Key Takeaways
- • Code & Co. principals Dan Bender and Kirby Montgomery argue that engineering output can now be measured in terms a board understands, with agentic AI workflows cited as the enabling shift.
- • Recommended metrics pair speed with stability and prioritize outcomes over activity, with the guiding principle "measure to learn, not to punish."
- • Technology diligence in private transactions is moving earlier, beginning before the letter of intent (LOI) is signed.
- • The guest warns that data-room dashboards presented during deals may not reflect production-grade systems, and that activity capture can be gamed while context cannot.
Summary
In a podcast episode titled "Measure Engineering Like You Measure Sales," host Jim interviews Kirby Montgomery and Dan Bender of Code & Co. on how investors and boards should evaluate software engineering organizations. The discussion frames the problem with a historical example: in 1982, Apple engineer Bill Atkinson improved the performance of the Lisa by removing 2,000 lines of code, yet the management form of the era asked how many lines he had written. The guests argue that organizations repeat this error today by measuring token consumption rather than results.
According to the episode description, agentic workflows make engineering performance measurable in financial and governance terms that directors can evaluate. The guests recommend tracking speed alongside stability, and outcomes rather than activity, so that productivity gains do not come at the cost of reliability.
The conversation also addresses transaction diligence. The guests state that technology diligence now starts before the letter of intent, rather than after exclusivity. They caution that polished data-room dashboards may not reflect the actual state of a target's systems, noting that some are running on localhost. They also contend that activity metrics can be manipulated while the underlying context cannot.
The episode closes with the principle "measure to learn, not to punish," which the guests present as the basis for effective engineering metrics. Code & Co. is identified as the firm the guests represent; further detail on specific metrics, deal examples, or quantified outcomes is not provided in the source material.
According to the episode description, agentic workflows make engineering performance measurable in financial and governance terms that directors can evaluate. The guests recommend tracking speed alongside stability, and outcomes rather than activity, so that productivity gains do not come at the cost of reliability.
The conversation also addresses transaction diligence. The guests state that technology diligence now starts before the letter of intent, rather than after exclusivity. They caution that polished data-room dashboards may not reflect the actual state of a target's systems, noting that some are running on localhost. They also contend that activity metrics can be manipulated while the underlying context cannot.
The episode closes with the principle "measure to learn, not to punish," which the guests present as the basis for effective engineering metrics. Code & Co. is identified as the firm the guests represent; further detail on specific metrics, deal examples, or quantified outcomes is not provided in the source material.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.