In the previous module I explained how KPIs kill software businesses.
The fuller truth is this: the wrong KPIs at the wrong stage are a symptom of a poorly constructed mission. And a poorly constructed mission is almost always a vision wearing a mission’s clothes.
If every item in your portfolio shares the same optimised measures, your mission is wrong.
Mission is the enduring reason an organisation exists. Vision is how you pursue it in an evolving world — shaped by technology shifts, regulatory change, market dynamics, and competitive pressure. Most software businesses conflate the two, and the consequences run in opposite directions depending on which way the conflation falls.
When vision changes
Measures reset. A new product development lifecycle has begun. The organisation recognises that the application of its mission to a specific moment has shifted and recalibrates accordingly.
This is the healthy signal.
When vision inverts to mission
When it takes the crown it has not earned, measures never change. The board loses the thread. Quality and velocity fill the gap.
Both are appropriate measures for what they govern. Neither is a governing instrument for the portfolio. Their presence at board level is a symptom of poor product management.
The Calcification
Many do not even get that far. When vision dresses in mission’s crown it does not change. It calcifies. The measures calcify around it. The product becomes the backlog, the backlog becomes the strategy, and the strategy becomes a more elaborate version of what already existed.
Nobody asks the mission question because nobody separated the mission from the vision in the first place.
When Agentic AI Arrives
When agentic AI arrives, it does not improve the software. It replaces the layer the software was built to govern. Most software exists in the administrative and operational space between humans and processes:
Every point where a person had to interact with a system to move something forward. Agentic AI absorbs that layer entirely. The interactions no longer require a human in the loop, which means they no longer require the platform that was built to manage them.
The businesses that conflated mission with vision spent years making that layer more elaborate. More configurable. More feature-rich. Strategic purpose compressed into a backlog of enhancements to something that can now be replaced. That is the SaaSpocalypse. Not a technology disruption. ~$285 billion in enterprise software market capitalisation — capital recognising that the kingdom it had been funding was a more sophisticated version of a layer that was about to disappear. Not because the business failed to move fast enough. Because it never built anything above the moat it was defending.
The Governing Sequence
Here is the sequence that actually governs this.
Mission
Defines the destination. Fixed, enduring, independent of market conditions or product roadmap.
The Mission Gap
The gap between where the organisation currently is and where the mission requires it to be determines what must be measured. Not the OKR framework. Not the sprint review. The mission gap.
The Stage
That gap, mapped against the lifecycle and maturity stage of each product in the portfolio, calibrates how it is measured. The stage does not determine what must be measured. The mission does that. The stage determines the instruments appropriate to measuring it.
This distinction matters more in software than anywhere else, because software is the only industry that never institutionalised it. No framework exists in most software organisations that separates two structurally different activities: managing the lifecycle of what already exists, and developing what must exist to serve the mission. Without that separation, everything looks like delivery. The board watches velocity and calls it governance.
Two Populations. One Framework.
A new feature is an in-life enhancement. It belongs in the PLM cycle. Delivery speed is a legitimate measure for it. Genuine new product development is governed by whether what is being built closes the mission gap at its current stage of maturity. Applying delivery metrics to that question does not produce a wrong answer. It produces no answer at all.
The New Product Development Framework enforces the distinction structurally. The moment you have a stage address for every item in the portfolio, you have separated what is in-life from what is genuinely new. Those two populations require different measures, different investment logic, different authority levels, and different board-level reporting.
The critical distinction
Velocity is a legitimate PLM measure. It is an actively dangerous NPD measure — because it tells the board that things are moving without telling them whether what is moving serves the mission.
One population supports an existing moat. The other builds a bigger, more defensible one.
The Proof
Nokia
Excellent delivery metrics. Handsets shipped. Cycle times strong. Carrier relationships healthy. No board-level measure ever asked whether the development pipeline still served why the business existed.
When it did not, there was nothing in the pipeline to replace it. The velocity had been real. The mission governance had been absent.
Apple
Ran a fifteen-year chip programme that no PLM metric would have approved. At every stage gate, one question: does this still advance our capacity to create tools that extend human capability through the integration of technology and design?
That is a mission measure. It does not live in a sprint board. It lives in the gate, and it protected a fifteen-year investment through conditions that would have killed it anywhere without that discipline.
The Diagnostic
The mission determines what must be measured. The New Product Development Framework determines the stage. The stage determines the appropriate measures. The gate uses all three to decide whether anything moves.
There is a diagnostic buried in this. When every item in a portfolio shares the same optimised measures, it is not a sign of consistency. It is a sign that the mission is wrong. Everything has been flattened to the same lifecycle stage because nothing has genuinely left it. The entire estate is in-life enhancement. The backlog is the strategy.
If that is where you are, you are not a product business.
You are a vision business. And vision businesses, as the last two years have made abundantly clear, do not last.
#ProductLeadership #CPO #ProductStrategy #ProductManagement #NPD #Innovation #EnterpriseTransformation #SaaS #AI
Peer Validation Report
M1b — Validated against Cagan, Martin, Christensen, Bain and IDC
© John Bowers 2026. All rights reserved. The Four Ms operating model, the New Product Development Framework, and the Architectural Formula (X×Y×Z×WN) are proprietary methodologies. The Product Leader’s Playbook is an original work in progress. Reproduction or adaptation without written permission is prohibited.
First Comment
The peer review attached to this piece is not decoration. It establishes that the argument connecting mission/vision governance failure to maximum agentic AI exposure does not exist in the published literature in this form.
Marty Cagan sees the conflation problem but not the AI consequence. Bain and IDC see the disruption but not the governance cause. Roger Martin’s Strategy Choice Cascade implies the conclusion but does not state it. Clayton Christensen frames the incumbent trap as a resource allocation problem rather than a mission governance problem — a distinction that matters because the interventions are different.
The NPD Framework is the intervention. Stage governance separates in-life enhancement from genuine new product development structurally, giving the board two populations with different measures, different investment logic, and different gate authority. Without that separation, everything looks like delivery and velocity becomes the proxy for strategic progress.
That is the architecture the SaaSpocalypse revealed as absent. Module 2 of this series covers it in full.