← All essays

The €220k a month nobody was counting

The client did not ask for an audit. The delivery data asked for it. We had been an embedded partner for over three years; releases went out on schedule, the board saw steady activity, and nobody was alarmed. That is precisely what made it dangerous.

Activity is the best disguise a leak can wear

When delivery is visibly broken, someone owns the problem. When delivery is busy, nobody does. Here, tickets closed and burndowns burned down, but when we asked one question per backlog item, "which business capability does this serve?", more than 90% of user stories had no answer. The work was real. Its value was unaccounted.

Priced against the commercial roadmap, the unaccounted work and the outcomes it displaced added up to roughly €220,000 a month of visible leakage, with exposure up to €315,000, plus a cost of poor quality running to $400,000 a year.

Velocity tells you the engine is turning. It does not tell you the car is pointed anywhere.

What actually changed

Not headcount. The delivery system was rebuilt around traceability: Ideas → Opportunities → Business Capabilities → Features → User Stories → Sprints → Releases, enforced as a gate, not a guideline. Three metrics replaced activity reporting: First-Time Yield, System Availability, and Work in Progress.

  • Every backlog item now traces to a capability, a roadmap priority, and a commercial outcome.
  • Investment decisions moved to before the work, not after the release.
  • Leadership reporting switched from "what we did" to "what it was worth."

If your dashboards are green and your roadmap still feels like a wish list, the leak is probably not in the code. It is in the connection between the code and the business. That connection is buildable, and it is worth building before it costs you a fifth of a million a month.

Was this useful?

Related serviceBusiness Case DevelopmentWe analyze your product, delivery system, market context, and technical foundation to uncover hidden revenue, savings, and growth opportunities, then turn them into a business case your board can act on.Explore Business Case Development

Related essays

LeadershipWhy rescues fail: the firefighting layerThe most common failure mode in troubled projects is not technical. It is the leadership team becoming the coordination layer.5 minDeliverySoftware Project Rescue: How to Know Whether a Project Can Be SavedA six-factor decision framework — business value, remaining work, architecture, team capability, vendor risk, time to market — for deciding whether to rescue, rebuild, replace, or stop.20 min

Related Expertise and Services

Engineering Operating System
For teams that have worked together for years but where progress feels slow and messy. We install the operating system: ownership, cadence, quality gates, and a straight line of sight from strategy to sprint to release.
Project Rescue
When deadlines slip, releases break, or delivery slows down without a clear reason, we step in to find what is blocking progress and quickly fix the highest-impact constraints across strategy, operations, and technology.
Architecture Transformation
Legacy pressure, modernization, due-diligence findings, or a platform that cannot carry the roadmap. We map the real risk, design the target architecture, and sequence the migration so production keeps shipping while the foundation changes underneath it.