Starting Backwards

September 13, 2014


If you were to draw a diagram of the entire product development process from start to finish, what would you start with? If you are like me, you’d probably start with the customer, or maybe sales. Then you’d probably pass the idea along to product management and onward to development. Last, but not least, the product would make its way to operations for deployment in production.

It’s pretty straightforward, front to back, end-to-end. Everybody knows how it works. And if we are going to try to improve this process, where do you think we typically start?

At the front? In sales? No.

At the back? With operations? No. Not there either.

Right smack dab in the middle? BINGO! Development always gets the love first.

Now what kind of lunacy is that? Now I’ve been part of a process improvement effort or two, so naturally I start to think I see patterns. Well, hallucinations of some kind anyway. Almost every time we start with the development teams. We do a good job, we get them up and sprinting, train a few scrum masters, console a few managers, and Bob’s your uncle: the dev teams are agile! Then what happens?

Downstream, the operations teams aren’t on board with the whole agile thing. They aren’t going to let you change their release processes just to satisfy some fad. Rapid change? Are you nuts? And what about the other end of the value stream, sales? They’re willing enough if it makes them money, but you’d better deliver (which isn’t happening with the operations guys, so you are screwed).

So lets take a step back and look at the value stream. Where do we typically see the most time spent? It’s not development – we’ve been squeezing development in one way or another for decades. However, you can find the most amazing queues in sales. With the rest of the organization moving so slowly, they naturally develop queues as they wait for the work to get done.

The question is, why don’t we start at the back? Why don’t we make the end of the value stream our focus first? We need to stop starting in the middle. Goldratt would have us chasing the bottlenecks. More power to him. If I speed up operations first, I may not see an immediate increase in productivity, but I have created the runway for success. I have them on board and bought in. Now, we move up the value stream to the Development teams. If we can get them performing, then we have already prepared the runway for them. No longer do we give them a fast car and ask them to drive it into the wall. Now they can deliver and they can do it with proper support from operations.

From there we can continue to move up to Sales and the front end of the value stream. They should be an easy sell at this point. So, the question is, why start in the middle?