Software company Atlassian is one the Australian IT industry’s success stories. Starting with a credit card back in 2002, the Sydney-based company now has offices in California and The Netherlands, turns over $60 million and recently received a multi-million dollar investment from Accel Partners.
Co-founder Mike Cannon-Brookes attributes some part of the company’s success to the ingenuity of its engineers. The company allows each worker to use 20% of his or her time each week to a project of their choosing.
Cannon-Brookes says this has allowed the company to generate new products and features it may not have otherwise used, and encourages other businesses to develop some type of program to help spur innovation.
How has the business been travelling since the investment?
We’re up to 320 people worldwide, going very well – everything is onward and upward. With Accel being there, it’s been great as we’ve had things to learn on each side. In some ways it’s been much more helpful in ways we didn’t anticipate, so thoroughly good all around. We’ve had three really strong quarters since the investment.
Where did the idea of freeing up time for employee projects come from?
We’ve been doing FedEx days quite a lot, and our internal innovation has a number of different angles; 20% time is one of those, and FedEx days are another that we’ve actually talked more about.
A FedEx day is a 24-hour coding exercise for all the engineers, so they get to build their own mini-product and project. It’s fairly intense, and at the end of it a winner gets a big trophy – we do that about once a quarter, and it’s a good morale booster.
The downside is that people start a project and get really disappointed when they can’t finish it, and the genesis of 20% time was to give engineers a little bit of independent time and autonomy to work on whatever it is they want to work on, whether it be a continuation of those projects or something else.
When did this all start?
About two years ago. When we started it was essentially a very large experiment. We said, let’s try it for six months, and we’ll share our rules about what we do on the blog. We put some guidelines up around how we run it, and how engineers should follow those, but we try and keep it as free form as possible.
Did you take some inspiration from Google on this one?
Google does talk a lot about its 20% time, but there isn’t a lot of detail about how it all works.
How does it work now?
It’s actually changed quite a bit since then. The way it works now is, you have to register your project – that’s actually a controversial one, but it just takes one minute to write down what you’re doing and register it. It’s so people have a directory of what they are working on and can get other people to help if they want.
You say it’s controversial – did that bother some employees?
It’s quite informal, people basically say, “I’m working on this specific thing,” and another engineer can say, “Hey, that’s cool, can I help?” It works well.
We also have a peer review system. You want this type of system to be autonomous and productive, and you want people to have scope to play around, but you do want that review. So if you have a 24-hour work period, you get other engineers to support you and what happens is reflected in your performance review.
That doesn’t mean you get a bad performance review if you fail – not at all. It just means that you need to be able to convince other engineers that the project is worth the time being spent on it.
How do you make sure what’s being done is actually productive?
That’s one of the problems. It’s a huge amount of time and money, our engineering force is 200 strong now, and 20% of their time is a huge amount of money every year, so you do want something out of it. But you want to give them scope to innovate.
We do try and have technical leads giving feedback on their project. We’re trying to do better in this area, having people identify something after a couple of days’ work and saying, “I don’t think this is going anywhere”. Otherwise they spend more time on it and it’s wasted.
Is that a hard balance to strike?
You do want to give them the opportunity to push on with it, and if they’ve got the passion for the idea and they have the drive, it’s good. It’s a tricky balance.
How often do these projects actually make it to market?
Quite a lot, actually. Often times they don’t even make it to us before they ship – what’s interesting is when Google talks about these huge projects coming out of 20% time, sometimes people look on this experiment as producing big, loftier goals.
But often it’s quite small, and engineers are looking at small pieces of an application but they’re the only one passionate about that particular aspect of it.
If you’re a growing business wanting to innovate, is 20% time something you should consider?
It’s hard. You do want to strike a good balance, between getting the amount of productivity there and giving people room to move.
We try and use this innovation in every department we have, but you have to tailor it a little for the job they’re working around. It’s relatively easy to have an engineer do something creative like that, because it’s their job. But in a different department, like the customer service department, it may have to be different to drive innovation.
But all through the company, we do have a thread of giving employees that autonomous time, when they are in complete control of what they are doing.
Patrick Stafford is a freelance journalist and a former deputy editor of SmartCompany.
