Agile Development Blog

This is #2 in a Series on the Culture of Innovation with guest blogger Lee Devin.

Failure and success are handy terms when we want to characterize closure in an industrial making process: We call a thing that works a “success,” and a thing that doesn’t work a “failure.” In an iterative collaboration that leads to an emergent result, they’re not so clear cut, not so handy. We come to closure on an iteration. When we test it, we find it doesn’t do some of what we need.

Thinking industrially, we say “This sucker won’t work. It’s a failure.”

But we need nuance here. It makes sense to observe that the thing failed the test. It makes non-sense to say that the thing’s a failure. As part of an iterative collaboration the current thing is a necessary part of a journey toward an innovation. Chances are pretty good that this iteration contains the seeds of the one that finally does the job.

We think of starting to build a new culture by reconceiving a couple of words because we believe that language is the key to our work; use of language is, after all, the fundamentally human action. Those old Greeks had this idea: they thought of language as a distinguishing feature of a human being. They knew creatures in the world who did not speak Greek, who made unintelligible sounds like “Bar, bar, bar, bar.” They called such creatures “barbarians.”

We won’t go that far, but we will suggest that language is our best tool for thinking and making choices, for knowledge work.

Before we make suggestions about how to address this difficulty, let’s revisit an important feature of any culture of innovation. The dominant way to make innovations is to run collaborative iterations. Get an idea of what you’d like to have; make one; test it and discuss it among the team and, if possible, with the end user; on the basis of this discussion, reconceive what you’d like to have and make new one; use everything you can from the previous iteration; chase new ideas to their end without predicting results; test and discuss; reconceive; make a new one; and so on until the project reaches closure. You recognize closure when anything you can think of to do makes the thing worse, not better.

Most of us are okay with the idea that the end product of an innovative process emerges from that process and is, finally, unpredictable. What we have not confronted is the idea that the words we use to think about processes and products may interfere with that and may need reconceiving.

Redefining ‘failure’ and ‘success’

If you can plan for and schedule a process, how new will the outcome be? Not very. But, how can you start on a project without identifying a goal, making a plan to reach that goal, and without confidence in your plan? We all know these keys to success, and unsuccessful equals failure. Right?

• Failure! “I’m no good. Better go out in the back yard and eat worms.”

• Failure! “Thank God, let’s drop this sucker and move on. Now that we’ve failed, and learned from our failure, the next idea will be a good one.”

Well, maybe not. Maybe instead of failing, you’ve taken an essential step along the way.

Maybe you haven’t reached an end point and suffered a defeat. Maybe you’ve moved toward an unpredictable closure. Some innovators, Tom Kelley of IDEO among them, believe that to succeed you must fail often. Works for him.

We think there’s a better way.

We can begin by noticing that our models for “failure” and “success” limit our work. This means that we need to make a cultural change, from industrial (plan, design, execute) ideas of making toward artful, innovative ones (prepare, iterate, test, iterate again). Think of a staircase. We don’t regard the first step as a “failure,” as “unsuccessful” because it doesn’t get to the second floor. It’s a step. One of many. Just so, we need to decide to put our attention on our process, conceiving it as a journey.

The change of mind here is: we can’t say, exactly, when we’ll complete the journey, or when we’ll arrive. In fact, sometimes (out of which times come the really good stuff) we’re not even sure where we’re going.  But what we can say is what we just learned and what we recommend that we do next.

At Rally and in many Agile teams, we use a notion from eXtreme Programing called spikes.  In a spike, the engineer sets out to learn what she does not know by conceiving of a simple test to prove or disprove a theory.  These spikes are used to help narrow an estimate, gather data on a continuum of choices or narrow a field of options.  By calling it a spike, the XP creators helped us RECONCEIVE the ideas of success and failure for a story and thus helped themselves and the rest of the team.

Suppose we decide to go further beyond just a small task like a spike and conceive of each iteration toward an emergent innovation as an essential step along the way. Suppose we decide to conceive success as a measure of progress, not closure. In our culture of innovation, this means we conceive product as a result, not a goal. We’ll know it when we get to it.

Here’s an idea that can help. “Nothing is lost, and wonders never cease.”

Artists live by this mantra: when the work reaches closure it contains everything done on the way. (You’ll see actors who don’t seem real. That’s because all they’ve done is learn their lines and blocking and a way to say the lines so that they’ll sound good. You’ll see actors who seem as if they are the character; you can’t believe they’re not personally like that. That’s because all they’ve done is spend hours and hours of thought and research into creating the given circumstances, a complete history, of that character.)

Instead of discarding work that didn’t reach a goal, reconceive the idea of “goal” into “result” and decide to use what you just made as material for the next iteration toward a result that you’ll recognize when you see it. The more of the current iteration you can find to use, the better. The harder you have to work to include everything, the better. In combination with the new ideas you (the team) get from discussion, and from the imaginative effort you spend, something unpredicted, something new, will appear in the next pass. This will happen.

The cultural principle here is: Collaborative iteration equals Innovation.

In this model, we can measure the progress of effort as it converges on the product.  What were the tests results with which stakeholders? What paths will we not follow any further? What potential design sets still need to be tested?

The failure of tests down a path is actually progress and a sign of innovation. Progress is a narrowing of options toward an optimal solution and failures are critical to narrowing.

By adopting the iterative process of Agile we increase the opportunity for innovations, but ultimately we need to prepare for improvisation by changing our idea about language. We need to use language; to decide what words mean. To use language, in other words, as a tool we control, not as a reality that traps us. And that’s a cultural change, not a tip we can quickly use.

We do have a tip, a simple (but not easy) way to begin this complex change. Never say no. Hang that on your wall next to the “No Sniveling” sign. “NEVER SAY NO.” This simple (but not easy) change cannot fail to increase the creative range of individuals, teams, and the organization. It’s not a final answer, but it’s a step along the way.

Next up in our series – Planning and Preparation

About the Authors: Ryan Martens is a goat cheese maker,  founding board member of the Entrepreneurs Foundation of Colorado, and Founder and CTO at Rally Software Development.

Lee Devin is a dramaturg at the People’s Light and Theatre, a Senior Research Scholar at Swarthmore College and a senior consultant with the Cutter Consortium. These activities often interfere with his fishing, and cause him to neglect his grandchildren.

Subscribe today to get free updates by email or RSS.