Process Improvement: How?

I have an apology to make to the 15.000+ students I have taught over the last 15 years since becoming a professor: my answer to your question about process improvement consulting has been woefully inadequate.

Let me explain.

Lazy Heuristics are fast but flawed

One of the key issues that I address when teaching concerns the application of theory to problems. I do this since managers are inevitably confronted with problems which require immediate action. Thus the tendency is to jump quickly from problem to solution. In some cases, this makes sense; often a quick if incomplete solution is better than a solution that takes weeks or months to work out. But if we forget to challenge the (lazy) heuristics that experience gives us we risk attempting to use our understanding of yesterday’s issues to resolve today’s problems.

When I teach, I inculcate my students with the importance of using theory to situate their analysis. It is only by doing this that one can come up with rich, well thought through answers. Indeed, nothing irritates me more than watching rabbit-like conclusions appearing out of magicians’ hats that lack any congruence with the analysis undertaken.

The same is true when I work as a business process improvement consultant. When confronted with a

• problem it is necessary to apply operations and process theory to comprehensively
• diagnose a situation before
• analysis is undertaken. It is only then that you can
• conclude what the issues are, and then
• solutions to the problem should be obvious.

Getting my students to avoid jumping from problem to solution is the biggest challenge I face as a professor.

So, the question I have been failing to answer adequately for the last 15 years?

Which theory or framework should I use to maximise gain?

I think there are two main reasons for my poor response to a perfectly reasonable question.

Firstly, there are existing approaches to structured problem solving which “more or less” work. When asked the quesiton above, I’ve shoehorned my answer into the:

• PDCA (plan, do, check, act) cycle or the
• DMAIC (define, measure, analyse, improve, control) cycle

Simple and incomplete answer: Use these and your world will be a happy place!

Secondly, the appropriate OM theory is ‘so obvious’ to me that the choices I make are unconscious. I see patterns in operations, whether I want to or not. For example, when I was out dining with my beautiful girlfriend Desirée a while back, I found myself looking at the layout and flow of the restaurant. Not the most romantic thing in the world to do, but the product of an improvement mindset that I can’t turn off. It was obvious to me their operation could be made more efficient before we'd been served our entrée.

My thinking was challenged over the weekend I read an article on the OODA loop (observe, orient, decide, act), first postulated by the USAF colonel John Boyd. I am generally against cutting-and-pasting military-inspired techniques into the civilian environment; having spent two years working as a soldier, I know first-hand that the world of commerce is a fundamentally different ‘battlefield’.

But it did make me think about the PDCA and DMAIC cycles – they are rich in how to solve problems, but poor in how to select the best technique to do so. By contrast, the OODA framework is rich at the front end – spend time observing and orienting yourself, but poor on explaining how decisions are made.

It led me to conclude that both the operations-oriented PDCA and DMAIC frameworks and the military-inspired OODA framework are necessary but not sufficient to do the job of a business consultant or a manager. Some further research allowed me to find a McKinsey article on the benefits of applying OODA in data analytics, although conclusions such as ‘make your output usable - and beautiful’ appear rather rabbit-like.

So I decided to combine the two myself – and as I postulated I picked up a felt-tip pen and drew my integrated framework for solving operations challenges. I realised that this was exactly what I had been doing for years, and it is exactly what I want my students to do.

Integrated Problem Solving

Remember I said that my operations brains never switches off? Well, I came up with this over the weekend in-between circus school and pizza-making with my kids. I drew this in my daughter’s sketch book with the aforementioned felt-tip pen since they were the only writing implements to hand. The first two steps I’ve called ‘diagnosis’, and I've called steps 3-5 ‘doing’.

Step 1: Observe and Define

This is an iterative process at the beginning of a piece of improvement work. It involves asking questions (why? why? analysis is a good place to start), getting to know the landscape that your client inhabits, probing, and listening. During this phase it is essential that you spent time understanding how you can create value. Often you identify issues that your client didn’t even know they had.

Step 2: Frame

At this point, it should start to become apparent what the key issues are; it is only then that you can start to frame them using operations theories. Here you’re starting to pin down the appropriate theoretical frameworks that could allow you to find solutions. Is it a problem where theory of constraints can help? Or a kaizen blitz? Or an understanding of the effect of batching? Or servqual? Or the importance/performance matrix? There are many, many theoretical constructs that can be applied to problems – the challenge is identifying the most relevant and helpful.

This is why I never apply a ‘formula’ to a problem; the old adage; “if your only tool is a hammer, everything looks like a nail” sings to me on this point. I am very against shrink-wrapped formulaic problem solving for this reason. I am highly suspicious of consulting companies that believe that they have a one-size-fits-all approach. They don’t.

Step 3: Measure / Gather

I’ve taken the ‘M’ from DMAIC and added the idea of ‘gathering’. Often companies have dusty warehouses full of underutilised data and thus gathering relevant data is often quicker and more appropriate than creating new data through measurement.

Step 4: Analyse

This step involves using the frameworks identified in step 2 to analyse the data from step 3. It goes without saying that step 3 only works if you’ve nailed the data requirements. What you’re doing here is taking the ‘hunches’ that you’ve created in the diagnostic phase and using data to (dis)prove them. This might mean that you loop back into the daignositcfs phase, if your original ideas prove inaccurate. This is a good thing; you’re learning-by-doing.

Step 5: Improve

In this step you implement the changes that have been identified from the previous work; if you show that x leads to y, then doing x is the obvious next step. So, get on and do it! You then loop back to step 1 if you believe in continuous improvement. Which you should.

 Does it work?

Steps one and two take time – this is where you are planning/defining your problem; it’s difficult yet important to get right. I worked recently with a UK FTSE-listed company in their Macedonia factory. They were convinced that their operation was world class and were about to invest heavily in new equipment to increase capacity. It was only after very careful diagnosis that it became clear that their understanding of OEE was flawed. When we ‘did the maths’ it became obvious that they were measuring the wrong things. The calculation was the easy bit – my eight-year-old daughter can multiply 3 numbers together. The diagnostic element that determined the relevance of this method took significant time.

Similarly, it was only through the selection of business process mapping that it because apparent why there were delays in treating patients at an A&E department (ER room, if you prefer American English). The map showed that 63 of the 84 steps in the process were non-value add. This diagnosis and analysis meant that the solution was obvious.

Reflecting on my professional experience of working directly with clients, this approach to problem solving is a good one. And when developing the intellectual capabilities of my students, this is the approach that I'm going to start teaching.

I just hope my little girl isn't too unhappy that I took a page of her sketch book to work it out.