Changing Paradigm

We are in the business of causing change. Change to the way organization flow work. By changing the way work is sequenced and when it is worked flow can be dramatically improved. Improved flow quickly brings attention to the constraint in the system. Focusing on the constraint starts the process of ongoing improvement. So far this all seems easy. Why would someone not follow these common sense steps?

The problem is in the existing belief system in the organization. Most often the people in the organization believe that the “right” way to keep things efficient is to make sure everybody is busy. They believe that “idleness” is a sure sign of wasted capacity that could be used to do something productive. They also believe that having work in front of people is a motivator to work harder. So the more work in the system the harder everyone will work. And the harder everyone works the better the organization will do. Are these beliefs valid. My experience, and that of countless other flow consultants and practitioners is that this is not a valid set of beliefs. Accepting that an organization only makes progress if it advances towards its goal. Accepting that work in the organization requires many dependent steps to be completed. Accepting that the flow of work will get constrained by some constraint. It is logical that the non constraints will have some idle capacity. So keeping everybody busy all the time is not possible. In fact idleness may not be “waste”, it may be necessary to have velocity in the flow. And people may be better motivated if they see how they contribute to the goal of the organization by the work they do, than by artificially pressuring them with piles of “work”. But these are challenging beliefs to change. As a result many, if not most, organizations remain mired in poor flow, distressed employees and inadequate outcomes for the organization.

We live in a world made of our beliefs. I have to be careful here. Not that the world is made of our beliefs. There is a “real” world. We just do not understand it fully. We have beliefs about the world. These are updated based on evidence we gather, and the strength of this evidence.

A Bayesian view of the world assumes that we do not truly know the world. We update our prior understanding of the world based on new evidence. Let us run through a simple thought experiment. I am handed a coin. It looks like a regular coin. It appears to be made by the mint. No reason to believe it is biased in any way. At this point my belief is that I have a fair coin. If I tossed the coin I would estimate that the probability of getting a head is 50%. To make the situation more interesting, the person who gives me the coin tells me that either the coin is fair or it is biased with a 90% probability of getting heads. At this point I am uncertain. I have an equal belief that the coin is fair or it is biased.

I toss the coin three times and get three heads. Should I update my belief about the coin. Should I estimate that the coin is biased? This is a somewhat tricky question because I do not really know anything new about the coin. I have not weighed it, tested its composition or done anything to understand the nature of the coin. For many people this implies that we should not update any belief. But stretching this thought experiment, what if I tossed the coin 20 times and got 20 heads. Now our intuition is quite strong. This outcome does seem to suggest that I should update my belief. To all appearances the coin is biased. How does one think about this problem of updating beliefs in a more rigorous way?

Let us start with the simplest case. We toss the coin and get a head. Should I believe that the coin is biased. Well if 20 out of 20 coin tosses made me quite sure that the coin is biased, how sure should I be after one toss. <Math alert 🙂 > Let us look at the two situations – if the coin is fair we have a 50% chance of heads. If it is biased it has a 90% chance of heads. I have equal belief that the coin is fair or biased before I tossed the coin. So for me the odds of seeing heads from a fair coin are 50% times 50% or 25%. The odds of seeing heads from a biased coin are 90% times 50% or 45%. (Now for the tricky part) We have a 45/(45+25) or 64% chance that the outcome is a result of a biased coin. So we now have a probability of 64% that the coin is biased.

What if I toss two heads in a row. Let us repeat this calculation. Our updated probability is 64% that the coin is biased. So the odds of seeing heads from a fair coin are 50% times 36% or 18%. The odds of seeing heads from a biased coin are 90% times 64% or 58%. We have a 58/(58+18) or 76% chance that the outcome is the result of a biased coin.

What if we roll head and then tails. Let us repeat the calculation. Our updated probability is 64% that the coin is biased after the first toss. So the odds of seeing a tail from a fair coin are 50% times 36% or 18%. The odds of seeing tails from a biased coin are 10% from 64% or 6.4%. At this point we have a 6.4/(6.4+18) or 26% chance that the outcome is from a biased coin.

What does this have to do with causing change? A big part of making a change in the organization is updating beliefs. The change agent uses logic, results, emotional appeals and political persuasion to change beliefs. Is it “right” that “idleness” is a sign of waste, or is it necessary capacity that ensures velocity in the flow of work? Is it “motivating” to see a lot of work that needs to be done, or more “motivating” to see how one can help the organization achieve its goals? These are beliefs and every argument, every result is the toss of a coin that updates these beliefs.

Nothing is certain. I may have very strong evidence by running a “pilot” that it is better to focus on establishing flow rather than keeping people busy. For the sake of argument let us say that we make some changes and the result is improved performance. Is it absolutely certain that the results were caused by the changes? Nothing is certain. There are many things that can cause results to fluctuate. So we can at best have some probability that the results are caused by the changes we made. This probability is hard to estimate. But for the sake of furthering this line of thinking let us assume that the odds are 90% that the results are caused by the changes. Should we expect people to change their beliefs?

It depends on what was the strength of their belief before we ran the pilot. Some beliefs are very strong. They are paradigmatic for a person. These beliefs are tied to a a lifetime of experience. Imagine that someone is 99% sure that idleness is “bad”. Let us calculate. For this person the odds of getting a good result because of changes are 90% times 1% which is about .9%. The odds of getting a good result by “accident” are 10% times 99% or 9.9%. So the updated strength of belief in the efficacy of changes is .9/(.9+9.9) = 8.6%.

On the other hand for a person like me who believes with 99% confidence that “idleness” can improve velocity of flow the same experiment has the following impact. For me the odds of getting a good result because of changes made are 90% times 99% or 89.1% and the odds that results are because of an “accident” are 10% times 1% or .1%. My belief as a result of the experiment is now at 89.1/(89.1+.1) or 99.88%. <End math>

So here we have it. For the change agent a very good test (90% chance of being right) updates their belief from 99% to 99.88%. For the skeptic with a strong belief (paradigmatic in nature) in the opposite the belief changes from 1% to 8.6%. This is the clash of perspectives in managing change. Evidence that offers certainty for one person leaves the other still quite skeptical. This is a very frustrating experience for all concerned.

So what can we do? Understanding this logic is a good starting point. It reduces frustration. It builds empathy and helps inform the best change management strategy. Every bit of evidence is useful in weakening the prior strength of belief before investing in a real pilot. Repetition matters. Repeating the experiment and being successful a second time would update the skeptics belief from 8.6% to about 50%. Logic is useful. It may be not have as strong an impact but it does update the prior strength of belief making the organization more likely to update their beliefs based on pilot results. Simulations matter in the same way. Appealing to even stronger beliefs also helps. Anything and everything that creates an opportunity to update strength of belief is important.

On the other hand if someone is dogmatic, they have a 100% conviction that they are right then no amount of data or argument will ever move them from this belief. But hopefully there are not too many people with that strength of conviction.

Organizations that are open to experimentation have a competitive advantage. They have room to update their beliefs with better ones. Healthy skepticism is good but dogmatism is a sure path to conflict.

Please do also read “90% right and 90% wrong at the same time“

Find out more about the TOC Club North America

To stay informed and continue this discussion

Join the Linked In group: TOC Club Bay Area

If you would like to read other articles I have written do visit my blog on LinkedIn and Focus. Please also join me on Facebook and follow me on Twitter (@KapoorAjai)