Lean Startup & Business Model Canvas

The two foundational frameworks for building new products under conditions of extreme uncertainty. Essential for every DigiTPME product.


Part 1: Lean Startup (Eric Ries, 2011)

The Core Problem Lean Startup Solves

Traditional management (Fayol’s POCCC, MBA frameworks) assumes you know your customers, market, and product. This works for executing a known business model.

Startups face a different reality: everything is a hypothesis. You don’t know:

  • Who your real customer is
  • What problem they actually have
  • What solution they’ll pay for
  • What price they’ll pay
  • What channel reaches them

The Lean Startup answer: treat a startup as a scientific experiment. Form a hypothesis → test it cheaply → learn → update hypothesis → repeat.


The Build-Measure-Learn Loop

                 ┌─────────────┐
    ┌───────────▶│    BUILD    │◀──────────────┐
    │            │  (MVP)      │               │
    │            └──────┬──────┘               │
    │                   │                      │
    │                   ▼                      │
  LEARN          ┌─────────────┐            IDEAS
    │            │   MEASURE   │            (Hypotheses)
    │            │ (data+KPIs) │               │
    │            └──────┬──────┘               │
    │                   │                      │
    │                   ▼                      │
    └───────────◀│   LEARN    │◀──────────────┘
                 │(validated  │
                 │ learning)  │
                 └─────────────┘

Minimize the time through this loop.

Validated Learning: the real output of a startup experiment. Not just “we built X” but “we learned Y about our customers with statistical confidence.”


What is an MVP?

MVP = Minimum Viable Product

NOT the smallest possible version of your product. It is the minimum version that lets you test your riskiest assumption about your business.

Riskiest assumption first: identify what would kill your business if wrong. Test that first.

MVP TypeHow it worksExample
Concierge MVPDo the service manually, pretend it’s automatedDigiTPME: manually create diagnostic report for first 5 clients before building the platform
Wizard of Oz MVPUser thinks it’s automated; you do it manually behind the scenesAI recommendation engine run manually by Ahmed
Landing page MVPDescribe product before building it; measure signups/interestDigiTPME landing page + “join waitlist” before any code
Prototype MVPNon-functional mockup for usability testingFigma prototype of SijiLab patient portal
Single-feature MVPOnly core feature, no extrasSijilPharma with only inventory management first
Pilot/Beta MVPReal product, limited users, high-touch supportSijiLab beta with 1 real lab
Smoke test MVPSell it before it exists; measure willingness to payOffer NOTQIN IEIA pilot contract before full build

Your current position: NOTQIN IEIA at Wave 20 is NOT an MVP — it’s a full product. The question is whether you validated the problem before building. If not, run discovery interviews NOW before Wave 21.


The Riskiest Assumption Test (RAT)

Better name for MVP in many cases. Answer:

  1. What is the #1 assumption that, if wrong, would kill this business?
  2. What is the cheapest, fastest way to test it?
  3. What data would convince you the assumption is true/false?

Example — DigiTPME Diagnostic Platform:

AssumptionRiskTest
Moroccan SME owners will complete a 30-min diagnosticHigh — they may be too busyBuild a 5-question version first, measure completion rate
They will pay for the detailed reportCriticalAdd a “buy detailed report for MAD 500” button on free version, measure clicks even before building the paid version
The maturity model is relevant to their actual problemsHighInterview 10 SME owners, ask what they’d want in a diagnostic

Pivot or Persevere?

Pivot: structured course correction to test a new fundamental hypothesis.

Pivot TypeDescriptionExample
Zoom-inOne feature becomes the whole productSijilPharma: maybe only the pharmacy inventory module has PMF
Zoom-outFeature is too small; product needs to be broaderSijiLab adding pharmacy integration
Customer segmentSame product, different customerDigiTPME diagnostic for industrial SMEs instead of retail
Business architectureSwitch from high-margin-low-volume to low-margin-high-volume (or vice versa)From consulting to SaaS
Value captureChange monetization modelFrom subscription to usage-based
ChannelDifferent distribution pathFrom direct sales to consultant referrals
TechnologyRebuild with different tech stack
PlatformFrom application to platformDigiTPME diagnostic becomes marketplace

Persevere: when data shows you’re on the right track, keep going.

