Product Management Frameworks

How modern tech companies decide what to build, for whom, and in what order. The role that bridges customer needs, business goals, and engineering capability.


What is Product Management?

Product Management (PM) emerged as a formal discipline at P&G (1930s brand managers), evolved through tech companies (Microsoft, Apple, Google), and became the central role in modern SaaS companies.

The PM’s job: Define and deliver the right product to the right users at the right time.

What PM is NOT:

  • Not a mini-CEO (common misconception)
  • Not someone who tells engineers what to build (that’s a manager/tech lead)
  • Not a project coordinator (that’s a Scrum Master)

What PM IS:

  • The person who understands customer problems deeply
  • The person who sets priorities and explains why
  • The bridge between business, user, and engineering

As a founder, you ARE the PM for all your products. Knowing these frameworks prevents you from building things users don’t want.


The Product Trio (Marty Cagan’s model)

Every product decision requires three perspectives:

  1. Valuable — does the customer want it? (PM’s job)
  2. Viable — does it fit our business model? (PM + Business)
  3. Feasible — can we build it? (Engineering)
  4. Usable — can users actually use it? (Design/UX)

The PM ensures all four questions are answered before committing to build.


Discovery vs Delivery

DUAL-TRACK AGILE
────────────────────────────────────────────────────
Discovery Track     ←→    Delivery Track
(Continuous)              (Sprints / Waves)

→ User interviews         → Building validated features
→ Prototype testing       → Testing and QA
→ Data analysis           → Deployment
→ Market research         → Monitoring
→ Assumption testing      → Iteration based on data

Goal: Never enter Delivery             Goal: Ship reliably
without validated learning             what Discovery approved
────────────────────────────────────────────────────

Biggest PM mistake: skipping Discovery and going straight to Delivery. This creates features users don’t use. Your GymOS is correctly parked — no Discovery signal yet.


The Product Discovery Process

Step 1: Opportunity Assessment

Before any feature, answer:

  1. What problem are we solving?
  2. For whom specifically?
  3. How big is this problem? (urgency, frequency, cost of not solving)
  4. How will we know we’ve solved it?
  5. What’s the alternative if we don’t build this?

Step 2: User Research Methods

MethodBest forHow
User interviewsUnderstanding problems, context, motivations30–45 min conversations, open-ended questions
Usability testingFinding UX problemsObserve users completing tasks on prototype
SurveysQuantifying what you already suspect5–10 questions, large sample, structured
Contextual inquiryObserving real work environmentWatch user in their natural context
Jobs to be Done (JTBD)Finding the “job” the product is hired for”When I [situation], I want to [motivation], so I can [outcome]“
A/B testingComparing two options with real usersStatistical significance required
AnalyticsUnderstanding actual behaviorFunnel analysis, retention cohorts

Step 3: Opportunity Sizing

  • TAM (Total Addressable Market): entire market if 100% captured
  • SAM (Serviceable Addressable Market): segment you can realistically reach
  • SOM (Serviceable Obtainable Market): realistic target in 1–2 years

For DigiTPME: TAM = all Moroccan SMEs (~600,000). SAM = tech-aware SMEs with 10–50 employees in Casablanca/Rabat (~15,000). SOM = 150–300 in Year 1.


Prioritization Frameworks

RICE Scoring

RICE = (Reach × Impact × Confidence) / Effort

FactorDescriptionScale
ReachHow many users affected per quarterNumber (e.g., 200 users)
ImpactHow much does it improve the metric?3=massive, 2=high, 1=med, 0.5=low, 0.25=minimal
ConfidenceHow sure are we?100%=high, 80%=med, 50%=low
EffortPerson-months to buildNumber (e.g., 2 = 2 person-months)

MoSCoW

  • Must have: without this, the product fails
  • Should have: important but not blocking
  • Could have: nice to have if time permits
  • Won’t have: explicitly out of scope (this time)

Kano Model

CategoryDescriptionExample
Basic needsUsers expect it; absence = dissatisfied; presence = neutralLogin/auth in any app
Performance needsMore = more satisfiedFaster lab results in SijiLab
Excitement needsUnexpected delight; absence = neutralAI-powered anomaly explanation in IEIA
IndifferentUsers don’t care either wayInternal implementation details

Opportunity Scoring (JTBD + Ulwick)

Survey users: “How important is [outcome] to you?” and “How satisfied are you with current solutions?” Opportunity Score = Importance + max(Importance – Satisfaction, 0) High importance + low satisfaction = high opportunity.


The Product Roadmap

What it is

A communication tool (not a project plan) that expresses product strategy over time. Shows direction, not a detailed schedule.

