Module-2 Experiment Design & MVP Testing

Blue Ocean vs. Red Ocean

Blue Ocean and Red Ocean are metaphors for market spaces. Red Ocean represents all existing industries — crowded, competitive, with companies fighting over the same customers (like sharks in bloody water). Blue Ocean represents unknown market space — uncontested, with demand created rather than fought over.

The core idea: even in existing businesses, you can create entirely new spaces by shifting focus from beating competitors to creating new value for buyers.

Definition – Blue Ocean: markets that do not exist today. Red Ocean: all industries in existence.

Why the construct matters (intuition)

• Startups often think about “doing something new”, but many blue oceans arise from re-imagining an existing service (e.g., maternity care → a joyous experience, not a clinical one).
• You don’t need a radical technology; you need a different value proposition that makes competition irrelevant.

---

Blue Ocean vs. Red Ocean – key differences

---

Cornerstone: Value Innovation

Value innovation is the central pillar of blue ocean strategy.

• Shift focus from competing to delivering a leap in buyer value.
• Make competition irrelevant by offering something so different that existing rivals become obsolete.

Exam tip: Value innovation ≠ technology innovation. It’s about what buyers value, not what engineers invent.

---

The ERRC Model (Eliminate – Reduce – Raise – Create)

A practical framework to craft a blue ocean:

1. Eliminate – factors the industry has always competed on but no longer add value.
2. Reduce – factors that are over‑delivered relative to what buyers really need.
3. Raise – factors that should be pushed well above industry standard.
4. Create – factors the industry has never offered – this is where new demand comes from.

---

Cloudnine (maternity hospitals)

• Problem: Childbirth is joyful, but hospitals are quiet, scary, clinical places.
• Blue ocean created: A dedicated mother‑child hospital that engages the family from the first trimester – exercises, nutrition, cohort, father included. Created a new category (mother‑child healthcare) where earlier only general hospitals existed.
• Result: Multiple cities now have such specialty hospitals. Cloudnine became one of the largest.

Y Combinator (startup accelerator)

• Before: Incubators in academic institutions, long programs.
• Blue ocean: 3‑month program with smart founders, top mentors, a demonstration day – created the accelerator category.
• Outcome: Changed global entrepreneurship; spawned Airbnb, Dropbox, Reddit, Meesho, RazorPay, etc.

Yellow Tail (wine)

• Problem: Wine labels are complex (grape, vintage, vineyard, tasting notes). Most buyers just want a drink – they find selection confusing.
• ERRC applied:

• What they did: Only red or white wine, slightly sweeter (friendly to casual drinkers), fun marketing (football ads, “uncomplicated”). Made wine selection easy.
• Result: Became the most‑drunk wine in the US for decades.

Other mentions (context only)

• Uber – blue ocean in taxi services.
• Airbnb – blue ocean in accommodation.
• Araku Cafe – value from shade‑grown sustainable coffee; made competition irrelevant.
• Darshini / Rameshwaram (Bangalore) – quick‑service restaurants for South Indian breakfast (stand, eat, go – cheap, fast, hygienic).
• Heritage hotels (Neemrana) – converted palaces/havelis into premium stays; created a new segment before Airbnb.

---

Key takeaways

• Blue Ocean = new markets; Red Ocean = existing, competitive markets.
• Blue ocean strategy centers on value innovation – leap in buyer value that makes competition irrelevant.
• The ERRC framework (eliminate, reduce, raise, create) helps systematically design a new value curve.
• Successful blue oceans often come from rethinking an existing industry (maternity, wine, travel) – not from inventing a new technology.
• Competition is not the enemy; lack of value innovation is.

Categories

Categories are market spaces that can be identified and occupied. Creating a new category is functionally equivalent to finding a blue ocean — a market where competition is irrelevant because the space did not exist before. Identifying categories is critical for opportunity recognition.

Coffee: How declining consumption created a growing market

Between 1950 and 2004, per‑capita coffee consumption in the US fell from 41 gallons/year to 24 gallons/year — nearly halved. Yet coffee sector revenues grew from $28\,billion to$47 billion between 1990 and 2010. The apparent contradiction is resolved by the emergence of new categories that attracted smaller, higher‑paying customer segments.

