21%

"Everything can be measured," my partner said.

"By whose standard?" I asked.

"Benchmark," he continued. "You know, a group of people agree on a few key metrics, and that becomes the standard. Like best car of the year, best football player in the world. Usually it's a group of professionals who decide what counts."

That was a random chat I had this morning with my partner when we were on the way to a local coffee shop.

It started with a tweet I showed him, someone calling an AI coding tool "the best software shipped in the last five years". He disagreed, strongly, and that is how we got here: by whose standard is anything the best, and how would you even measure it?

"What gets measured gets managed." Most of us have heard that one, and I have repeated it myself, more than once. But in my own work, the measuring was rarely the hard part. The hard part sits earlier: why we are measuring at all, and who decides what we should measure. Once those ones are settled, the number more or less takes care of itself.

That took me straight back to a meeting I had yesterday.

A salesperson was showing me a demo of an AI visibility platform. On the screen, the example brand had a visibility of 21%, its competitor over 90%. Behind those two numbers sat a stack of choices someone had made: which prompts to ask, which AI models to ask them to, what counts as being mentioned.

Then he paused, probably expecting me to ask about the gap.

Instead I asked something that was probably unfair to ask on a sales call: how was the 21% calculated? Compared to whom? Against what metric? And who decided this is what visibility means?

He said he would get back to me.