1 of 88

LEAN STARTUPS�Introduction & Key Concepts

McDonald Institute Professional Development Series

Edward Thomas

2 of 88

Acknowledgement

This presentation is based on a four-part seminar developed �in 2017 by Edward Thomas for �Queen’s University’s Office of Partnerships and Innovation �and its Regional Innovation Ecosystem unit

3 of 88

$

FIND EARLY ADOPTERS

OFFER HELP

TRANSACTION

UTILITY

SCALING

If I can start this little talk with my conclusion … HERE’s THE PUNCHLINE: I’m going to build towards an iterative method for finding customers with problems and developing products that can sustainably and scalably meet their needs.�

engaging people trying to solve problems (and failing to do so satisfactorily, such as someone dissatisfied with their smile who is wearing a set of candy wax lips to mask themselves)
offering to help them accomplish the outcome they want (better smile)
Setting up a transaction to confirm they value this outcome (payment)
Provide the value with an iterative MVP loop (starting perhaps by helping them pull their face into a smile) … taking customer feedback (“your hands are really annoying”) and iterating the solution until the value realized and the method of delivering it are good enough to take to new customers (perhaps a regimen of smile-building face-muscle exercises).
Wash, Rinse and Repeat until I have validated a scalable and sustainable way to deliver value-creating solutions to customer problems.

4 of 88

Matthew Kupfer/CBC

5 of 88

6 of 88

Innovation vs. Invention

(and other definitions)

7 of 88

INVENTIONS ≠ INNOVATIONS, IF they are are NOT used create VALUE > 0.�
INNOVATION is not necessarily dependent on invention -- created value can be novel, even if the means of creation are not�
INNOVATION is core to entrepreneurship, INVENTION is not (but it can be an advantage).

8 of 88

Invention or Innovation?

Tying your shoes with L-sided knot
Graphene (1 atom-thick C allotrope)
Inside-the-shell egg scrambler
Human beat-box & bluegrass opera

9 of 88

Startups vs. Businesses

10 of 88

STARTUPS ARE NOT BUSINESSES�
LEAN PRINCIPLES, intro to VALUE-CHAINS�
LEAN MODELS applied to MVPs and METRICS�
STEP-BY-STEP (logical Lean Startup steps)

LEAN STARTUPS are entrepreneurship projects for identifying and proving a process model for innovating in a scalable and sustainable way. But let’s start with: WHY LEAN?

I’d like to tell you that startups are simple. But, NOPE. All business models are horribly complex machines, with design choices, builders, suppliers, buyers, financial models and risks -- each with a degree of uncertainty. The least complicated startup enterprise in the world with only 9 fundamental business assumptions would have at least 36 important correlations (9C2) to track at any given time. Realistically, an early stage startup would have in the order of 126 important correlations (9C4). The truth is that no one really has every facet of their business completely under control at any time. Startups start with ZERO facets of their business under control.
Startups begin as one giant pile of uncertainties. Our job as entrepreneurs is to move from total uncertainty toward acceptable uncertainty.
But which uncertainties should we tackle first? And how should we keep track of “progress”? And how do we communicate this progress with co-founders, customers and investors?
Lean thinking emerged in the 1950s from the problems of managing uncertainties in established businesses. It has recently gained a following among entrepreneurs dealing with the much deeper uncertainties facing startups. If you have ever given yourself a headache trying to figure out what is and isn’t “the real deal” in your startup idea, this might just be the medicine you are looking for….

