Most AI failures get blamed on the wrong thing.
This week Amazon Web Services committed a billion dollars to build a Forward Deployed Engineering group: thousands of engineers embedded directly inside customer companies, building and shipping AI systems in 45 day cycles. Microsoft answered within days, committing 2.5 billion dollars and 6,000 engineers of its own, working in small pods inside client environments with full access to company data.
A new PYMNTS Intelligence report explains why both companies are moving this fast. 71 percent of executives at billion dollar companies say organizational readiness, not the technology, is the primary barrier to AI performance. Only 11 percent blame the tech itself. A separate Workplace Intelligence survey of 2,400 executives found that 97 percent report some benefit from AI, but only 29 percent see significant organizational ROI.
I’ve spent 25 years in change management, and I recognize exactly what’s happening inside these engagements. A small team walks into a company, sits in the real meetings, learns the informal power structure, and builds trust with the people whose daily habits have to change.
The job title says engineer. The work is human.
Amazon proved it at Lyft, where an embedded team cut driver support resolution time by 87 percent, not by shipping a better model, but by changing how the people doing the work actually worked.
You don’t need a billion dollars to apply the same lesson:
- Start by mapping your human success factors. Identify the three or four specific people whose daily behavior needs to change before you deploy anything, and bring them into the design process from day one.
- Assign an internal change champion. Someone whose job is tracking and removing the barriers people hit in their first 90 days, and who is genuinely good with people, not just systems.
- Track workflows instead of seat activations. Thirty days after any deployment, you should know whether the target workflow actually changed, and whether you’re saving real time or money.
This Week
Review your organization’s active AI deployments and ask a single question: who owns the human side? For each initiative, identify whether there’s a named internal person responsible for behavioral change and business results, not just training delivery or license management.
Who owns that question inside your organization right now?
I publish Leapfrogging the Headlines to help leaders cut through the noise, gain clarity on what’s happening, and make smarter decisions. Subscribe below to get each issue delivered free.
Join 30,000+ other leaders