Reader Question: Asset Optimizers and Point Optimizers
It's time for another in my Reader Questions series, this question comes from Brian via email. If you have a question or topic that you'd like me to address, you can contact me via the web or by leaving a voice mail.
Today's question looks for clarifications about the terms “asset optimizers” and “point optimizers.” Lean thinkers should be neither of these two things.
I have a question that you can freely use in a blog post. No rush in answering.
Can you help me understand what James Womack was referring to in his presentation, Slide 31 “Healthcare managers are asset optimizers, doctors are point optimizers” (from 1st GLOBAL LEAN HEALTHCARE SUMMIT – June 2007 Managing a Lean Organisation – PDF link)
I think asset optimation is ensuring all assets are used as effectively as possible. In other words, managers do not want staff, equipment, and such to be underutilized.
But point optimizers? Is that someone who improves only their area but but sub-optimizes the rest of the system? In other words, someone that makes their world efficent but maybe adds waste to the downstream?
I asked a few of my peers and did a google search but couldn't quite determine what he meant by this!
It's often hard to get the whole story and context from presentation slides, isn't it? I'll try to fill in the gaps, not speaking on Jim's behalf, but talking to the slides as if it were my presentation.
Yes, you have the gist of “asset optimizers.” It's a very traditional management viewpoint to want to keep resources busy, with the goal being 100% busy. This was traditional management in the manufacturing world (keep machines busy, keep people busy) and it's also often the mental model in healthcare (keep beds full, keep doctors or MRI machines busy).
The dysfunction with that, of course, is when you want resources to be 100% utilized, then flow really really suffers (a factory has a lot of “work in process” inventory build up or patients wait a long time in a hospital or doctor's office). Waiting times often get really long when resources are even just 75 or 85% utilized.
This idea is pretty core Industrial Engineering course content, if you study “queuing theory” (read more about queuing theory on the Shmula blog) or there's an outstanding operations text book called Factory Physics that teaches this stuff brilliantly. You listen to my podcast with one of the co-authors, Mark Spearman PhD, here.
The exact level of utilization at which you get this so-called “WIP explosion” depends on the amount of variation in a system – and healthcare settings tend to be high variation systems, so we have to be careful about balancing utilization and flow.
Really low utilization is bad – you get no throughput or flow. But really high utilization can be bad too – it harms flow.
When I did some work with a radiology department, their starting point MRI machine utilization was about 40% (and this was calculated based on a denominator of just 10 working hours per weekday).
To get the patient backlog down, we had to increase utilization. But I knew, intuitively, that we couldn't aim for much more than 70% MRI machine utilization or else patients would be potentially waiting a long time in the waiting room if there were any hiccups in the system.
With process improvement, the team was able to get the utilization up to 70%, thanks to better scheduling processes, better communication with patients, and better end-t0-end value streams. Asset optimization is not the only goal – you also need good flow. Again, you have to find the balance point, the “sweet spot” if you will.
Yep, point optimizers look at just their work, their department, their sub-department, etc. They often suboptimize the overall system is this is where value stream mapping (and, better, value stream thinking) is so beneficial – we are able to optimize or improve the entire system.
We saw a lot “point optimization” in the pre-Lean condition in that radiology department. It wasn't really “point optimization” as much as it was “point focus.” It's not that anyone involved was a bad person or a selfish person, but nobody understood the whole system or the whole value stream. People couldn't see (there's a good reason the core LEI value stream mapping book is called Learning to See: Value Stream Mapping to Add Value and Eliminate MUDA) and people weren't communicating, so they couldn't improve.
Each “point” (or you could call it a silo) was trying to optimize its own piece of the system, including:
- Scheduling office
- Front desk / registration
- Anesthesiologists (this was a children's hospital, so this was very much required)
- Sedation room
- MRI techs
- Respiratory therapists
- Recovery room
The key to Lean improvements was to get a cross-functional team to look at the WHOLE system. They were starting the first case, on average, an hour late every day.
When we pieced together the entire process, all of the stakeholders and their work, we found that the system was perfectly designed to start an hour late.
As Dr. Deming's students would recognize:
Every system is designed to get the results it gets.
So we had to design a NEW system that could actually start on-time. This wasn't really that difficult once you helped everyone see how all the pieces of their work fit together. It wasn't anybody's fault — the system had just evolved. The good news is that it could be fixed.
These are powerful Lean thinking lessons, I think — don't be a “point optimizer” or an “asset optimizer.” Have you been able to change those mindsets in your organization??