Monday was our publication day at gap intelligence.
Every week, we published millions of data points across upwards of fifteen market intelligence reports covering industries from consumer electronics to home appliances. All of it, on a single day. The work was so intense that Tuesday through Friday existed for one purpose: to make sure the next Monday went off without a hitch.
In the early 2000s, that meant a Herculean effort. Manual processes. Manual writing. Manual proofreading. Things that AI has since rendered almost quaint. As a result, many of our people, myself included, would arrive at eight in the morning and not leave until ten at night. I will not be naming this to the Department of Labor.
For years, our goal as a leadership team was to get everyone out the door by six on Mondays. Despite our best intentions, we never could crack it. We looked at processes. We looked at shortcuts. We explored automation. Nothing stuck. People were still leaving in the dark.
It reminds me of how many of us are applying artificial intelligence to our businesses today. We are spending money on consultants and AI integrators and we are simply not seeing noticeable traction on those investments.
This is what we did to leave by 6pm back in 2008.
A colleague passed along a piece of advice he had picked up from a Navy shipyard: what gets measured gets done. Meeting deadlines drives behavior and when you set a deadline and make it public, human nature does the rest.
The observation was spot on. We were not measuring what we were actually doing on Mondays. There were dozens of steps required to finish all of those reports. But we had never sat down to determine when each step needed to be completed, who owned it, or what a reasonable time expectation looked like. We were flying blind.
So I did what any slightly desperate CEO would do. I took things literally into my own hands.
I bought giant poster boards, one for each of our fifteen report categories. In marker, I drew a timeline of when individual milestones needed to be completed, and who owned each one, by name. If a particular set of work, say PC retail data, needed to be done by eleven in the morning, and Joe owned it, Joe’s name went on the board next to that milestone. I made fifteen of these charts. And then for the next two months, every Monday, I walked the halls of gap intelligence at the right times and asked each person whether they had hit their mark.
A lot of yeses. A lot of nos. We investigated the nos, which usually revealed another milestone that was not being measured, or not being done efficiently. Week by week, the picture got sharper. The milestones were owned. The owners were accountable. And after a few months, we were all out the door well before six.
This is not just a story about poster boards. It is the same principle Toyota built an entire production philosophy around decades ago. Measure your output. Find ways to improve it. Whether the metric is productivity, cost, or quality, the act of measuring is itself the intervention.
I share this because it is the exact same challenge we face with artificial intelligence today.
We have purchased the Claude subscription. We have hired the integrator and sent them running through our companies to make us more efficient. But just like our early attempts to leave by six, we have no idea where to begin, because we are not measuring anything. We do not know what tasks are being done, how long they take, or what good looks like. We are applying AI to processes we do not even understand ourselves.
The goal is not to leave by six. The goal is simply to get better and to know when we have.
So here is what I implore Esteemed members to do. Do not hire the AI integrator who will whimsically apply technology to processes no one has ever mapped. Instead, hire an intern. A college student from a liberal arts background who can meticulously document what your people actually do, step by step. With that information in hand, you can feed those workflows, those processes, those standard operating procedures into AI. With clear direction and defined steps, AI will turn those tasks into a machine gun of efficiency. And because you measured before, you will know exactly what after looks like.
AI is the latest shiny thing promising to revolutionize our businesses. Even in that excitement, we must never forget the fundamentals.
What gets measured gets done.
Give it a spin.
Be Esteemed.