I was recently inspired by Henrik Kniberg but needed an image of my own that I could split up and move around to make my points in the Zoncon Keynote this year. Yup, my PowerPoint drawing skills suck or so said the lady sitting next to me on the flight to Seattle. So I asked her, what would make this picture better in your mind? She replied that she was also no artist but the colors weren’t very appealing. For the next 20 minutes, we worked on it together and made it better but it still wasn’t much to write home about. We further explored why the picture sucked and then realized that while it wasn’t all that good, the message behind it was powerful.
Failing fast is the essence of true innovation. Discovering why something works or doesn’t work and then learning from it helps to solve problems faster and better. During my ride to Zoncon, I learned that colors matter, size and shape help distinguish importance, placement could destroy understanding, and finally that it’s easy to criticize but hard to help.
Upon further reflection, everywhere I go these days, I hear people talk about failing faster. It’s an empowering concept if done right. And yes, failing fast has the ability to transform an organization. Yet I often find myself cringing at the message because while the word is getting out, the way its being framed makes it hard to grasp the concept of applying science to the art of meeting customer demands. Failing is different than absolute failure. It is important to distinguish between absolute failure and sending the message that experimentation is “ok".
I’m fortunate to work for Intuit where I get the opportunity to learn and create amazing teams. Here I’ve learned a lot about accelerating learning. Failing fast is mostly about applying a continuous scientific approach to the work we do. It’s about creating a hypothesis and testing the factors that influence a solution. It’s about testing whether you have a good enough solution to invest in. During the creative process, its about applying science to art so that the best innovations can be developed. And finally, failing fast is about bringing about the power of collaboration.
This past year, we’ve had the fortune to work with Tom Chi, http://www.tomchi.com. Tom is a great example of a leader that truly understands why failing fast and creating continuous science is important. His efforts on Google Glass and stories about how he and his team worked at Google are inspiring. He provides a set of visceral examples about applying continuous science to everyday problems to understand what solutions work and don’t work without over-investment.
Likewise, I attended the Rugged DevOps day at RSA this year and had the fortune to listen to Dan Glass talk about the R.O.A.D to Rugged DevOps for a Major Airline. I thought the most brilliant thing he said was that none of us in the audience would want the Airline industry to complete continuous delivery of software to an airplane while it was in the air flying. I tend to agree and reflected to realize that the airline industry has applied experimentation to their environment and that they have passed down lessons in a much better way than most other organizations. In fact, they have largely adopted accelerated learning and matured it in their industry so that it is baked into the way everyone works. And Gary McGraw also made somewhat of the same point about SCADA systems in his discussion during the DevOps Throw Down at RSA, albeit much more animated. In these cases, science has been applied for years and continuous improvement moves at a prudent pace balanced by the safety required.
What I have taken away in all of this input is that science matters. Failing fast comes from developing a hypothesis and testing it in service to success and innovation. When we talk about DevOps, Lean, Agile, Scrum or other; in essence, we are talking about testing something before investing too heavily in the invented solution. We are developing a step based approach towards solving a problem and we are learning our way into creating a solution that not only solves the problem but is worth investing in. Organizations have many constraints that need to be understood and customers which vary. But the approach taken is one of experimentation and collaboration. And yet, to achieve these outcomes, applying science and experimentation within an organization needs many things to succeed; here’s the top five so far:
1) Blameless Culture
Blaming someone for failing is the biggest mistake that executives make when promoting a culture of innovation. Science requires experimentation and testing which is best supported by a blameless culture dedicated towards furthering understanding and collaboration. Imagine for a moment a big transformation, perhaps the integration of Security and DevOps. Security has typically been a detailed process that comes at the end of the Software Supply Chain and looks somewhat like a point inspection for a car coming off an assembly line. To migrate this capability earlier in the supply chain, you’ll need to conduct lots of experiments with environment change, tasks, outcomes, goals, processes, tools and mindsets. An environment that focuses on understanding has the ability to allow work to be shifted in whatever direction is required so long as commitments can still be made. Executives can make an initiative successful by understanding experimentation and creating the framework that makes it possible to remove blame from the system.
2) Good Coaches
This leads me to my second ingredient for a successful science based approach, good coaches. Good coaches have the ability to help executives create a blameless culture and drive work so that it progresses and doesn’t get slowed down. A good coach understands the work required, works out their understanding of what it will take to get the job done, and then works within the team to produce the lessons and skills needed to progress rapidly on solving a problem. A good coach is patient, resolute in their role, and passionate about inspiring the passion necessary to get the job done. They are observers, role models, teachers, and students rolled up into a great leader. Often they are also managers, sponsors, and the tie breakers. But most importantly, in an environment of uncertainty they are the brave souls that help to define guardrails and defend the space necessary to achieve speed and scale in a continuously improving environment. They form the right relationships to get the job done and they are boundary-less in their approach. They are an inspiration.
Inspiration is visceral. It’s the magic that often times helps spark imagination, provide insights to a problem and create the passion necessary to invest in finding an answer to a really hard problem. Organizations that embrace a leader-teacher culture tend to create more inspiration and achieve better results. The leader-teacher culture provides opportunities for everyone to participate in problem solving and reduces politics. Problems get solved faster. Teams are happier. Customers demands get met with less investment. Inspiration can make you do and achieve things you never thought possible. It can lead to faster understanding and greater collaboration. And ultimately, it leads to sharing and collaboration.
Capturing lessons is critical; sharing them is priceless. I often tell teams that if something feels slow, you’re likely doing things wrong. Iteration should feel like an exponential increase in time that helps an organization go from a paper idea to a maturing product. Your first experiment should help you to learn within minutes whether an idea should be given your time and whether it solves the right problem. Gathering feedback is essential towards observing a problem from a different perspective. Listening to someone else tell you whether you nailed it or flubbed entirely is really eye-opening and can provide the information needed to refine and make a solution better. A fast feedback loop is critical and building it into your framework can accelerate learning.
5) Flexible Operating Mechanisms
Finding a way to operate effectively to deliver against timelines and make commitments is also quite important. Companies that already employ science in the development and delivery of their products know well that it is difficult to create commitments until some of the initial work has been done. Time-boxing can work to help with controlling timelines and also make it possible to gain benefit from innovating. Having weekly operating mechanisms that help create a fast feedback loop for everyone involved can also increase communications and reduce surprise. We use weekly videos to keep each other up to date and highlight problems worth solving. Demos and reviews provide can be used to share information, gain feedback and provide additional inspiration within a team environment. Ultimately, creating flexible operating mechanisms that help with communications is dependent on the culture, people and work being done.
I’m sure you have more ingredients for applying experimentation and collaboration in your environment. If your organization is seeking big changes and impactful product innovation, there are many ways to approach it. Allowing teams to experiment and promoting an accelerated learning environment goes a long way towards engagement. Creating a culture that allows for rugged software and products to be built is in essence why innovation is accelerating.
Isn’t it time to embrace continuous science at your organization and learn how to fail fast? It would be great to hear your story.