Real-world Insight on The MIT Nanda Report on GenAI
Generative AI (GenAI) has been one of the fastest-moving technologies in history. New tools and announcements arrive every week, and the hype cycle feels unstoppable. Yet, according to research from MIT’s Nanda Project, the reality for most organizations is very different: only around 5% of GenAI projects deliver measurable business impact.
That number is striking — and it raises a simple but critical question: why are so many organizations struggling to realize value from GenAI?
In this video, Matt Willsmore, Principal Architect at Pureinsights, shares his perspective on the MIT findings. Drawing on decades of experience helping enterprises build production-ready AI and search applications, Matt highlights the common pitfalls that lead to failure and the practical steps that increase the odds of success.
Four Recurring Themes from the MIT Nanda Report
Based on the MIT Nanda Report PDF and Matt’s commentary, four recurring issues explain why so many GenAI projects stall:
- Expectation mismatch – Employees are already using tools like ChatGPT on their own, while only a minority of companies provide official solutions. This creates gaps between expectations and reality.
- Weak use cases – The most successful projects started with a clear business driver and measurable success criteria. Many failed initiatives lacked this focus.
- Go-it-alone builds – In-house efforts only reached deployment about one-third of the time. Companies that partnered with vendors or expert consultants doubled their odds of success.
- Workflow misfit – Even the best tools fail if they don’t fit seamlessly into how employees actually work. Adoption depends on natural integration into workflows.
At Pureinsights, we see these patterns every day. The organizations that succeed with GenAI aren’t chasing hype — they’re embedding adaptive, workflow-aligned solutions that deliver real business results.
Related Resources
You may find these related blogs or articles helpful.