This week I had a conversation in which a friend said: “Bad results begin with bad assumptions”. We scrabbled around with the theme and got from “Assumptive Bias” to “BIASsumptions” (which is not an actual word unfortunately). A fun to have conversation, but I believe underlying is a serious matter that brought me to write this post.
Let me start by saying that some form of guessing or interpreting facts is inevitable when you are embarking upon new roads, trying out stuff, experimenting or analyzing e.g. process failures and their consequences. ” Assumptions” are not the problem. It is how we treat assumptions and what we believe are good assumptions that worries me. My 3 main worries about assumptions are:
- Assumptions are often built upon “common knowledge or understanding” that are presented as facts
- When used for building business cases assumptions are hardly ever checked in hindsight
- Attempts to validate assumptions lead to “biased” research aimed only to prove one’s assumption is correct
Assumptions based on Common Knowledge or Understanding, not supported by facts
I wrote about one of the most sticky assumptions, which is regarded knowledge in the Customer Services arena, before: Why keep chasing the wrong Goose? The assumption I’m talking about: One should pick-up the phone within (on average) 20 to 30 seconds to keep customers satisfied. As this might be true for a specific company, it is most likely not to be true for all companies. It actually doesn’t matter if it is true or not in general. What does matter is if it is true for you, or better, if you can change it to: what is customer waiting time tolerance for your company? For a good example on how one could prove what’s true for your company, take a look at this post by McKinsey. Research based on the facts & data of your company is the best you can get. You have it, why don’t you use it? As the example shows: a lot of waste can be prevented.
Assumptions are hardly ever checked in hindsight
Much attention gets paid to the assumptions in preparing a business case. It therefor has always surprised me how little time is taken to review afterwards, whether the business case came through or not. In both cases this is missed opportunity. First of all you can learn from the validation of your assumptions. It might be your assumption came through, but the result of your business case didn’t, or the other way around. There is always a great learning potential by checking in hindsight. Best case: you have proved your assumption for this business case. Now you know it is a better assumption for your next business case. But never forget: even if it came through 2 times or more, you need to check every time. Don’t miss-out on the opportunity to learn.
Biased research
The first two are about failing to check, my last one is about poor validation. The issue here is that most people tend to look only for the positive proof of the assumptions they made. We often disregard completely the proof against the assumptions. It is not only scientifically better to make a real effort to not prove your assumptions, it also makes your assumption (much) less of an assumption if it comes out as true (or highly correlated at best actually). True validation is about checking both sides of the coin. It will provide you with great (personal) satisfaction if you do, even if your assumption proved false. At the same time you might have saved your company from adding more waste to the bottom line.
Bad results begin with hypotheses poorly validated
The best way to take the turn away from Assumptions in your business cases, plans or analysis, is to change from Assumptions to Hypotheses. As discussed above, assumptions are regarded truths without the facts supporting it, are poorly checked and often a result of bias. Hypotheses are not about that. Hypotheses are screaming to be challenged. Change your assumptions to hypotheses the next time you need one and dig in the facts.
And, if you get any of those facts from Customer Surveys, don’t take your Customer’s word for it: you have another Hypothesis in your hand..
What do you say?: will you start working with hypotheses and walk away from assumptions?