11 of 88

12 of 88

CUMULATIVE COST OF FAILURE

Find Customers �Trying to Solve Problem

Offer Customers Help

Set Acceptable Price

Demonstrate Product Solves a Problem

Build the Product

UNCERTAINTY

5

1

2

3

4

13 of 88

CUMULATIVE COST OF FAILURE

Find Customers Trying to Solve a Problem

UNCERTAINTY

1

14 of 88

CUMULATIVE COST OF FAILURE

Find Customers Trying to Solve a Problem

Offer Customers Help

UNCERTAINTY

2

1

15 of 88

CUMULATIVE COST OF FAILURE

Find Customers Trying to Solve a Problem

Offer Customers Help

Set Acceptable Price

UNCERTAINTY

3

2

1

16 of 88

CUMULATIVE COST OF FAILURE

Find Customers Trying to Solve a Problem

Offer Customers Help

Set Acceptable Price

**Demonstrate Product’s Solution �to Problem

UNCERTAINTY

4

3

2

1

17 of 88

CUMULATIVE COST OF FAILURE

Find Customers Trying to Solve a Problem

Offer Customers Help

Set Acceptable Price

**Demonstrate Product’s Solution �to Problem

Build the Product

UNCERTAINTY

5

4

3

2

1

18 of 88

CUMULATIVE COST OF FAILURE

Find Customers Trying to Solve a Problem

Offer Customers Help

Set Acceptable Price

**Demonstrate Product’s Solution �to Problem

Build the Product

UNCERTAINTY

5

4

3

2

1

Sneaky!

19 of 88

Steve

Eric

Alex

Ash

Justin

The two previous slides pretty neatly illustrate the difference between “classic” and “Lean” startup thinking. Of course, they represent extreme variants of “non-ideal” vs. “ideal” startup progression, so we’ll spend the rest of this presentation unpacking my favourite approach for dealing with our somewhere-in-between realities.

�Just a brief moment here to acknowledge that my own Lean Startup thinking is massively based on the thinking of THESE five thought leaders:

1. Steve Blank started advising his business students at Stanford to treat startup businesses as a set of unproven assumptions rather than a complex clock-work model that was going to work with just a few tweaks here and there.

2. Steve later borrowed on Alex Osterwalder’s idea to collapse the ungainly business plan into a single-page business model as a “living document” that allowed entrepreneurs to re-write their assumptions on the fly in response to customers’ and suppliers’ behaviours. �3. Eric Ries (a serial entrepreneur and Steve’s former protege), seized on the Lean Management Principles used since WWII by firms like Toyota to make the process of testing, validating and refining business model assumptions as efficient and painless as possible. �4. Ash Maurya has been refining the Riesian methods and modifying the Osterwaldian tool into an efficient and robust “full stack” approach for startup ventures -- his “Lean” variant of the Business Model Canvas figures heavily in my current approach.

5. Justin Wilcox, co-founder of Customer Development Labs in San Franscisco, has, in my mind, approached the Blankian ideas and Osterwaldian tools with the first truly lightweight, immediately actionable step-by-step framework for the earliest-stage startups. I’ve been adapting his core ideas about product-market fit to Ash Maurya’s Lean Canvas since early 2017.

20 of 88

Key Partners

Key Activities

Key Resources

Value Propositions

Customer Relationships

Customer Segments

Channels

Cost Structure

Revenue Streams

Business Model Canvas

www.businessmodelgeneration.com

Let’s start with that Business Plan and it’s one-page replacement, The Business Model Canvas, which was invented by Alex Osterwalder, circa 2008. Basically, Alex realized that pretty much any business model you could name really addressed only nine important kinds of sub-models, or domains (each containing many specific variables within them, of course, but depending, in all likelihood on two or three truly fundamental variables). This realization flows naturally once your realize that all traditional business plans follow essentially the same format. The difference was Osterwalder’s insistence on limiting the verbiage to no more than a couple of strategically important statements in each of the model’s domains. What made simplification so enticing was acknowledging that most traditional business plans do not pass the test of time (they rarely survive first-contact with customers). A simple “canvas” allowed teams to have a nimble, flexible document that could be easily changed when necessary. The canvas actually reflected the reality that every business plan starts as set of unproven assumptions. THAT realization owes a huge debt to the work of Steve Blank…..

21 of 88

Key Partners

Key Activities

Key Resources

Value Propositions

Customer Relationships

Customer Segments

Channels

Cost Structure

Revenue Streams

Business Model Canvas

www.businessmodelgeneration.com

GUESS

FACT

GUESS

FACT

GUESS

22 of 88

Key Partners

Key Activities

Key Resources

Value Propositions

Customer Relationships

Customer Segments

Channels

Cost Structure

Revenue Streams

Business Model Canvas

www.businessmodelgeneration.com

FACT

….and THIS … is an established business. (flipping back and forth while making point). Steve Blank’s big insight was that full-on business plans only make sense for established businesses … That is, they are things you would LOGICALLY write only AFTER your business was ALREADY THRIVING and sustainable, and you would write them to help the team EXECUTE on a PROVEN BUSINESS MODEL. He also realized that CUSTOMERS are the most uncertain and vexing RISK to any business model, so it made sense to start with them first and work toward a PRODUCT from there.