Roadmap Types

TypeBest forTimeframe
Now / Next / LaterEarly-stage startupsFlexible
Quarterly themesMid-stage SaaS3-month horizons
Feature-date roadmapCommitted enterprise contractsSpecific dates
Outcome-basedOKR-driven teamsQuarterly outcomes
NOW (this quarter)
→ SijiLab Wave 3 (Billing, Invoice, Quality modules)
→ DigiTPME diagnostic questionnaire UI

NEXT (next quarter)
→ SijiLab Wave 4 (Reports, Portal)
→ DigiTPME PDF report generation
→ SijilPharma ML fix (overfit R² from 0.31 to 0.7+)

LATER (6+ months)
→ DigiTPME benchmark (cross-SME comparison)
→ SijiLab multi-lab (multi-tenant)
→ NOTQIN IEIA external pilot

Anti-patterns

  • Feature factory roadmap: packed with outputs, no outcomes
  • Waterfall disguised as roadmap: rigid dates 12 months out
  • Roadmap as commitment: stakeholders treat it as a contract

Jobs to be Done (JTBD) Theory

Author: Clayton Christensen (Harvard, 2003), refined by Bob Moesta and Alan Klement.

Core insight: Customers don’t buy products — they hire products to do a job in their lives.

Classic example: McDonald’s milkshakes. Research found customers hired milkshakes for their morning commute (long, thick, can sip slowly, one hand, not messy). Competitors were other milkshakes. But the real competition was… bagels, bananas, and granola bars. The job was “keep me busy and not hungry during a boring 30-minute drive.”

JTBD Interview Questions

  • “Take me back to the day you first realized you needed [product]. What was happening?”
  • “What were you doing before you found [product]? What was painful about it?”
  • “What made you finally decide to switch / buy?”
  • “What would you use instead if [product] disappeared tomorrow?”

Applied to SijilPharma

The job pharmacy owners hire ERP software for is NOT “manage inventory” — it’s “not get caught with expired drugs and be able to close my books at month-end without errors.” This reframing changes what features matter most.


Product-Led Growth (PLG)

Definition: A go-to-market strategy where the product itself — not sales and marketing — drives user acquisition, conversion, and expansion.

Examples: Slack, Notion, Figma, Canva, Dropbox.

PLG Mechanics

FREE / FREEMIUM → AARRR funnel
    Acquisition: user discovers product without salesperson
    Activation: user experiences "aha moment" quickly
    Retention: product is useful enough to come back
    Revenue: user hits limit → upgrades to paid
    Referral: user invites others (viral loop)

Aha Moment

The moment the user first gets the core value. Must happen as early as possible.

  • Slack: send your first message to a colleague and they respond instantly
  • Dropbox: see your file on multiple devices
  • SijiLab: see a patient’s complete lab report generated automatically in 60 seconds

Design your onboarding flow to reach the Aha Moment in <5 minutes.

PLG Metrics

  • Time to Value (TTV): time from signup to first Aha Moment (lower = better)
  • Product Qualified Lead (PQL): user who has experienced enough value to be sales-ready
  • Activation Rate: % of signups who reach Aha Moment
  • Feature Adoption Rate: % of users using a specific feature

The North Star Metric

One single metric that best captures the core value your product delivers to users. Everything else is either a leading indicator (inputs) or lagging indicator (outputs) of the North Star.

Examples:

CompanyNorth Star Metric
AirbnbNights booked
SpotifyTime listening per user per day
SlackDaily active users who sent messages
WhatsAppMessages sent per day
FacebookDAU (Daily Active Users)

For your products:

ProductNorth Star Metric
SijiLabLab requests processed per month
SijilPharmaPrescriptions dispensed per pharmacy per month
DigiTPME DiagnosticDiagnostics completed and converted to paid action
NOTQIN IEIAEnergy anomalies identified + acted upon

Product Analytics Stack

LayerToolPurpose
InstrumentationMixpanel, Amplitude, PostHog (OSS)Track user events and flows
Funnel analysisFunnel charts in AmplitudeWhere do users drop off?
Retention analysisCohort tablesWho comes back after Day 1, 7, 30?
Session replayFullStory, HotjarWatch real user sessions
Error monitoringSentryTrack client-side errors
Feature flagsLaunchDarkly, Unleash (OSS)A/B test and gradual rollouts

For your current scale: PostHog (open source, self-hosted) covers everything. Add it to SijiLab and DigiTPME diagnostic from day one.


Product Spec Formats

PRD (Product Requirements Document) — what you already use

Captures: problem statement, user stories, functional requirements, non-functional requirements, acceptance criteria, open questions.

One-Pager

Shorter: problem, solution, why now, success metrics, risks. Used for quick alignment before writing a full PRD.

User Story Map

Visual: personas × journey stages × stories. Helps see the whole product as a user flow.


Connecting to Your Role

As a founder-PM-engineer:

  • Wear the PM hat during Discovery (interview users, shape features, say no)
  • Wear the engineering hat during Delivery (write code, review PRs)
  • The danger: you’ll be tempted to skip Discovery and build what you imagine users want
  • The discipline: minimum 2 user interviews per wave before writing any code for new features

See Also