The hardest question: when to pivot (many founders persevere too long on a dead hypothesis, or pivot too early on a good one). The answer comes from validated learning, not intuition.


Innovation Accounting (measuring Lean Startup progress)

Traditional accounting doesn’t work for startups (no revenue yet). Innovation accounting tracks:

  1. Establish baseline: current state of each metric (conversion rate, retention, etc.)
  2. Tune the engine: experiments to improve each metric
  3. Pivot or persevere: if metrics aren’t improving, reassess the hypothesis

Cohort analysis: don’t look at aggregate metrics. Track behavior of users who joined in the same period (a “cohort”) to see if each new cohort performs better than the last.


Part 2: Business Model Canvas (BMC)

Author: Alexander Osterwalder & Yves Pigneur — Business Model Generation (2010)

The BMC maps an entire business model on one page using 9 building blocks.


The 9 Building Blocks

┌──────────────────────────────────────────────────────────────────────────┐
│  KEY PARTNERS   │  KEY ACTIVITIES │ VALUE         │ CUSTOMER    │CUSTOMER│
│                 │                 │ PROPOSITIONS  │ RELATIONS   │SEGMENTS│
│  Who helps us?  │  What do we do? │               │             │        │
│                 │                 │ What problem  │ How do we   │Who are │
│                 ├─────────────────│ do we solve?  │ interact    │we for? │
│                 │  KEY RESOURCES  │ What value do │ with them?  │        │
│                 │                 │ we deliver?   ├─────────────│        │
│                 │  What do we     │               │  CHANNELS   │        │
│                 │  need?          │               │             │        │
│                 │                 │               │How do we    │        │
│                 │                 │               │reach them?  │        │
├─────────────────────────────────────────────────────────────────────────┤
│  COST STRUCTURE                   │  REVENUE STREAMS                     │
│  What are our major costs?        │  How do we make money?               │
└──────────────────────────────────────────────────────────────────────────┘

BMC Applied: SijilPharma

BlockContent
Customer SegmentsIndependent Moroccan pharmacies, 1–3 staff, Casablanca/Rabat, paper/Excel-based today
Value PropositionsReplace paper tracking: inventory management, prescription dispensing, expired drug alerts, monthly closing without errors
ChannelsLinkedIn, referrals from pharma wholesalers, direct sales demos
Customer RelationshipsFree 30-day trial, WhatsApp onboarding support, monthly NPS survey
Revenue StreamsMonthly subscription MAD 199 (Starter) / MAD 399 (Pro). Annual plan at 2 months free.
Key ResourcesReact + Spring GraphQL codebase, Ahmed’s pharma domain knowledge, ML model for demand forecasting
Key ActivitiesSoftware development, customer onboarding, ML model maintenance, customer support
Key PartnersPharmacie du Maroc (regulatory), pharma wholesalers (distribution channel), payment processor (CMI)
Cost StructureAhmed’s time, server hosting (VPS/cloud), payment processing fees, marketing spend

BMC Applied: DigiTPME Diagnostic Platform

BlockContent
Customer SegmentsMoroccan SMEs (10–50 employees), owners/GD seeking digital transformation, gov-aligned
Value Propositions”Know exactly where your business is digitally and what to do next.” Sector-specific maturity score + actionable roadmap aligned with Morocco Digital 2030
ChannelsFree online diagnostic (self-serve), LinkedIn thought leadership, CRI/ADEREE partnerships, consultant referrals
Customer RelationshipsFreemium (free basic), paid advanced report, consultant marketplace
Revenue StreamsFree diagnostic → MAD 500 detailed report → MAD 2,000 consulting sprint → MAD 500/mo platform subscription → Gov aggregate dashboard MAD 50k/year
Key ResourcesMaturity model (8 dimensions), NestJS+React platform, AI recommendation engine (Claude API), Ahmed’s SME knowledge
Key ActivitiesProduct development, content marketing, SME interviews (discovery), report generation, government partnership
Key PartnersCRI (Centre Régional d’Investissement), ADEREE, ANPME, DigiTPME program (MCNET), consulting firms (resellers)
Cost StructureDevelopment, hosting, AI API costs (Claude), content creation, government partnership development

