The objective of management is to make improvements—not only to products and services but also to entire processes. The difficult thing is to know what needs improving, and then whether your interventions have made a positive difference. This course provides an overview of the basic tools used for process improvement, such as statistical process control, and how to use these tools to improve the three most critical aspects of your business process: time, quality, and cost. Chris Croft distills the best practices from process improvement frameworks such as Lean and Six Sigma, and combines them into lessons that will help take the core of what your business does, measure it, and do it better.

Over 25 years ago, Stephen Covey wrote the book 7 Habits of Highly Effective People. The 7 Habits are universally applicable habits that can help you be more effective personally & professionally.
  1. Measuring and Managing
  2. Measure the Right Thing
  3. How to Measure Process Improvement

Measuring and Managing

by Chris Croft

 

Before you can improve a process, you need to know whether it needs improving. And then, as soon as you start to intervene, some might say interfere, you need to know whether you've made a difference, or rather, you need to know whether the changes were due to what you did or due to something else or just due to pure chance. After all, there's a 50-50 chance that this year will be better than last year even if you do nothing. So we need to measure what's going on with our processes. In fact, it's been said that if you can't measure it, you can't manage it. But do you think that's really true? Can't we manage something without being able to reduce it down to a number? In fact, can everything be measured? What about quality or service or design or culture? Can these be measured? Well, I would say yes, but then, I started out as an engineer, so I'm bound to say that. And I do believe that if you're not measuring something, then there's a risk that you'll end up assuming things are okay when maybe they're not and you won't take action until they've got really bad and the problem has become visible even to someone who's not measuring anything. It would be good to not leave it that late. So I'm going to be suggesting on this course that we work out the key things and measure them so that we can monitor and then optimize them. And optimize doesn't mean maximize or minimize. The best place to be is often at 80% rather than 100%, as we'll see. So, it all starts with scientific management. Decide where you want to be and then measure your progress towards that point. My heart sinks every time I hear a company say, "We're having a reorganization," because I just know that that means they're going to have several years of chaos, expensive chaos, with the good people leaving and the bad ones waiting in the hope of redundancy money. And then, when all the pieces they threw up in the air have fallen back down again into a different pattern, yes, it's different, but is it any better than before? Some bits are probably better, but some are also probably worse. So, much better would be to measure everything, work out where the problems are and just target those areas, and then measure whether you're getting those problem areas sorted out or not. So, the first question of process improvement is: Are you measuring enough? What are the black holes, the processes that apparently can't be measured in your area of work? If you had to measure them, how would you do it?

Measure the Right Thing

by Chris Croft

 

If you try to measure everything, that could take a lot of work, and you might end up not knowing which numbers to focus on. Often measurements are self-contradictory too. You might be wanting to increase quality and reduce costs and be measuring both. But then which one do you give priority to? And of course, if you do get your costs down, you need to make sure it's not affecting your quality or anything else important. Never measure or try to interpret just one number without context. If your turnover has gone up, what about your profit? Maybe someone's been cutting prices in order to sell more. And if your numbers employed have gone up, is that good or bad? It depends on your turnover. Even something like complaints may have gone up because you're getting better at listening to your customers or doing some ambitious work. So never measure just one number. There's something hard to resist about numbers. When you hear that your local school is 5% worse than the one in the next neighborhood to you, immediately you feel unhappy. But actually, what does worse mean? And is 5% really significant? Was it just a blip when usually it's better than other schools? And maybe your school is worse at one thing but they're better in other ways that are more important. How do we know that the school down the road isn't just better at playing the game so they're getting a better score at the cost of areas that aren't measured? A recent example of this in the UK was that the government is trying to measure doctors. And they published a league table of scores, but it was discredited and then discontinued because some quite important factors weren't in the scoring. And also the weighting of the factors was debatable. And it was hard to take into account the affect of living in richer and poorer areas. And most importantly, the doctors who were prepared to take on the most difficult patients were getting lower scores despite being probably better. So should we give up and ignore measurements? Well, that would be a shame. Without measurements we haven't really got anything at all. But the thing is not to judge people from the measurements but to take measurements as a starting point, a start for asking why. Why has that doctor got a lower score? Why does that school get better results? This will tell us what's really going on so that we can understand the causes of the scores and then decide whether they really matter. And the best way to ask why is to first get a ratio. For example, infections per operation, days absent per pupil, accidents per passenger mile, errors per report, that kind of thing. And then compare that ratio either with similar people or departments or companies or with yourself at the same time last year. Comparing ratios is the only valid way to interpret statistics like this. 20 infections means nothing. 20 infections per hundred operations is better but still doesn't tell you enough. When you find that everyone else has only 10 infections per hundred, now you're getting somewhere. So it has to be ratios and then compare.

