Why are DOE fans so passionate? There aren’t many evangelists for statistical process control after all. A sceptic once told me, “It’s like a religion for DOE people like you.” True, those of us who have seen the light preach the message with great zeal. We love telling people that one factor at a time does not work.

DOE is truly transformative.

You may have heard how DOE can double the speed of R&D or halve the costs. That’s also true, but you can’t fully capture the value in numbers. Instead, I’ve found a simple way to connect the message with people at different levels of organizations: predictability, productivity, and professional advancement. Later I’ll also let you into a secret about DOE.

But first, the three P’s of DOE that everyone needs to know:

Predictability: Delivering on promises

At an organizational level, what matters most is predictability. In 2018, I met with CEOs and MDs of companies in the North East Process Industries Cluster on a tour of Newcastle and Teesside in England. I was nervous about talking to people at this level of an organization. What would they want to know? Surely they would want to know they could reduce R&D costs with DOE, so I was shocked to hear that they didn’t really care. Turns out, it’s not that difficult or expensive to hire more Ph.D. scientists. What they really cared about was predictability.

Unpredictable R&D timelines are a killer. It’s not good enough for the answer to be always just around the corner. Businesses need to make timeline commitments to customers with confidence. Not delivering projects on time is hugely costly in delayed revenue and damaged reputation. Also, it’s cheaper to spend effort in R&D to fully understand and mitigate against potential problems than it is to have them appear in manufacturing, where the costs are now orders of magnitude higher.

DOE provides a rational, structured alternative to the haphazard trial-and-error approach. Results are unambiguous, enabling faster, more reliable decision making. DOE de-risks R&D and provides the predictability to confidently deliver on promises.

Productivity: Knowing you are investing your time wisely

Now a story from even further back in my career to illustrate that, for managers and team leaders in R&D, DOE is all about productivity.

After my chemistry Ph.D., I went into industry and fairly quickly found myself leading a team in formulation development. What an experience! Every day, I had to decide what each experimenter on the team needed to do. We’d quickly review the incomplete results from the last attempt and decide where to go next to keep up the momentum and ensure no person or machine was idle. We didn’t get the chance to think beyond the next trial. It was easy to lose track of how much effort we had burned and how little concrete progress we had made. We had some successes, but it was relentless and overwhelming. I made mistakes and I struggled to keep my head above water.

DOE would have been a lifesaver had I known about it. If I could do it all again, I would have the whole team learn how easy it is to design and analyse powerful multifactor experiments. Together, we could design a plan of what we needed to do over the coming days and weeks that would ensure we were using our time, people, materials, and equipment where they would have the most impact. Each day, we would all see how our efforts contributed to solid, unambiguous results and tangible, incremental progress.

DOE is a project management tool to maximise team productivity.

Professional advancement: Becoming a better scientist

At a personal level, DOE means professional advancement. It makes you a better scientist or engineer.

At my farewell lunch for my last job before I joined JMP, one of my soon-to-be former colleagues shook my hand and thanked me for helping them with a recent DOE project. He told me that it had finally got them the promotion they had been seeking for years.

This experimenter had hypothesised that it should be possible to improve the efficiency of a purification step that had become a bottleneck in the manufacture of the latest product line. Together, we worked on a designed experiment to understand all the factors affecting the cycle time of this process step. We were able to show that we could optimise the settings of pH, temperature, and concentration to halve the washing time required to achieve the required high purity. Finding a way to double the kilograms per hour of a valuable product without having to invest in any additional hardware is a result that will get attention.

I’ve seen this a few times now. DOE provides compelling results that can get you noticed in your organisation. It’s not magic and does not guarantee you will find a solution – one of my earliest successes was a designed experiment that unambiguously showed I was not going to find an answer. That’s a story for another time.

DOE’s best kept secret

You may be surprised to learn that DOE is not that hard to master. That’s a secret I wish more people knew. Admittedly, it requires more than just learning a new tool. It’s a mindset shift – and there is a bit of mathematics and technical jargon that you need to understand – but anyone working in science and engineering will have had to struggle with much more difficult concepts in their education. (Would any chemists care to explain the Franck-Condon principle?) The lack of exposure to DOE in most people’s training is really disappointing, but it does mean that with a little knowledge, you can quickly become an expert amongst your peers.

Become a DOE evangelist

If you’ve experienced the transformative benefits of DOE and you want to convert others to the cause, then use the three P’s – predictability, productivity, professional advancement – to connect with people at different levels. You could be the person that makes DOE a strategic advantage for your organization.

If you’re new to DOE, I’ve created a downloadable case study based on the purification example mentioned above. Explore JMP’s Easy DOE platform for yourself – just download a free trial of JMP and open the provided data set. I’ve even created a short video to help you get started.