The Lean Project

The following is an article I published on LinkedIn in February 2015.  Thought I would republish it here as my first post.  See the link after for the original.

I've recently had the opportunity to lead some strategic projects for the company I currently work for. These range from improvements in product lines to intricate cross-functional processes and my "customers" are all internal to the company. I've lead several of these types of projects in the past, mostly on a smaller scale, and have struggled a little with the sluggishness of these larger projects. In business school I was exposed to the concepts in Eric Ries' Silicon Valley bible The Lean Startup and recently had the chance to read it more thoroughly. It dawned on me that these same principles could be the key to turbocharging these projects. How did I implement this?

To start with, the customer's problem has to be well defined. As in the book, we need to make sure we are actually solving a problem, and not inventing one. Data should define the current state, not opinion or anecdote. The basic needs of the customer should be well understood, so that a minimum viable product or process might be developed. Here I differ from Ries a little. For this type of application, I feel the "perfect world" state of the customer should be understood and kept in mind as well.

The next step is to create a vision for the solution, not the solution itself. It's important to avoid anchoring here. The team should brainstorm possible solutions to meet the minimum viable product concept. There are volumes of literature on this process; I won't expand on this step here. The goal is to narrow down to the one or two least risk or highest probability of success alternatives that meet the customer needs. If there's some difficulty in selecting, try to retain options which are more scalable or extensible, to hopefully get close to the "perfect world" result in the end.

Build an initial prototype of the product or process and test it in a simulation. This is where the concept of validated learning comes in. The prototype could be a quick CAD model, physical model, or computer simulation. Create actionable metrics around the customers’ needs, hopefully the same metrics that were used to measure the current state. Measure the results and get feedback from the internal customer. If there are multiple viable options, an A/B test might be applicable here. The goal here is to be fast and flexible. Iterate on the options and continue to measure feedback and results. This is the real benefit of this approach. The goal isn't to think it to death from the beginning and try to hit the mark the first time. This rarely works and wastes scarce resources. Results come much sooner, and if you need to pivot, at least you haven't wasted most of your budget in the meantime.

Another important concept from The Lean startup is that of innovation accounting. Similar to the way a startup can use this accounting to show progress to investors, a project manager could use these iterations and measurements to show progress on the project. IR&D funding can be a difficult thing to secure, and even more difficult to hold on to. Testing early validates R&D spending and shows direct evidence of innovation progress. Instead of a status report that says "we've completed this phase" or "we're 50% complete with the project", you can actually show that we "hypothesized this, tested it, achieved these results, learned this, incorporated that into the plan, and are moving on to this". Innovation accounting also helps to retain design intent for future reference.

Once the minimum needs are met, you can begin adding features and enhancements to get as close as you can to the "perfect world" solution. I've found this process has added structure and agility to some of my projects. Obviously, it's not going to be applicable to every case, but I think it's a great way to frame a project at the start. It also helps to keep the team engaged and not get too bogged down in the minute details early. I'd love to hear any comments on this process or other approaches you've tried.

Original post: https://www.linkedin.com/pulse/lean-project-john-cino?trk=prof-post