Startups and businesses, Blank understood, AREN’T THE SAME THING AT ALL.

A STARTUPS’ mission is to find out a business model that is actually true. BUSINESSES’ mission is to operate a viable business model. He GOT IT that the entrepreneur’s mission is NOT to run a successful company, but to VALIDATE A SUSTAINABLE, SCALABLE BUSINESS MODEL which is a necessary (if not sufficient) condition for any successful growth enterprise. So if startups “business plans” are just a bunch of assumptions, he reasoned, then the process of becoming a business must be equivalent to the process of moving from an initial set of assumptions toward confirmed facts. The problem wasn’t HOW to represent a business model, but how to move from an UNCONFIRMED model to a CONFIRMED model. Steve urged his students to “get out of the building” and engage with customer needs and work from there. It was one of Steve’s serial-entrepreneur students, Eric Ries, who later seized on a long-lived management paradigm (LEAN) that was really good at delivering value under conditions of uncertainty … the Lean production and management system.

23 of 88

Startup or Business?

Twitch vs. YouTube
Uber vs. Zipcar
NASCAR vs. Formula E
New food truck vs. New French bistro

24 of 88

Lean

25 of 88

Edward

Taiichi

Micheal

James

Shigeo

LEAN THINKING is a sub-discipline of business management and has been in constant development since the late 1940s. We will not delve deeply into its theories or practices, but I do want to point out some of the foundational thinkers that have been important to my own understanding of Lean:

1. Edward Deming was hired to ramp up U.S. wartime production in WWII. He came to understood that effort and belief are not a guarantee of success in value-added goods manufacturing and he also figured out that the production of goods that deliver tremendous value is far more uncertain than business leaders wanted to admit, so he designed a humane management philosophy founded on elevating measurable, tangible evidence of performance over personal authority within organizations. He is considered by most to be an important driver of management-oriented statistical process control and continuous improvement methods in industry and by many others to be the spiritual father of Lean Thinking.

2. Taiichi Ohno was an early advocate of iterative learning in industrial practice and became the father of the Toyota Production System which is the industrial “lodestone” of Lean Management learning worldwide. Toyota’s contribution to Lean is foundational … there would be no Lean Management philosophy if Toyota had not adopted continuous improvement as its core operating principle.

3. Shigeo Shingo provided a sound theoretic framework underneath Toyota’s early Lean development and is responsible for knitting together Deming’s engineering-heavy system with Ohno’s management principles into a unified and coherent Lean framework.

4. James Womack is among several important Lean thinkers who have spent careers analyzing, and in many ways, translating and popularizing Toyota’s (and others) corporate knowledge base into a widely understood and tangible approach to management and customer value problems.

5. Michael E. Porter, is not a Lean Management thought leader, but his globally influential work on competition and value-chain/value-stream analysis has found its way into most Lean Thinkers’ approach to understanding products and customer value. His key contribution has been a kind of hybrid visual/quantitative language for understanding how entire supply chains are changed to drive productivity and value for entire ecosystems of producers and consumers. Many Lean advisors appropriate aspects of this language to look at micro-ecosystems of startups and their customers.

26 of 88

1. Customer-value drives purpose (pull)

4. Cycle through observation-testing-evidence-action

2. Eliminate waste (defects, effort, downtime, UX)

3. Transparency in the value-chain

There is a massive body of knowledge in Lean management. Lean management traces its roots to the post WWII era, and was most famously adopted by Toyota, and has spawned its own terminology. Its proponents use lots of exotic-sounding concepts like Kanban, Kaizen, Andon, Hoshin Kanri, SMED, etc., but take my word for it that they are all very neat expressions of customer-outcomes-driven common sense. Lean is not a linear management method, so much as it is an intellectual management framework. It is not a science, but an evolving body of practices (leaning heavily on scientific method) that help organizations act on evidence more than on assumptions of authoritative “leaders”. My purpose here is not to teach you Lean Thinking and Methods, but to bootstrap you to one actionable approach for validating startup businesses that is based on a self-consistent Lean framework. To grossly simplify Lean concepts as they apply to startups (or to translate Steve Blank’s mission statement into Lean-ese): “The entrepreneur’s ITERATIVE task is to efficiently gather the KNOWLEDGE required to ACT in the way that creates the most possible customer VALUE for the least possible cost.”

