A contractor named Ashley Stroud heard me speak at a conference a while back and walked out irritated. She wrote about it later, and she was generous enough to let me publish it:
Everything she was talking about — AI, business intelligence, where technology was headed — felt light-years ahead of where I was standing. It was overwhelming. In fact, I remember thinking, "This is a terrible sales presentation. No one in this room even knows where to begin processing all of this. How unrelatable!"
I have thought about that line more than any compliment I have ever received, because "light years ahead" is usually meant as a complaint, and the complaint is fair. If you are standing in a room and someone is describing a world you cannot see yet, that is not inspiring. That is alienating.
But here is the part I want to correct, and it is the reason I am writing this. Ashley experienced it as distance. It is not distance. It is not vision, it is not talent, and it is definitely not me being smarter than the room.
It is a loop. I run it on purpose, and it is boring, and anyone can run it.
The loop
Four steps. I have been running some version of this since 2008.
- Scan — pay attention to what is changing at the edges, not the center.
- Small bet — spend a little money and a lot of attention on one thing.
- Production test — put it in front of real customers with real money on the line. Not a pilot. Not a demo.
- Double down or kill — decide on evidence, and actually kill things.
That is the whole system. What makes it look like prophecy from the outside is that the loop has a cycle time, and most companies never start the clock.
Scan the edges, not the center
The center of an industry is a lagging indicator. By the time something is discussed on the main stage at your trade association, it is not an opportunity — it is a compliance requirement. Everyone learns it at once, which means nobody gains anything.
The edges are where it is visible early: what the platform companies are shipping, what tooling developers are complaining about, what your own customers are doing in the workaround they built because your industry has not solved it yet.
In 2008 the edge signal was that people had stopped asking their neighbor for a contractor and started asking a search box. That was not a prediction. It was just watching where the question went.
Make the bet small enough to be wrong
This is where most owners break the loop, and they break it in one of two directions. Either the bet is so small it cannot teach you anything, or it is so large that being wrong ends the company — which means you will not place it, and if you do, you will not kill it when the evidence says to.
A good bet is one where the downside is a quarter of wasted effort and the upside is a decade of position. You should be able to place several a year without flinching.
Tony Jeary taught me a four-quadrant time matrix years ago — important and urgent, important but not urgent, urgent but not important, neither. The discipline is not the grid. It is what you do with the second quadrant.
Nearly every good decision I have made lived in important but not urgent. The 2008 search bet. Narrowing the company to the trades. Rebuilding on our own platform. None of them were on fire. That is exactly why there was room to make them well.
Production is the only real test
An internal pilot tells you whether your team likes the tool. It tells you nothing about whether it survives contact with a customer who is paying you and expects it to work on a Saturday.
So the test is always the same: put it into production on our own house first. We rebuilt this company's entire web platform and ran it on ourselves before we sold a single client site on it. We went from a few hundred people to a few dozen and rebuilt the work around an intelligence layer, in public, with our own revenue as the thing at risk.
That is not bravery. It is just the only test that produces trustworthy information. If you are not willing to run it on your own business, you have not actually decided anything — you have formed an opinion.
Kill things, out loud
The step nobody talks about is the kill. A loop without a kill step is not a strategy, it is a collection. You end up maintaining every experiment you ever started, and the maintenance cost quietly eats the capacity you needed for the next bet.
Killing well means saying it plainly, in front of the team, with the reason attached. That is also what makes the next bet cheap — people will follow you into an experiment when they have watched you end one honestly.
Why it reads as distance
Run four steps with a twelve-month cycle time and after a decade you have made roughly ten corrections. The company that never started has made zero. The gap between you is not intelligence. It is accumulated iterations, and it compounds the way anything compounds — slowly and then not slowly.
From inside the loop it never feels like being ahead. It feels like being unsure and deciding anyway, most of a decade in a row.
From outside it looks like light years, and it sounds like a bad sales presentation, because I am describing the output of ten iterations to a room that has not run the first one.
What Ashley did next
Here is why her letter matters more than a testimonial. She did not go home and take my word for it. She wrote:
Then I started asking hard questions. Not just to Jennifer, but to the marketing companies and technology partners we were already working with. The more questions I asked, the more I realized that what Jennifer was describing wasn't science fiction — it was already beginning to happen.
That is the loop starting. Her scan step was interrogating the vendors she already paid. And the question she landed on is the one I would put to any owner reading this:
Was I going to become someone who criticized change because it felt uncomfortable, or was I going to become the kind of leader who learned fast enough to keep up with it?
Start the clock
If you feel behind, the instinct is to look for the big catch-up move. There isn't one. The owners who are ahead of you are not ahead because of a decision — they are ahead because their loop has been running and yours has not.
The fix is unglamorous:
- Pick one edge signal you keep noticing and cannot explain away.
- Place a bet small enough that being wrong costs you a quarter.
- Put it in production on your own business, not in a pilot.
- Set the date you will decide, and honor it in both directions.
Do that four times and you will stop needing anyone's predictions, including mine. You will have your own.
That is also why I publish my calls with dates attached in the Prediction Ledger — an operating loop that is never graded is just a story you tell about yourself.
Light years is not a distance between people. It is a distance between the number of times you have run the loop. Start the clock.