Where should we invest?
Priorities, capacity and the cost of another commitment.
Writing
On the decisions that sit between delivery data and business value.
Essays for technology leaders on investment, operating models and AI adoption, with the flow and measurement detail underneath for those who want it.
33 essays since Aug 2025
Latest · 6 Sept 2026 · 5 min
From AI productivity to economic value
An engineering team reports that AI made part of its work faster. What can the organisation legitimately conclude from that?
Previous: The next investment decision is also a capacity decision · 31 Jul 2026
Start with the question
What prevents implementation?
Operating models, decision rights and organisational change.
What actually changed?
Evidence that can support the next decision.
What benefit can we defend?
The distance between local productivity and business value.
Longer form: The Predictability Loop (Apress, 2027) · The Art of Creating Self-Organizing Teams
The library
All essays
- From AI productivity to economic valueWhat benefit can we defend?5 min
- The next investment decision is also a capacity decisionWhere should we invest?5 min
- Why your team stopped reporting problemsWhat actually changed?8 min
- Your AI strategy is incoherent — here’s how to tellWhat prevents implementation?7 min
- Copilot made your velocity number meaningless. Now what?What actually changed?7 min
- Before a forecast becomes a commitmentWhat actually changed?3 min
- Choose the intervention the evidence supportsWhat benefit can we defend?3 min
- You don’t have an Agile problem. You have a power problem.What prevents implementation?9 min
- Your team isn't slow. Your system is.Where should we invest?8 min
- Five stages your delivery process passes through before it breaksWhat actually changed?8 min
- The metrics that protect careers and break delivery systemsWhat prevents implementation?8 min
- Why frameworks failWhat prevents implementation?8 min
- A date is not a forecastWhat actually changed?7 min
- When teams game metrics: what anomaly detection reveals about cultureWhat actually changed?10 min
- When fixing one bottleneck reveals anotherWhat actually changed?3 min
- The delegation problem: why AI made intent the new bottleneckWhat prevents implementation?9 min
- AI generates faster than you can review. That's your real constraint.What benefit can we defend?7 min
- When your historical data stops describing your current systemWhat actually changed?7 min
- The dashboard delusionWhat actually changed?7 min
- Aggregation lies: what your portfolio dashboard isn't telling youWhere should we invest?7 min
- Specification-driven delivery: governing what AI buildsWhat prevents implementation?10 min
- AI coding tools moved the bottleneck. Most teams haven't noticed.What benefit can we defend?8 min
- Resistance is relational, not just personalWhat prevents implementation?10 min
- The seven failure modes of a stuck delivery systemWhat actually changed?9 min
- Process behaviour charts for people who hate statisticsWhat actually changed?8 min
- The one equation that explains most delivery slowdownsWhere should we invest?8 min
- Percentiles, not averages: why your forecasts are wrongWhat actually changed?7 min
- Your improvement worked. Or did it?What actually changed?8 min
- The promise you can actually keepWhat actually changed?7 min
- Your delivery system is running on debtWhere should we invest?8 min
- Brownfield first: why you can’t just let AI rewrite your codebaseWhat prevents implementation?9 min
- Portfolio flow: why your fastest teams are hiding a shared problemWhere should we invest?8 min
- AI is an amplifier — your engineering culture determines your ROIWhat benefit can we defend?8 min
Get new essays by email
New essays, books and Flow Observatory news. Occasional, and you can unsubscribe at any time.