3 Comments
User's avatar
VerbumEng's avatar

This was a great read; I've always appreciated your data product mindset and it seems even more critical now with AI in the mix. Looking forward to reading more!

Harjaap's avatar

Technical investment compounds for six months, then plateaus because the org was never redesigned around the new shape of work. That matches what I see. The metrics are all tool-side: tokens, throughput, accepted lines. None of them says whether the downstream team can absorb what got produced. Faster tickets feed the same review queue, the same release process, the same approver. I would add one measure to the technical set: where does the extra output wait? If you cannot name the queue, the speedup is inventory, not value. Operating model is the right name for the other half, and it needs an owner who can change how work is decided, not just report on it.

Liam Croash's avatar

This describes how my approach has changed in the last few months. Scoping and validation are the 2 largest components. Analyses are not built for someone, they're built with someone. It seems like we're becoming context curators and creativity enablers. Its a weird new world but its fun to experiment with different stakeholder experiences curated for how they like to receive information.