The evolution of coffee categories is framed as three waves:

Third‑wave coffee has fragmented into countless sub‑categories: single‑source, peaberry, high altitude, shade‑grown, sustainable coffee. Each commands a premium.

Example: Nespresso Nestlé’s Nespresso initially struggled. It succeeded only after positioning as an exclusive, niche experience — hard‑to‑get boutiques, premium pods (~₹100 per pod in India). It created a category of “high‑convenience, high‑quality espresso at home/work”.

Example: Araku Coffee (India) Araku positioned itself as shade‑grown, sustainable coffee — contrast with Brazilian coffee linked to Amazon deforestation. It created a category that appeals to ethically‑conscious, premium buyers. “If you want the best coffee without cutting forests, there’s only Araku.”

How new categories create new markets in other industries

The principle extends far beyond coffee:

Identifying a category opportunity

Ask:

• Which group’s problem is not being addressed by existing offerings?
• How can you create a space where you become the center of that category?

Categories do not require completely novel inventions. Repositioning or bundling attributes (e.g. “first‑wave → third‑wave”, “lollipop → healthy pop”) can carve out a fiercely loyal customer base.

Key takeaways

• A new category is a blue ocean strategy applied to market spaces.
• Categories can emerge even in declining consumption, as coffee illustrates – smaller, higher‑paying segments are the key.
• Successful categories address unmet needs (childcare in incubators) or combine attributes into a superior proposition (sustainable + premium).
• Incumbents in crowded markets (milk, eggs) can be disrupted by redefining the category (ethical, free‑range, traceable).
• To spot categories: study local communities, identify people whose problems are ignored, and imagine what “center of that space” would look like.

MVPs as Antidotes to Failure

Ventures fail not only because of a bad product, but also because they are too early for the market. The classic example is Better Place (2007), which built a network of battery-swapping stations for EVs in Israel. The core assumption – that battery swapping would accelerate EV adoption – was correct, but the timing was so premature that the venture collapsed and set back EV conversations for years. Tesla succeeded later by positioning EVs as luxury, not utilitarian. The lesson: correct assumptions can still lead to failure if the market is not ready.

The Probabilistic Nature of Entrepreneurship

Entrepreneurship operates under uncertainty, not calculable risk. One common belief is that first ventures often fail, and the experience gained makes subsequent attempts more likely to succeed (e.g., second or third ventures at NSRCEL). This makes entrepreneurship a probabilistic game – more attempts increase the odds of eventual success. However, there is a method to mitigate uncertainty systematically, beyond traits like hustle or risk-taking propensity (the myth that only IIT/IIM founders succeed is an "urban legend").

The method involves moving through three stages of fit:

• Problem-Solution fit
• Solution-Product fit
• Product-Market fit

Failure can occur at any stage. Even when a market exists, the solution-to-product transition can fail (e.g., Ziva and Justdial). The goal is to fail fast, fail smart – incur small, recoverable failures that teach quickly.

Exam tip: The "too early" failure mode is distinct from "bad product". Being early means the assumption is correct but the market timing is off – a key nuance in understanding venture failure.

Minimum Viable Product (MVP) as a Learning Tool

An MVP is the simplest version of an idea that allows you to learn something important from customers. It is not a mini version of the final product; it is a low-cost experiment to test one risky assumption at a time. This aligns with the corridor principle – once you enter a specific corridor (path), you may not be able to return, so test cheaply before committing.

MVPs can be used to test any element of the Business Model Canvas (or Lean Canvas): customer channels, pricing, partnerships, value proposition, etc. The focus is on learning, not on launching.

Dropbox MVP Example

Dropbox’s first MVP was a simple explainer video showing how the product would work. The team measured whether people signed up for a waiting list. This validated demand before building the complex product.

Key takeaways

• Ventures can fail because they are too early, even with correct assumptions.
• Entrepreneurship is probabilistic; systematic experimentation improves odds.
• Three fit stages: problem-solution → solution-product → product-market.
• An MVP is a low-cost experiment testing one assumption, not a mini product.
• Corridor principle: test early to avoid irreversible commitments.
• Dropbox’s video MVP validated customer interest without writing code.

