A framework for AI-augmented organizations
Core Perspectives
01
You have a use case management problem, not an AI adoption problem.
Stop debating which AI vendor is best. Start understanding which processes benefit from AI.
02
Your best AI idea is already in the building.
Frontline employees know where the pain is. Give them a framework to capture it.
03
Bottoms up over top down.
Don't buy $20M of Copilot then figure out why. Build use cases first.
The Momentum Manifesto
In building AI products and driving AI adoption at large organizations, we have come to understand that success is predicted through adherence to the following values:
Process Realityoverproof of concept
Business Innovationovertechnical innovation
Augmented Staffoverautonomous execution
Bottoms Upovertop down
While there is value in the items on the right, we value the items on the left more.
What This Means
Process Reality over Proof of Concept
Understand the process deeply — don't just demo. AI built on a simplified version of how work happens will fail in production. Go to where the work is done and document what actually happens.
Business Innovation over Technical Innovation
Find innovation that changes the work. Business Innovation alters roles and responsibilities, changing what is possible. Technical Innovation delivers the same product slightly cheaper. If go-live looks like your team copy-pasting problems to a chatbot, something has gone very wrong.
Augmented Staff over Autonomous Execution
Make the humans better. AI should transform what people are capable of, not just eliminate headcount. Most processes have exceptions, edge cases, and tribal knowledge that only humans can navigate — that's why the role exists. Start by defining what the human's new, better role is. The headcount conversation follows the capability conversation, not the other way around.
Bottoms Up over Top Down
Ask the people who do the work. The best AI use cases are discovered by the people closest to the process, not mandated from the boardroom. Top-down alignment determines what gets done, not what is possible.
The Best AI Ideas Are Already
in Your Building
“Truth does not need to be invented, only recognized.”
Already started?