Learn and improve
The 1% Board
Improve every part a little
Pursue many small compounding improvements to how the team works with AI instead of chasing one transformation.
Borrowed from British Cycling under performance director Dave Brailsford, who called the approach "the aggregation of marginal gains." Read the original source.
Ongoing habit · Days or ongoing effort · Anyone on the team can start it
Try this first
Pick a few small improvements to how the team works with AI, make them, and track them where people can see.
Use it when
The org is chasing a single big AI tool as a single dramatic fix.
Skip it when
Sometimes a real step-change is warranted, so this is no rule against ever doing something big. And the small improvements need a direction. A scatter of tweaks is not compounding.
How to introduce it
Break the way the team works into parts and improve each a little, tracking the gains where people can see them. British Cycling built a program around one-percent improvements. It fits the evidence too: value comes from many complementary changes to how humans and AI work together, not from one dramatic tool.
How to show up
Help the team find and track many small improvements, and keep attention on the accumulating habits. The outcomes lag the habits, so the scoreboard has to show the habits.
How long it takes
Ongoing. Small improvements tracked over time.
What makes it hard
Small improvements are unglamorous and the hype pulls toward the big transformation, so valuing the incremental takes discipline. Because outcomes lag, the accumulation feels like it is not working before it pays. Hold the course.
What it looks like when it's working
The team makes and tracks small improvements and the gains accumulate. Chasing one transformation while the small changes go unmade means it is not being applied.
How long until it sticks
Weeks to months before the compounding gains become visible and the approach proves itself.
How you know it stuck
The team pursues compounding improvements rather than waiting for a transformation.
The idea behind it
Big AI value rarely arrives in one move. It accumulates from small changes to how the team works. Chasing the single transformation misses where the gains are.
Where it comes from
British Cycling. The hype pushes one big transformation. The evidence says the value comes from many small changes to how people and machines actually share the work. Break the work into parts and improve each by a little. Results follow the habits, so the gains show up later than the effort.