One in ten people on your AI program are driving most of the activity. The rest drifted back six weeks in.
The question is whether you know who they are, and who others turn to when work changes. Most programs never ask. This one does.
users drive the majority of AI activity after six weeks. The rest have reverted to what they knew.
months before adoption changes how work actually gets done, if it changes at all.
question surfaces who others turn to when work is unfamiliar or changing. Most programs never ask it.
You can see login counts. You cannot see whether anyone changed how they work. Or who is shaping what the team believes is worth their time.
We read the environment before we approach a single person.
Four conditions determine whether change can take hold: strategic clarity, decision authority, information flow, and psychological safety. We assess all four before designing anything.
Two data sources, one question. Adoption data shows who is already doing the work. The network question finds who others go to when work changes. Together, they surface four personas.
Each persona gets a different sequence. Champions anchor. Ambassadors bridge. Signal Carriers get diagnosed, not recruited.
Influence is not on the org chart. The network question finds it. Two axes, adoption depth and network reach, place every person into one of four types that determine whether a program spreads or stalls.
Strong adopter. Deep trust within their team. Anchors adoption in their immediate group. Works through direct relationships.
High adoption Deep reachStrong adopter. Cross-team connector. Spreads practices across functions. The program's most effective bridge.
High adoption Broad reachNon-adopter. High network influence. Already shaping what peers believe about AI, often without realizing it. The most overlooked persona.
Lower adoption Deep reach Broad reachAll three traits present. Rare. When found, build the activation sequence around them first.
High adoption Deep reach Broad reachReach comes in two forms: locally trusted within a team (Anchor) and cross-functionally connected across functions (Translator). Both matter and require different activation sequences. The playbook covers both in detail.
"When the repetitive part is handled, what does this person do with that time? That's the question most programs never get to ask."
Seat counts and login rates are proxies. The goal is a different kind of work: more focused, more analytical, less administrative.
When you free up the right work at the right level, people have a choice about what to do with that time. We help make that visible before anyone is asked to change.
An enterprise AI pilot across tens of thousands of employees. Roughly one in ten users drove more than three-quarters of all activity. Users who worked across multiple tools, not just one, saved significantly more time per week. One division adopted at a dramatically faster rate than the others. The difference: leadership held a town hall before the rollout, not after.
Enterprise scaleChange management tells you what to do. This work tells you who to work through, and whether the conditions exist for anyone to change anything at all. The difference is the network layer: the people and relationships that don't appear on any org chart.
A field guide to finding the right people and sequencing the right moves before the program loses momentum.
Built for COOs, Chiefs of Staff, and Transformation Leaders running AI programs at scale.
Free. No sales call required. Just a working email.