AI Adoption
Understand what is shaping AI adoption and adjust the rollout as it evolves.
Latent Variables keeps the AI program invested where value is forming: it identifies where AI is changing work, informs decisions about workflow, policy, and investment, helps transfer working practices between teams, and returns as the technology and the organization evolve.
What leadership has throughout the AI program
A current view of AI usage and work practices
As the technology and the organization change, each wave updates the same context, so the program works from current information.
Investment decisions with evidence attached
Findings connect to the workflow, policy, or capability decision they affect, with a response specific enough for the relevant team to act on.
Transfer of working practices between teams
Practices that work in one team are brought to comparable teams through targeted engagement, and the adaptations each team needs are identified during the transfer.
Leading indicators of adoption and value
Repeat conversations show where value is developing or stalling before aggregate adoption metrics reflect it.
How Latent Variables supports AI adoption as the program evolves
Establish the baseline
Conversations across the program’s reach map where AI currently fits the work: which workflows are changing, which conditions are supporting adoption, and which assumptions the program should test before expanding.
Build the response
Findings are connected to the program’s decisions: a workflow redesigned, a policy adjusted, capability built, or investment redirected. The response addresses the specific condition the evidence identified.
Transfer and support new practices
Working practices move to the teams positioned to use them, with the reasoning explained. The adaptations each team needs are identified as the practice is introduced.
Measure and adjust
As models, workflows, and expectations change, repeat conversations record what held, what emerged, and where the next investment should go. The process repeats as the program and the technology continue to change.
How findings change workflow, policy, and rollout decisions
A working practice worth transferring
The finding
One team had redesigned a handoff around the model and was measurably faster. The program’s telemetry did not distinguish them from teams using the tool without changing their process.
The response
Leadership codified the practice and brought it to adjacent teams doing comparable work, transferring the process redesign along with the tool.
The follow-up
Conversations with the receiving teams followed the transfer, identifying what each needed to adapt for the practice to hold.
A policy boundary blocking legitimate use
The finding
A policy written early in the program was blocking a legitimate, high-value use in one function, and people were resolving the conflict individually.
The response
The policy was adjusted with explicit guardrails for that use, replacing individual exceptions with a sanctioned path.
The follow-up
The next conversations confirmed the sanctioned path was being used, and monitored for new edge cases as usage matured.
High usage without corresponding results
The finding
A function showed strong usage and little measurable result. The conversations traced this to a process that had not changed around the capability: the tool had been added to the existing workflow.
The response
Leadership paired the deployment with a redesign of the underlying process and enablement for the people running it.
The follow-up
Later waves measured whether cycle time changed, rather than whether usage remained high.
How the adoption and value evidence is produced
Every conversation contributes to the same campaign context, so the same AI capability can be compared across teams and workflows as evidence accumulates, and new uses, verification practices, and policy conflicts can be followed as they emerge.
One instrument across teams and workflows
Each conversation adapts to how the participant actually works with the capability while following a consistent evidence protocol, so usage accounts can be compared across the organization.
Differences examined before the rollout expands
If one team reports significant value from a use case while another reports additional work, the system compares the operating conditions around both and identifies what differs before leadership expands the rollout.
Context that persists as the program evolves
Repeat conversations build on the earlier evidence, so the program can distinguish a condition that was resolved from one that repeated as usage matured.
What shapes AI adoption across teams and workflows
AI programs can measure deployment, usage, and spend. The conditions that determine whether AI becomes part of how work is done are harder to observe.
The technology and the organization change at the same time
Model capabilities change while teams are still absorbing the previous change. Program decisions need current information about both.
Usage metrics do not explain adoption
The same usage number can represent redesigned work, added verification steps, or minimal engagement. These differences determine whether value appears, and usage metrics alone do not distinguish them.
Work practices develop faster than programs can observe
Teams adapt tools and workflows continuously. Important developments, both useful and risky, often occur before the central program becomes aware of them.
Bring us one live AI adoption decision.
We’ll describe what we would test, what a response could involve, and how you would measure whether value is developing.
Discuss this application
Bring one transformation decision in planning or execution.
Pick a time on the rightbelow to learn more about our research and explore whether it may be relevant to your organization.
30 min
Web conferencing details provided upon confirmation.