How to Measure Process Improvement

by Chris Croft

How to measure process improvement from Process Improvement Foundations by Chris Croft

 

All the time, you should be questioning the data. Ask yourself if the person giving the numbers has a reason to be biased. How reliable is their source? Could it be happening by chance? Are they looking at the whole picture? And we be sure that A caused B? So these are my five tests of information. Number one, is it likely to be biased? Number two, is it a number or a ratio? Number three, is it significant? In other words, could it have happened by chance? Does the number vary so much from one time period to the next, this could easily just be a chance blip? And what aren't they measuring? And finally, can we be sure that one thing really did cause the other? Causation is particularly tricky. Often, numbers are related, so that if one goes up, the other one goes up. But they aren't actually causing each other, there's just a common cause. For example, children from educated homes are more likely to read, and also more likely to do well at school, but that doesn't mean that it's the reading that's making them do well, it could be their attitude in the classroom, or the homework that they do at home, and this distinction is important, because if you discover a correlation between reading and doing well at school, you might think, "Right, every child must read more," and that might be a waste of time. It might not, of course, but the point is we don't know. Proving that A is caused by B and not C is tricky. You'd have to do a controlled experiment and find middle-class children who don't read and compare them to those who do. Or get some children from a deprived area to read more and see if they improve more than an identical group who don't read. Just finding an existing correlation is lazy and doesn't prove anything at all. Similarly at work, finding out what makes you sell more could be tricky. If you reduce the price and sell more, could it of been something else that caused it? Did anything else change at the same time? So as an exercise, I'd like you to think about what you measure at the moment. Are you measuring just one thing out of context? Are you measuring too many things, to the point where you don't know which ones to focus on? Are you measuring ratios rather than straight numbers? And are you comparing them with similar people or time periods? And are you sure about your causations?

Optimizing your quality, costs, and time tradeoff

by Chris Croft

 

In every process you've got quality, cost, and time. The quality of what you're making or doing if it's a service, the cost of the process, and the time it takes. And by the way, time isn't the hours spent doing the work, because that's cost. It's the lead time. How long customers have to wait after placing an order. And time could also be on time delivery, although that's usually a function of variability which is more likely to be to do with quality. If things vary each time you do them, maybe having to be done twice, then your on time delivery will be affected. So really when you boil it down, you've got quality, cost, and time at the root of your processes. And the problem is that you can't have all three. You can get your costs down by reducing your quality. Higher quality usually takes longer to produce. Or less obviously, it costs you more to produce things quicker, and I'll explain that in a minute. Many organizations haven't really thought this through. They want to deliver great quality at a low price, and as soon as the customer wants it. And often this is what's promised to the customer by the sales guys or by the advertising campaign. And the plan is to deliver it by either putting lots of pressure on the workforce, maybe even shouting at them, or by putting one person in charge of each thing. A quality manager, an operations manager, a sales manager, and a finance manager, and letting them fight it out. The risk with this is that you might get a sub-optimal mix. Whoever is strongest will win. Or you'll end up with a lukewarm mixture of quality, cost, and time, when maybe the type of market that you're in means that you should be going for maximum quality or maximum speed of delivery and making a conscious, realistic decision to be less good at the other two factors. Generally in most markets there's been a trend from cost to quality and then to time. Until maybe the '50s, it was all about being the cheapest. And then quality became an issue. Just think about Japanese cars and how they went from cheap to high quality. And now that everything's well made, the low-quality producers have mostly gone bust. The focus is mostly on time. However, the Chinese have maybe moved the emphasis back to cost since they can produce things so incredibly cheaply because of their lower wage cost and their economies of scale. And another reason why cost is a focus at the moment is that the internet has made shopping around on price so easy. But this could be short term. There is research that shows that in the end its the high quality producers, think German cars or Apple laptops, that slowly build market share and make more profit per unit. So they make the most profit in the end. In the next section I'm going to show you why this might be true by looking at the relationship between quality, cost, and therefore profit. And we'll come back to lead times later. But first, maybe it would be interesting to think about your organization or your market. How cost sensitive is it, and would you do better to be focusing on low costs and prices or high quality or rapid delivery times? What are your competitors doing? And should you be changing your overall strategy do you think?