Types of MVPs

An MVP (Minimum Viable Product) is the smallest, cheapest experiment that tests a specific hypothesis about a problem, solution, or willingness to pay. Different MVP types de‑risk different assumptions. Entrepreneurs have developed several canonical forms:

---

Landing Page MVP

Presents a simple website or email describing the proposed service and asks for sign‑up / interest. No product is built until demand is confirmed.

• Dropbox – Created a landing page explaining cloud storage with a signup button. Thousands joined the waitlist in 24 hours – validated demand before writing a single line of code.
• Zoojoo.be – Sent PowerPoint decks to HR departments describing a health‑improvement app. Gauged interest without building any software.
• Amagi (Impulse Soft) – Emailed potential clients asking if they needed Bluetooth drivers. Only started development after receiving “yes” replies.
• Zoojoo.be (upgraded) – Used Photoshop mockups to simulate a real app. When a client expressed interest, they asked which features mattered most and built only those first.

Exam tip: A landing page MVP can be as simple as an email or a mockup. The goal is to learn whether the problem is real, not to deliver a polished product.

---

Concierge MVP

The service is delivered entirely manually. The customer experiences the full proposed value while the founder does all backend work by hand.

• Zappos – Nick Swinmurn took photos of shoes at local stores, posted them online; upon order, he bought and shipped the shoes himself.
• Flipkart – Copied book images from the internet; when a customer ordered, founders bought the book from Sapna Book House, wrapped it, and sent it via speed post. They also distributed pamphlets to acquire repeat customers.
• BigBasket – Uploaded grocery pictures; collected orders, bought in bulk at Metro Cash & Carry, sorted manually, and delivered using rented tempos.
• Drop Cafe – Offered free coffee in Koramangala via Facebook posts. When demand surged (200 → 400 people), they tested willingness to pay at ₹80 – demand fell to 75, revealing that customers would pay but expected hot coffee, a logistics challenge.

---

Wizard of Oz MVP

The front end appears fully automated, but all operations are performed manually behind the scenes.

• Early food delivery apps – Customers thought orders were processed by a system, but founders manually called restaurants and coordinated couriers.
• Unifor (course case) – Posted notices asking customers to call; staff entered data manually and returned information, making the process seem automated.

Exam tip: Wizard of Oz is ideal for testing an automated experience before building the technology. The key is to fake the automation convincingly.

---

Paper Prototype

Sketches on paper (or an iPad) used to test user flows and interface design. Observers note where users stumble.

• Fintech apps – Early versions tested flows by placing paper sketches in front of users; cheap, fast, and revealing.
• Modern equivalent: interactive iPad mockups that users can touch.

---

Choosing the Right MVP

Match the MVP to the riskiest assumption:

The core principle: never build a product without testing the riskiest assumption first. Build in conversation with customers, use MVPs to guide decisions, and iterate.

---

Product‑Market Fit: The Ultimate Goal

When even zero marketing spending cannot kill demand, a product has achieved product‑market fit.

• Thumbs Up (Parle, later acquired by Coca‑Cola) – Coca‑Cola tried to kill the brand (stopped advertising, reduced production), but demand remained so strong that they were forced to relaunch it. This demonstrates extreme product‑market fit.

Key Takeaways

• MVPs are experiments, not scaled‑down products. Each type tests a different hypothesis.
• Landing page / email MVPs test demand; Concierge tests solution value; Wizard of Oz tests automated experience; Paper prototypes test usability.
• Never build a full product before testing the riskiest assumption.
• Use the matching framework: problem → landing page; solution → concierge / Wizard of Oz; payment → pre‑orders.
• Product‑market fit occurs when demand persists even with zero marketing – illustrated by Thumbs Up.

Structuring an MVP Testing Experiment

Every MVP test needs a deliberate structure — without it, you get noise, not insight. The goal is to reduce uncertainty as fast and cheaply as possible.

The three essential components

The criteria sets a quantitative threshold. In the Rent the Runway case, the founders didn’t just ask “will people rent?” — they tracked what was rented, how items were returned (condition, timeliness). Each metric fed back into venture design.