BMC Applied: NOTQIN IEIA

BlockContent
Customer SegmentsMoroccan textile factories 50–500 employees, EU export pressure, energy cost > 10% COGS, CBAM-impacted
Value Propositions”From sensor to certificate.” Real-time energy anomaly detection, CBAM compliance data, MAD-priced scenario analysis, automated energy audit
ChannelsDirect LinkedIn outreach to factory directors, CGEM events, AMEE partnership, OCP Usine du Futur network
Customer Relationships3-month paid pilot → annual contract, dedicated customer success, quarterly energy reviews
Revenue StreamsMAD 5,000/mo pilot → MAD 15,000/mo annual contract per factory. CBAM reporting add-on.
Key ResourcesNOTQIN Core IoT platform, IEIA FastAPI + Claude API backend, Energy engineer domain expertise, ONEE tariff data, CBAM regulation knowledge
Key ActivitiesPlatform development, pilot deployment (physical sensor installation), customer success, regulatory tracking (CBAM), research papers
Key PartnersONEE (energy data), AMEE (efficiency agency), sensor vendors (Crouzet, ThreadCloud), ENSA Berrechid (talent), CBAM legal consultants
Cost StructureHardware (ESP32, sensors), cloud hosting, Claude API, Ahmed’s time, pilot deployment costs, travel to factories

Value Proposition Canvas (Osterwalder’s zoom-in)

Zooms into the Customer Segment ↔ Value Proposition relationship.

CUSTOMER PROFILE                    VALUE MAP
┌─────────────────────┐    FIT?    ┌─────────────────────┐
│  Customer Jobs      │  ←──────▶  │  Products & Services │
│ (what they try to   │            │ (what you offer)     │
│  accomplish)        │            │                      │
├─────────────────────┤            ├─────────────────────┤
│  Pains              │  ←──────▶  │  Pain Relievers      │
│ (frustrations,      │            │ (how you eliminate   │
│  obstacles, risks)  │            │  pain)               │
├─────────────────────┤            ├─────────────────────┤
│  Gains              │  ←──────▶  │  Gain Creators       │
│ (outcomes, desires, │            │ (how you produce     │
│  aspirations)       │            │  desired gains)      │
└─────────────────────┘            └─────────────────────┘

FIT happens when your Pain Relievers address real Pains, and your Gain Creators produce desired Gains.


Lean Canvas (Ash Maurya’s adaptation for startups)

Replaces 4 BMC blocks with startup-specific blocks:

ReplacedWith
Key PartnersProblem (top 3 problems you’re solving)
Key ActivitiesSolution (top 3 features)
Customer RelationshipsUnfair Advantage (what can’t be copied)
Key ResourcesKey Metrics (what you measure)

Lean Canvas is more useful than BMC at the idea/pre-PMF stage.


Connecting Lean Startup + BMC to Your Operating Context

From your vault’s operating rules:

“When evaluating a new idea: Does it produce a case study, reusable component, or sellable diagnostic? Can it be delivered in a bounded wave?”

This IS Lean Startup thinking applied. Bounded wave = time-boxed experiment. Case study = validated learning. Sellable diagnostic = revenue test.

Your riskiest unvalidated assumption today:

“Moroccan SME owners will complete a digital maturity diagnostic, find it valuable, and pay for a detailed report.”

Cheapest test: a Google Form with 8 questions + a MAD 200 PayDunya link at the end. If 10/50 people click “buy the full report,” you have signal. Build the platform after that.


See Also