The Sales Measurement Problem

I was reading Benedict Evans’ newsletter this morning, and it got me thinking. He wrote about AI and how nobody really knows how to measure intelligence. As I was reading it, I couldn’t help but draw the same line back to sales.

Sales has a measurement problem, too.

Tech professionals often believe their job is the most challenging. Sales professionals do the same thing. Everyone thinks their quota, accounts, and territory are uniquely difficult. I’ve been guilty of that myself.

As Benedict Evans notes, that kind of thinking built Uber. It also built WeWork.

With AI, we applaud the output, but rarely do we stop to ask if we’re measuring the right thing. As sales professionals, we obsess over pipeline, dials, demos, and forecasts. Our busywork may appear to be progress, but is it moving the needle?

Are you capable of sitting across from a CEO and earning their trust?

Can you dig into the real challenges a business faces and reframe the problem so clearly that inaction is no longer an option?

That’s selling. Everything else is just noise.

Recall the story of Clever Hans – the horse people thought could do math until they realized it was just reading its trainer’s body language. Sales is full of Clever Hans reps. They don’t practice real salesmanship, but they confidently mirror what they are taught is success and log activity that looks good on paper.

AI passing the bar exam doesn’t make it a lawyer. Just as a rep hammering out 500 cold calls doesn’t make them a closer.

The job isn’t outputs. The job is outcomes.

Revenue. Relationships. Reputational capital.

And here’s the real question: what happens when something is 10x better than a human at one thing, but useless everywhere else?

An AI that can spit out a perfect exam answer but falls apart when the question is asked a different way. A rep who can flood inboxes with emails but can’t generate interest, trust, or win the deal.

That’s the measurement problem.