People on your team have already decided AI is not for them.
They are good at their jobs. Every rollout plan you have read assumes they do not exist.
Jim’s cousin got let go last month when Oracle announced a 20,000 person layoff. Oracle’s leadership (i.e. crisis management PR team) claims AI is making the company more efficient. Jim’s pissed. He is not going to open a chatbot on Tuesday because you sent a memo.
Suzanne tried ChatGPT in 2024. She asked how many R’s are in “strawberry.” It told her two. She has not been back and doesn’t see any reason to. Fool me once, right?
Zoe heard on a podcast that data centers are making her water bill go.
These people exist. They’re great at what they do. They’re not using AI anytime soon. And when you check the usage dashboard, you find what nearly every company finds in year one: a small fraction of the company has actually made AI part of how they work. The licenses went out. The memo went out. The number did not move.
You are not buying seats. You are buying habits.
Here is what makes AI different from every system you have deployed. With traditional software, everyone learns the same tool and gets the same answer. With AI, the way each person learns is different. The answer each person gets is different too.
So you cannot tell who is using it well. You cannot tell who is using it at all. And you cannot forecast the spend.
That gap does not close with a license and a video.
Adoption stall curve. Source to be confirmed.
It closes when someone sits down with your regional GM and learns how work actually works: who creates it, who hands it off, who finishes it, and who decides it is done. Then that same person talks to finance, to HR, and to IT.
They need to know the business. They need to know the people.
We hire for that: the patience to sit with someone who does not want to be sitting with you, and the judgment to say when AI should stay out of it entirely.
When should AI stay out of a process? We’ve met estimators that can look at a hole 20 feet deep and a football field wide and give a price estimate for the construction job.
Workflow diagram. The rectangles and arrows we build in every assessment.
Adoption is a gym membership, not a surgery
Nobody gets in shape by going to the gym once. You get in shape by going two or three times a week for a year, and then continuing to go.
AI works the same way. Companies buying a one and done rollout, a keynote, a training day, a launch email, will feel great in month one and be back where they started by month twelve. The tools will have changed twice by then. The people who were excited will have drifted. The small fraction will still be a small fraction… or smaller if the motivated and curious found greener pastures.
What ten hours of listening produces
We interview the people who actually do the work, and we survey the whole team to find out where they really stand. Then you get a readout.
Most leaders arrive with a list of five to ten AI projects already written. Most of those lists were built on assumptions nobody tested against the people who would have to live with them. The readout is where you find out which ones survive contact with your own team.
01
Where work stalls today, by role and by handoff
02
Which opportunities are worth doing first, and which are not worth doing at all
03
An honest baseline: who is using AI, who is not, and why
What six to eighteen months looks like
Baseline and first wins. Interviews, surveys, and a workshop where nobody leaves without building something. We measure where every person starts.
Habits. Group trainings and one on ones every month. The first skills and artifacts move off one person's laptop and into something the whole team can use. We measure again at 45 and 90 days.
Ownership. Named internal owners. Governance, permissions, and cost tracking sorted out with IT and finance. Onboarding and offboarding for AI, so people arrive equipped and leave without taking your work with them.
Independence. Your team builds without us. We maintain what needs maintaining, and we tell you honestly when you no longer need us.
From Google Docs to scalable pipeline
Kojo moved sales and delivery out of scattered documents and founder memory, so both teams could see the pipeline and carry work forward.
View Case Study