Global Consultancy
The AI worked. So why weren’t people using it?
Key metrics
240+
Users research group
20,000
Users engaged
500+
Beta users
20+
Journeys mapped
Activities:
AI Strategy / Product Design / Research / JTBD / UX / Agent Design / DesignOps / Adoption / Workflow Integration
From AI capability to human adoption
Challenge
The platform was powerful: chat, agents, citations, document retrieval and PowerPoint integration, but adoption lagged. Why? Research with 214 users showed that some of the most time-consuming and frustrating tasks in consulting were exactly where AI usage remained lowest. The technology could help. People just weren’t choosing to use it.
Process
We focused on the human side of AI adoption. That meant improving trust, making specialist agents easier to discover and integrating AI into tools people already used. The key question wasn’t: Can the model do this? It was: Will someone under deadline pressure actually use it?
Outcome
We treated adoption as a design problem, not a training problem. We looked at where people lost time, where trust broke down and where the product didn’t fit the way work actually happened. The research directly shaped leadership priorities for the next platform roadmap - focusing investment on client preparation, knowledge sharing, proposals and content review. No made-up ‘27% productivity gain’. Just much better evidence about where AI could create real value.
What We Learned
STRATEGY - start with the work, not the technology / AI - capability and adoption are different problems / BRAND - internal products have reputations too; trust is built interaction by interaction.
