Build, Measure, Learn – in Startups, Healthcare, and Careers
One reason I love applying Lean thinking in new arenas as my career progresses is that each shift in industry leads to new challenges and new experiences, which forces you to challenge your thinking… and each change in Lean direction reinforces how the fundamental principles are consistent and pretty universal.
When I got involved with healthcare in 2005, after 10 years in manufacturing, I knew this would be a great learning adventure. After learning about “Lean Startups” at talk by Eric Ries in late 2009, I found another new way to stretch my Lean thinking (leading to my involvement with a startup, KaiNexus)… but it ends up reinforcing the same classic lessons.
A core part of Lean is the iterative problem solving model (from Shewhart and Deming) of PDCA or PDSA. In this approach, we Plan a change (which includes studying the current state), we “Do” in terms of a small test of change, we “Check” or “Study” the impact of the change, and we “Act” or “Adjust” based on an honest assessment of that experiment. In the PDSA mindset, there's a chance that a change (something different) isn't necessarily an improvement (something better). If things aren't better, we can go back and try again, focusing on what we learned from the effort rather than punishing somebody for failing.
Ries adapts this to what he calls a “Build, Measure, Learn” cycle, as I took a picture of at MIT:
This is a highly iterative approach to software development and, more importantly, a highly iterative approach to building a business model and a company.
The traditional approach to a startup includes the upfront step of writing a big huge business plan document that takes a long time to write… and is often followed by an inflexible execution of that plan. Following the business plan is more like a “top down” or “command and control” approach to building a company.
Lean Startups start by understanding customer needs, releasing a “Minimum Viable Product” relatively quickly and getting feedback from the market rather than working in secrecy for a few years and then releasing a “big batch” version 1.0 that may or may not meet the market's needs. In the Lean Startup approach, you get to market early (with something that provides real value) and then iterate… it's a much more flexible approach and more likely to be successful (or so it seems).
Having worked with a lot of hospital department redesigns over the past six years, there's been a similar revolution in hospital design and construction – the “Lean” approach. Instead of the architect giving a “best practice” design to the department or hospital, the people working in the department develop their own layout that fits their patients' needs and their needs as employees.
Instead of a “big batch” of design, Lean healthcare design is a much more iterative model.
Lean Startups call for “Build, Measure, Learn” – it's the exact same approach in healthcare. Instead of “code,” we actually have something physical to build.
But, wait, isn't software much more flexible than a hospital building? Isn't software easier to change?
Well, of course — but that's why our first “measure, learn” cycle isn't done after we have already built the new department or hospital.
We “build” and test the new design in small, iterative cycles. Much as Eric Ries advocates, the goal is fast cycles of learning, not just moving fast.
The hospital design process might follow these cycles:
- Build (create paper layout drawings or “paper doll” models and concepts)
- Measure (simulate the different flows, including patient flow, provider flow, and material flow to see what seems to work best based on dimensions like time and productivity)
- Learn (make adjustments and choose a “best model”)
- Build (do cardboard mockups, as shown in the book about Lean at Seattle Children's, and try them out)
- Measure (test the layouts and make improvements based on working in the simulated space)
- Learn (make adjustments and start building for real)
This process might take longer and might be more time intensive, but the resulting “Lean construction” will go more quickly and will have fewer change orders and less rework. It's a new example of the Lean notion of “go slow to go fast,” where Toyota would take more time upfront and would initially ramp up production more slowly… and then BOOM, they'd ramp up faster and higher than their competitors.
There's a lot of learning to be done before the real space has been built. The last time I visited ThedaCare and saw their newest patient tower, they emphasized the little details that were injected into the design based on their prototyping… little details that generally delight patients, things that might not have been noticed until it was “too late” (or more expensive) to make changes in the design. With space design or software, you can't plan it all perfectly up front. You just can't anticipate everything, so you have to be flexible.
Lean design hospitals are being built smaller (because of process improvements and improved utilization of beds and space). Lean design hospitals are being built more quickly and at a lower cost.
“Build, measure, learn” continues AFTER the new space has opened, of course, through the Kaizen and continuous improvement concepts.
As with software, changes to the building or rooms are generally cheaper to make earlier as opposed to later.
And speaking of fast ramp ups, the Lean Startup approach is “go slow to go fast,” as well – where a company takes more time to develop it's business model and sales process before, BOOM, you ramp up faster (sorry for the booms, I'm channeling John Madden, apparently).
Lean factory design = lean startups = lean hospital design. What's different is really pretty much the same.
Also the same is the tendency for Lean talk to often not meet up with Lean hype, where you have “Lean Startups” building CRAP product (instead of a Minimum VIABLE Product) and you have some healthcare architects also latching onto Lean as a buzzword without a whole lot of substance. “Real Lean” will beat “Fake Lean” (as Bob Emiliani calls it) in any industry.
One other place you have to be adaptable and iterative – your own career, as this piece by Thomas L. Friedman points out: “The Startup of You.”
From the piece:
To begin with, [LinkedIn founder Reid] Hoffman says, that means ditching a grand life plan. Entrepreneurs don't write a 100-page business plan and execute it one time; they're always experimenting and adapting based on what they learn.
It also means using your network to pull in information and intelligence about where the growth opportunities are â€” and then investing in yourself to build skills that will allow you to take advantage of those opportunities. Hoffman adds: “You can't just say, ‘I have a college degree, I have a right to a job, now someone else should figure out how to hire and train me.' ” You have to know which industries are working and what is happening inside them and then “find a way to add value in a way no one else can. For entrepreneurs it's differentiate or die â€” that now goes for all of us.”
Finally, you have to strengthen the muscles of resilience. “You may have seen the news that [the] online radio service Pandora went public the other week,” Hoffman said. “What's lesser known is that in the early days [the founder] pitched his idea more than 300 times to V.C.'s with no luck.”
My original “grand life plan” was to work my way up to be a COO or senior leader in a big manufacturing company like a General Motors or Intel. So much for that. I started taking detours in my career when I joined my first startup, Factory Logic, in late 2000. I've had moments that required resilience, like being laid off by that startup in 2002, to be rehired a few months later and then moving on to better things, including healthcare.