Why you aren't getting the results you expect (part 3 of 3)
In part 1, we looked at the say-do gap, what we say and what we do aren't always aligned. In part 2, a dive into confirmation bias, the quick definition is that we have a tendency to seek out and believe data that support our preconceived notions.
But the question is, so what? The reason to read on is that these two things are connected, and they are affecting your bottom line. Think of this, you say something, it's fair to presume that you believe what you say. You then recall that corporate training that told you to "speak with data", so you go out to collect information, low and behold you were right the whole time! Yet a view of the monthly figures keeps coming up empty, nothing is moving! Time for a re-org, or a new initiative, or fire some people, anything other than questioning the method in the first place.
Say-do gap applies to your customers as well, just because they say something is most important to them (like price) doesn't mean that's how they actually behave!
A side note, if you want to see a great display of confirmation bias treat yourself to College Humor "If Google Was A Guy (part 3)" (at 45 seconds you'll see a textbook display of this bias).
What can we start with to combat the say-do gap?
Faster planning cycles. In part 1 I mentioned the planning fallacy, which is to say that we are poor estimators of how much time something will really take. This does not make planning a bad idea, but stop the weeks or months worth of getting the plan just right and mark some major milestone, an end point, and start moving!
Data, in a scientific manner. Don't forget you have a confirmation bias to combat. Just having numbers doesn't mean you are automatically better off. Data is there to answer a question, know what the question is, then go get the data. Here's a tip on a way to collect:
Statistics call for TRUE randomization. Don't call selecting your top 10 customers as "random". There is a simple formula in excel "=RANDBETWEEN" that can generate a random number between any range. Anyone in your company can take a set of data in excel, assign a number from 1 to whatever in column A and use this to select a random set. The key here is that every piece of data has the proper and likely chance of being selected.
How about that pesky confirmation bias?
As you may recall in part 2, the image below:
No matter how many times you see it, the bottom line will always look longer than the top (although it truly isn't). The fix to your bias isn't in awareness, the fix is what you do about it. Question your gut instinct, it's incorrect more often than you'd like.
The null hypothesis (default in statistical analysis) says that there is no statistical significance between two variables. Don't blink out yet! The importance of this is that statistical analysis is setup to start with believing the researcher or experimenter is WRONG! As most of what you will go out to prove or test, if you truly start with this in mind, the data you go to seek should disprove your theory, be your own devil's advocate. When you can't, the alternate must be true.
Don't underestimate the combination of the say-do gap and confirmation bias being root cause to lacking results at your firm. All of us are affected by these to some degree, knowing their out there and staying curious about change can make all the difference.