The messy, very human overlap of data with the real world.

The data we work with does not exist in a vacuum. It is being interpreted by human beings, and human beings are inherently irrational. We are affected by biases and emotions.

Data is a proxy for the truth, not truth itself. We use data to try to smooth over how complex and contextual real problems are, but that means we inevitably don't get the full picture. If someone can't tell me what they don't know and can't articulate the limits of the data, I'm not going to trust what they say they do know.

Nothing matters more than the last mile between data and human beings. I've seen millions of dollars spent on incredible analysis that never gets used because someone didn't consider how to communicate with their audience. You need to understand the people using the data first.

People sometimes create a false dichotomy in which design and data exist in separate worlds. They do not. There is nothing but overlap in these disciplines.

Technical data skills are common; common sense is not. I've never had a problem finding people who can use Excel, write Python, or handle complex SQL. The real struggle is finding people who understand the context of the data — where it comes from, why we track it, what it actually can (and cannot) tell us.

It took five minutes of research to find out they were working with one of the two big AI brands — and then to get hit by the realization that there was a timer ticking down, much more quickly, on the defensibility of my work.

I offered to license training data to them, to no avail. What I've come to is that there's not much to be done that's going to protect my work. I send out tens of thousands of free files every month — everything is out there already. Plus my whole brand is sort of built on teaching people how to replace me and my work.

All data presentation is a form of communication design, and the goal is to help the data communicate more clearly, not to obscure it. The idea that data can "speak for itself" is a myth — all data is presented in some context, with some visual choices, and those choices either help or hurt communication. The question isn't whether to make design choices, it's whether to make them intentionally.

A report that doesn't get used isn't of any value, no matter how good of a report it is. People make a lot of assumptions about the credibility of the data they're seeing based off of how well designed it is, how well laid out it is. We might not want it to be that way, but that's the way it is. People are irrational creatures, and we have to consider that.