Generative AI projects largely depend on change management and culture to move from pilot to production, according to BT150 members. In addition, performance management is an area where AI could be a big help.

Those are some of the takeaways from Constellation Research's August BT150 meetup.

Generative AI, budgets and proof-of-concepts. With 2024 nearly 75% complete, the jury is still out on generative AI proof-of-concepts getting to production. Anecdotally, the lack of returns for many genAI projects is an issue since AI took funds from other critical enterprise projects.

One issue for genAI projects is that there are vanity proof-of-concept genAI projects. These projects were championed more for resume building than use cases. In addition, many companies have lacked the data strategy to execute well on AI.

BT150 members noted that in their experience culture for change and change management was a more important indicator of success than vendors, models and other technology.

BT150 recaps

A BT150 member said AI-driven management coaching software has been helpful to employees and has boosted performance at a double digits clip. This bot works though scenarios and difficult conversations. The upshot is that one return from AI products may be behavior change and thinking through scenarios differently. CxOs could start thinking in terms of AI apps beyond just doing tasks faster.

Performance management software is a tough sell even if it has returns. "Performance management is highly relevant at this time; however, time are tough, the economy sucks and no want wants to pay for anything," said one BT150 member. "Everyone realizes there's a huge problem with performance management, but not one use case was allowed to go forward. Our use cases are either things that save money, increase margin or are customer facing."

Perhaps the biggest issue is that enterprises can't measure the ROI for effectiveness well. Instead, enterprises fall into the trap of efficiency where cuts are more important than effectiveness.