Exam tip: A criterion with no number is not a criterion. Always define the threshold before running the test.

Test cards — document your experiment

A test card formalises the three components above into a single sheet that forces clarity and makes the experiment reproducible.

Writing it down prevents memory bias — people tend to remember results that confirm their hopes and forget disconfirming ones. A written card provides evidence others can audit.

Learning cards — turn data into action

A learning card is the reflection step after a test.

• What did we learn?
• Did it confirm, refute, or partially confirm the hypothesis?
• If partial: what revision to the hypothesis is needed?
• What will we change next?

A learning card ensures every experiment leads to a concrete next step — not just a pile of data.

The entrepreneurial learning cycle

Each loop reduces uncertainty. The faster and cheaper the loops, the higher the chance of building a successful venture — this is failing smart: spending little to learn a lot.

Exam tip: The phrase “faster and cheaper your loops” is a core principle of lean startup. It ties directly to the concept of build-measure-learn feedback loops. Be ready to explain why speed and cost matter for survival.

Key takeaways

• Every MVP test must define a variable, a metric, and a success criterion.
• Test cards document the hypothesis, test, metric, and threshold before you run the experiment.
• Learning cards force post-test reflection and a decision (confirm, refute, or revise).
• The hypothesis → MVP test → learning cycle is the engine of scientific entrepreneurship.
• Each iteration reduces uncertainty; aim for fast, cheap loops to maximise learning per unit cost.

Customer Development

Customer development is the process of systematically discovering whether a product solves a real problem that customers are willing to pay for. The core mistake entrepreneurs make is falling in love with the solution instead of the problem – leading to products nobody wants or products introduced too early.

Why Customer Development Matters

• Zero to one vs. one to n: inventing something new (zero→one) does not automatically translate into a scalable business (one→n). Example: Google created search but DoubleClick (acquired) figured out monetisation via AdWords. Similarly, current AI has buzz but hasn’t yet achieved one→n – analogous to the late-1990s internet euphoria before the dot-com bust.
• Peter Drucker: “The purpose of a business is to create a customer.” No customers = no business.
• In a startup, no facts exist inside the building – only opinions. Validation must come from customers, not from friends, teammates, investors, or market research. Data trumps opinion.

Exam tip: The phrase “no facts inside the building” is a hallmark of Steve Blank’s customer development methodology – expect it in exam questions about early-stage validation.

The Edison Phonograph: A Classic Caution

Edison invented sound recording (the phonograph) but had no idea what to do with it. His ideas:

• Record dying words for remembrance.
• Audiobooks for the blind.
• Answering machine for telephones.
• Music recordings – though he worried artists would lose income.

This shows that even a breakthrough invention requires customer development to find a viable application.

Core Principles

The Customer Development Process (Steve Blank)

1. Identify initial customers – determine if the opportunity is important to them.
2. Conduct qualitative interviews – open-ended, no pitching.
   • Do not ask: “Would you like to buy if something like this existed?” (yields false positives).
   • Ask: “What are the challenges you face? How are you currently solving it?”
3. Use qualifying questions – filter for actual target customers. Example: For a left-handed product, ask “Are you left-handed?”.
4. Seek proof – get an email, a commitment to buy, or an expression of interest (“Let me know when you launch”).
5. Close with a referral – “Do you know anyone else with this problem?” A referral is a form of validation.

Interview Techniques

• Open-ended questions only – get them talking. Example: “What’s hard about being a freelancer?”
• Flip to advice – “As an expert, what do you think?” (reduces perceived selling).
• Record if permission is given (e.g., “as part of an IIMB program”).
• Use ChatGPT (or similar) to refine question lists – check for yes/no traps.

Exam tip: Validation ≠ asking “Do you like it?”. Real proof is a concrete action (email, payment promise, referral). If they say “I’ll buy if you have it,” that’s proof.

Key takeaways

• Customer development prevents building products nobody wants by focusing on the problem first.
• Zero→one does not guarantee one→n; commercial viability requires separate discovery.
• Inside the startup there are only opinions; data must come from customers.
• Early-stage interviews are qualitative, open-ended, and avoid pitching.
• Use qualifying questions, seek proof, and ask for referrals to validate demand.