April 13, 2026

Conflicting Data

As you've been told, I write a Weather Newsletter. I also write a Pickleball Newsletter, but that's a topic for another day.

About three weeks ago an epic heatwave gripped the Western third of the country. I received two themes of feedback.

  1. This is what happens when climate change is amplified. The future is worse than this.
  2. This weather pattern is a normal weather pattern with an extreme outcome and has nothing to do with climate change.

Both sides are very passionate about their argument. They're quick to educate, they're in some ways overconfident ... they surely believe they are "right".

Forty years of data analysis suggest that humans struggle with conflicting data. We struggle with feedback loops. We struggle with interactions.

It happens in ecommerce. Remember our Category Impact table.



Category 17 is the biggest sales driver (sales are not shared in this table). It is also the biggest source of new customers. It also negatively impacts Category 6. This means when Category 17 does a good job, Category 17 hurts Category 6.

Assume you are the CEO of this brand. The merchant in charge of Category 17 looks great. She's responsible for a high-volume category that brings in new customers. And yet ... every time she does a good job, the poor merchant in charge of Category 6 looks bad. His sales decrease.

Should the CEO fire the guy running Category 6 if sales decline in Category 6? Or is the guy running Category 6 at the mercy of the woman running Category 17?

It's a nuanced situation, isn't it?

Just like weather / climate is a nuanced situation.

Obviously, you need the right data to even consider a reasonable response.

Almost none of you have the data to answer this question.




No comments:

Post a Comment

Note: Only a member of this blog may post a comment.

Conflicting Data

As you've been told, I write a Weather Newsletter . I also write a Pickleball Newsletter , but that's a topic for another day. About...