← back
Moving away from Agile: What's Next – Martin Harrysson & Natasha Maniar, McKinsey & Company
Takeaway
AI coding tools deliver individual 10x gains but enterprise impact stalls at 5–15% until Agile-era team processes are redesigned around AI throughput.
Summary
- McKinsey's Software X practice surveyed ~300 enterprises and found that despite individual demos showing days-to-minutes wins, company-wide productivity gains average only 5–15% because Agile-era processes are the bottleneck.
- Argues AI-driven software development is a paradigm shift on par with the move to Agile 20 years ago and that team-level collaboration and review workflows must change, not just tooling.
- Cites Carnegie Mellon research that AI-generated code amplifies existing tech-debt and quality issues when human review stays manual.
- Speakers propose new operating-model patterns: scaling individual productivity to whole teams, then to organizations, with redesigned review, ownership, and value-measurement loops.
enterprise-aiagiledeveloper-productivity
Original description
Most enterprises are not capturing much value from AI in software dev to date (at least relative to the potential). The reason is that most are adding AI tools to their dev teams without changing the people and operating model aspects (i.e., limited changes to ways of working, team configurations, role definitions, stage gates, etc.). Many core aspects of software development haven’t changed in the past 10+ years, and that’s holding us back from moving to the new paradigm of software development! We will share examples of what makes the difference. https://www.linkedin.com/in/martinharrysson https://www.linkedin.com/in/natasha-maniar-945276107/