How organisation size crushes innovation - and what to do about it!

Remember working for that start-up? Things were good. You responded quickly to change. You could, and often did, roll out new programs within weeks or days. Your boss approved quick changes with a simple nod. And you got results - fast.   

Then you switched your start-up gig for an important role at a big enterprise. Things were different - slower, costlier, stuck in red tape, less tangible, less experimental. That's because big organisations are complex. And when humans get accosted by complexity, we get anxious. We need certainty and coordination - in the form of structures, policies, responsibilities and rules - to push that fear away.

This need applies to organisations both big and small. In a start-up, the boss (usually the founder) knows more or less what's going on in all areas. So things are still complex, but not threateningly so. But if you're a VP in a publicly-listed vertically-integrated multinational, things are likely to be different.  There are a lot of roles with a lot of complexity that all need coordination.

We can't change our fear of complexity, nor our desire for control. So, what can we do to keep our organisations agile -  even as they grow?  How can we ensure innovation doesn't get crushed?

1. 'Practical Guidelines'. Not policies

Growing organisations typically try to align their people though policies and procedures. A set of rules provides consistency and, so the theory goes, coherence throughout the organisation.

We shouldn't try to fight the natural tendency for coordination and control. As a matter of fact, few people will argue against the benefits of introducing consistency. But there's a better way than rigid rules.

At Atlassian, we've grown from 100 to 700 staff in the last four years. We try to limit hard rules - we have guidelines instead. What's the difference? Take, for instance, our Atlassian Design Guidelines, which is an online resource with design components and principles. It provides guidance on how to build awesome Atlassian products and add-ons. The tool throws in JavaScript for layouts, menus, and headers so that our developers (and our third-party allies), can develop new features themselves.

The best bit is that the tool doesn't interfere with the developers' design process, rather it helps them work faster (just cut and paste code) so they can experiment freely while still retaining a base level of consistency. Unlike rules - which say "you must do this" - guidelines will give your people latitude to get the job done better and faster.

2. Ritualise Autonomy

Our extreme openness invigorates projects and produces better outcomes - an engineer recently implemented a new Facebook style @mention feature, days after reading a customer feedback blogpost from our Product Management team. A recent change to our induction program wasn't just driven by HR, our developers also pitched in their ideas.  

4. Give customers solutions, not perfection

The most daring - and arguably frightening - way to respond faster is to give employees easier access to customers. This gets your solutions into the real world faster, and lets you improve iteratively based on the feedback you've just received. Feedback loops get tightened.

At Atlassian, we introduced a "growth hacking team". The team's job is to release smaller experimental features to customers often (since inception, they've launched countless tests) and analyse the billion of lines of data that come back every week. They're the A/B-testing, analytics-breathing front-line troopers who put our ideas into practice, see what the market says, then use it to make our stuff better.

Every quarter, Atlassian goes through an exciting ritual called ShipIt Day. The organisation halts at noon on a Thursday and every participating team starts working on an innovative pet-project of their choice. 24 hours later, teams present their idea to the rest of the organisation and everyone votes for a winner. Over the course of six years, a stunning 700 ideas have improved Atlassian's products and made their way into production. Also, a few totally new products were developed as a result.

Giving people autonomy - to pursue their own projects and interests - encourages them to anticipate new ways of doing things and test their ideas without repercussions. But when should you take your innovation time if you have a list of tasks to do?

ShipIt works because rituals or routine give us certainty - which, as we realised before, is the antidote to the fear of the complex. But it does so in an exciting, playful way. And the results have a tangible impact on the "serious business" - the new products and features end up either being shipped or used internally.

Not only engineers do ShipIt days. The concept has been successfully implemented in marketing companies, hospitals, even high schoolers!

3. Feed our Social Appetite

Few ideas are created by one brilliant mind. Most great ideas are the product of many great minds. Thankfully, as humans, we're hungry for social interactions. We also crave access to information to help us fit the pieces together. For larger organisations to adapt quickly and power innovation, the easiest way is to encourage those natural needs. 

Email stifles transparent dialogues and discussions. And disjointed software systems limit access to the information people yearn for (like sales data). 

That's why innovative organisations increasingly use social collaboration platforms (like Confluence, Jive or others), Group Instant Messaging tools (like Hipchat), or social coding tools for technical teams (like Stash, Bitbucket or Github) to make this happen. Such tools help us discover ideas, display data, discuss projects, update teams...in short, do everything we need to do. Together! 

Social collaboration lets us crowdsource content and fine-tune the way we run our company. At Atlassian, instead of emailing the yearly company strategy plan to staff, we openly discuss its content on our social collaboration platform Confluence. I love it when a super smart graduate openly disagrees with our CEO. And - as always with crowdsourced ideas - the best ones survive, regardless of who offered them. 

By introducing this growth hacking team - and their focus on getting solutions in the hands of customers faster - we made "experimentation" and "growth-hacking" part of our company's common vocabulary. It has become a natural aspect of how we think and work across all teams. 

At Atlassian, we introduced a "growth hacking team". The team's job is to release smaller experimental features to customers often (since inception, they've launched countless tests) and analyse the billion of lines of data that come back every week. They're the A/B-testing, analytics-breathing front-line troopers who put our ideas into practice, see what the market says, then use it to make our stuff better.

Anyone, not just engineers, can experiment and iterate with customer solutions. I believe every team should think outside the box, disregard status quo, and discover new ways to solve problems. Most people don't like to endlessly work on tasks where they can't see the results of their labour. Growth hacking makes work interesting, solutions more daring and innovative, and teams more agile.

Larger organisations especially need to be reminded to get things out and in the hands of customers quicker. They need to keep moving fast and iterate often. It feeds innovation!