28Stone Says 230-Person Boutiques Beat Big Firms in AI Rollouts as 95% Consensus Misses 5%
Updated
Updated · Computerworld · Jul 20
28Stone Says 230-Person Boutiques Beat Big Firms in AI Rollouts as 95% Consensus Misses 5%
1 articles · Updated · Computerworld · Jul 20
Summary
28Stone’s founders say smaller, domain-focused consultancies can outperform AI labs and large integrators in enterprise rollouts by pairing agentic tools with human oversight rather than treating AI as a one-size-fits-all fix.
That model keeps business analysts, developers and product owners in the loop, with 28Stone arguing requirements discovery and code accountability cannot be handed to a single “unicorn” operator or a product team alone.
The firm says AI is reshaping competition: its 230-person capital-markets consultancy can now meet cost and scale demands that previously favored offshore delivery centers or firms with 300 to 400 staff on a project.
Governance remains immature, especially on spending — one client burned through $1 million of tokens in eight weeks — while enterprises still hesitate to commit as delivery forecasts keep falling and buyers wait for cheaper equilibrium.
For 28Stone, the broader risk is “vibe coding” and DIY enthusiasm souring clients on AI after failed projects, even as lower costs make proof-of-concepts and previously unapproved work more feasible.