Taking Our Own Medicine

What Would A Process Improvement Function Look Like If It Drove Down Waste And Variation In It’s Own Way Of Working, Not Just In The Business Process?

As Process Improvement practitioners, or “the Lean Six Sigma types” as some of our clients might call us, we get hired to drive down waste and variation in the business. What if we hired ourselves, gave frank advice, and then listened to it? Can we then drive down cost of transformation and time to benefits realisation? Can we increase the certainty of benefits actually being realised? Can we identify and realise more benefits than otherwise? And last but not the least, can we ensure longer sustainability of the changes implemented?

In the first of a series of articles, we explore these questions and their implications. Our methodology tells us to start with looking at how to measure success and define value from the customer’s perspective. Accordingly, this first article focuses on what we would need to measure to tell if we are doing well or not. The performance of the Process Improvement (PI) function can be measured across various dimensions:

Effectiveness
How much improvement potential is identified and successfully delivered. Process Improvement begins with a problem or an idea, so a good way of measuring the effectiveness of the PI function is to see how many problems or ideas were identified and how many reaped the benefits through the solutions provided by the PI function. Thinking of the “idea funnel”, one can also think of rolling throughput yield of that funnel. Another perspective could be the effectiveness of planning and costing. A First Time Right measure can identify the proportion of projects that come in on time and within budget. On this dimension, many organisations do a fair job of tracking their ideas engine or funnel, but many don’t. As for first time right, it is not something that gets enough attention. The IT industry publishes figures on project performance in terms of on-time and on-budget, and incidentally ends up highlighting a clear need to improve. But as a function, Process Improvement does worse in our opinion, in that we don’t track and publish such data beyond individual pockets within enterprises. Taking our own medicine, it can therefore be reliably said that if we did gather this data across the industry, it would be unlikely to be flattering!

Efficiency
An obvious starting point is return on investment for individual projects, as an application of the very definition of efficiency which is output per unit input. This gets fair attention from management, as it should, and that’s a good start. Another perspective can be cost per project. A smaller number implies more granular projects, and hence less risk and more control. Practitioner resource utilisation and benefits realised per practitioner can provide insight into resourcing as well as governance efficiency. Given the scarcity of expertise, one can also track span for expert resources. Given the cost differential between internal and external resources, one could add cost of external expertise to the set of efficiency measures. Last but not the least, cost of tools and platforms used by practitioners is important as well.

Speed To Market
This can be measured as cycle time to reach certain key milestones in the life cycle of an improvement opportunity: time to identify, validate and prioritise an idea; time to generate actionable insights for improvement; time to implement the change; time to realise benefits and time to institutionalise the learnings. This is an area that has really not got the attention it deserves. Speed to market not only influences cost directly, but also influences returns by bring forward the benefits. And yet, there is very little evidence of systematic tracking of these cycle time measures across PI projects.

Reusability
This is a measure of leveraging insights, learnings and intellectual property from past projects; can be further broken into reuse of data collected, reuse of insights generated and reuse of solutions implemented. Some organisations have a good culture of encouraging reuse through incentives to share what you know and to ask others what you don’t. Likewise, some do a good job of capturing learnings into corporate memory and making it accessible. But by and large, this is an underdeveloped capability and leads to frequent reinvention of wheels. Apart from the obvious loss of efficiency, this also creates variation in outcomes, as the same situation is solved differently by different business units.

Sustainability
Theoretically, an improvement is sustained forever, or at least until a new disruptive change makes the older improvement irrelevant. In practice, improvements decay over time, which is why we need “continuous improvement”! To apply a measurement lens to this is difficult because of the protracted impact of decay. It is impossible to say at what time the improvements decayed down to a level as if the original change for the better had never happened. It is also impossible to analyse where the natural decay would have landed the process, had it never been improved. We want to recommend a different perspective to tackling sustainability, and that is to ask the question – “did we permanently change the way the Operator of this process thinks about their job”. If we did, the improvement impact will stay forever, or at least until the next disruptive wave. There are many tools and techniques to measure if Lean thinking, for example, has been absorbed into daily practice in the operations unit that performs the process. In the end, benefits are sustained by people, so instead of measuring the process for sustainability of benefits, we can measure the people and their capabilities as a closer proxy.

Variation
Let us now turn to the last, but some would argue the most important of all dimensions – variation between projects. Most if not all the above measures are transactional and not aggregate (a transaction in this case is a PI project), so they will also have the potential to show insights through variation analysis. For example, we could analyse if a given business unit, or practitioner unit, or industry, or project manager, is consistently better than their peer on number of projects, RoI, cycle times, reuse of IP and sustainability of change. That can lead to actionable insights on learning from the best and applying to the rest.

Measuring the performance of the PI function is just the first step towards improving the PI function as a whole. As with our advice to our clients, we will also need to change some work practices, devise and use tools and techniques to enable that change, and put in place a culture of continuously improving ourselves over time. In the subsequent articles of this series we will expand on these ideas.