanted two things: (1) you wanted us to answer your issues for your product in your time zone and; (2) you needed our service to be reliable.
To achieve this we needed to make some pretty big organisational changes. We recently opened a new office to cover the US East Coast and grew our entire Support team by more than 50%. What this means is that by October 2010, you will have a support engineer for your product (i.e. all products) in your time zone.
In terms of reliability, we have already published our target Initial Response Times but over this next 6 months we will publish exactly how we are doing against these targets. Providing transparency to our response times is a great way to demonstrate increasing reliability. We also created a highly customised Shared Support Queue that helps us better prioritise our incoming issues, to ensure the most efficient and effective delivery of support.