Product Management

Product management owns the what and why of a product — which problem to solve for which users and why it matters to the business — as distinct from Project Management, which owns the how/when/who of delivering a defined piece of work. A product manager can (and often does) work with a project or program manager rather than being one.

Discovery vs. delivery

• Discovery — figuring out what to build: user research (see User Centred Design), validating a problem is real and worth solving, testing a solution concept cheaply before committing engineering time.
• Delivery — actually building and shipping the validated solution.

The common failure mode is skipping discovery and treating every stakeholder request as a delivery item — a fully-staffed, well-run delivery process building the wrong thing is still a failure, just an efficient one.

Strategy and roadmaps

• A product strategy states the target market/user, the problem being solved, and how winning is defined — the filter every roadmap decision should trace back to.
• A roadmap is not a project schedule with fixed dates for fixed features; the more useful form communicates themes/outcomes and rough sequencing ("now / next / later"), leaving room for discovery to change the specific solution while the outcome stays fixed. A roadmap that's actually a Gantt chart of features has usually stopped being a product artifact and become a delivery commitment.

Prioritization frameworks

• RICE (Reach × Impact × Confidence ÷ Effort) — a scoring model forcing explicit numbers for each factor, useful for comparing otherwise-incomparable ideas on a common scale.
• MoSCoW (Must / Should / Could / Won't) — a simpler categorical framework, common in scoped, deadline-driven delivery contexts.
• Kano model — classifies features as basic (expected, absence causes dissatisfaction), performance (more is better, roughly linear satisfaction), or delighter (unexpected, high satisfaction impact) — useful for noticing that piling on more "performance" features has diminishing returns while a missing "basic" feature is disproportionately damaging.
• No framework replaces judgment about strategic fit — a low-RICE-score item that's a hard regulatory requirement still ships.

Success metrics

• North Star metric — one metric that best captures the value a product delivers to users (not just revenue), that the team rallies around; supporting metrics explain movement in it.
• OKRs (Objectives and Key Results) — a qualitative Objective paired with a small number of measurable Key Results; commonly used to connect product work to business outcomes at the level a roadmap theme should also trace to.
• Beware vanity metrics (raw signups, page views) that move without reflecting real value delivered — prefer metrics tied to the actual problem the product claims to solve (activation, retention, task success).

Common pitfalls

• Feature factory mode — measuring product management success by output shipped rather than outcomes achieved; a roadmap full of shipped features that didn't move the North Star metric is a warning sign, not a win.
• Treating every customer request as a requirement — a specific request is evidence of an underlying need, not necessarily the right solution; the underlying need is what discovery should validate.
• Confusing a roadmap with a promise — committing specific dates for specific unvalidated solutions removes the flexibility discovery is supposed to provide, and sets up a broken promise the moment discovery reveals a better approach.
• No prioritization criteria, only politics — whoever's loudest or most senior wins the backlog fight; an explicit framework (even an imperfect one) gives a shared, defensible basis for saying no.

Learn more

• Marty Cagan: Inspired — widely-referenced practitioner text on the discovery/delivery model.
• Intercom on Product Management — practical prioritization/roadmap guidance.
• Project Management for the delivery-execution discipline product management partners with.
• User Centred Design for the research methods discovery depends on.