27 of 88

A bit about “value-chains”

+0.3

+0.9

OK, a quick aside about value-chains, because it’s really important to getting good at Lean Thinking. In a nutshell, I’m not a big fan of thinking about “products” per se. I prefer to think about VALUE-CHAIN disruptions.� � * A PRODUCT is anything that gets the customer’s problem solved (whether the customer is aware of that problem or not).� * A VALUE-CHAIN is everything that touches the customer’s experience of getting their problem solved. Disrupting this value chain is the job of INNOVATORS, including entrepreneurs, intrapraneurs, hackers, etc. Disruption may occur via a product, product-modification, a service, or sometimes even simple addition or elimination of one or more value-chain steps.

Learning to think about products (or services) as part of a value chain is an INCREDIBLY POWERFUL HABIT because it allows you to discover value creation & destruction in all kinds of unanticipated places. It often allows you to LEARN INEXPENSIVELY BY MODIFYING EXISTING PRODUCTS or SIMPLY DELIVERING THE VALUE YOU SEEK THROUGH MEANS OTHER THAN THE PRODUCT YOU “THINK” IS RIGHT.

THE TAKEAWAY IS - GET FAMILIAR WITH THE CONCEPT OF VALUE-CHAINS AND GO FURTHER AS AN ENTREPRENEUR!

28 of 88

PLAN: Where is the customer value?

INSIGHTS & ASSUMPTIONS

OK, back to the iterative Lean cyle of observation, experimentation, measuring and value-generation. Lean says that customer value is the highest calling of an enterprise. Toyota’s famous “just-in-time” delivery system is a manifestation of this philosophy, as it orients the entire company’s manufacturing base towards customer value (products they want when they want them) rather than the business's selfish goals (keeping things simple and cheap by continuously banging out widgets and storing them in inventory until someone can be pressured into buying them). Toyota saw that massive-scale inventory keeping of auto parts at a factory did nothing of value for customers. So it stopped doing so, with the added benefit that it stopped manufacturing cars with equipment options that no one was demanding to buy. The whole world-changing idea of “just-in-time” emerged from observing what was and what wasn’t delivering value to customers.

To respond to customer value, startups have to know (not guess, not think, but KNOW) what that value is that customers are seeking. It’s one thing to make an observation about customer value and have an assumption. It’s another thing entirely to KNOW something is true about customer value. That’s where the Lean cycle of PLAN-DO-CHECK-ACT is so powerful, because it gives us a recipe for moving from Assumptions to Knowledge.

29 of 88

+0.3

+0.9

Observed behaviour

To illustrate the concept of a value chain approach to observations & insights, suppose your startup wants to solve the problem of where customers put coffee when they’re driving. You see commuters carrying coffee cups in their hands all the time. Clearly coffee is important to them, but surely the risk of dropping the coffee in their lap and the inconvenience of only having one hand on the steering wheel must be causing problems, you think. The “product”, you would imagine, is an original-equipment cup-holder installed at the factory. But the value-chain is the whole sequence of things that add value to the customer’s experience of managing a cup of coffee while they drive. That chain doesn’t have to just be about your “product”. There are external parts of that value chain, like the care and attention of the detailing team, the interior layout chosen by the auto manufacturer, the glare from the windows or the smoothness of the ride that depends on well-inflated tires and properly aligned wheels and supple suspension. In addressing the “where to put the coffee” problem, you will change this value chain in some way and your task as an entrepreneur is to find out how to change it, how much it gets changed and what that is worth to the customer.

30 of 88

DO: Efficiently test that value proposition.

EXPERIMENTS

The easiest way to go from uncertainty to knowledge (or acceptabel levels of ignorance) is the scientific method. The thing about scientific method is that it requires effort (i.e., cost). So how do we get from ignorance to knowledge in the fastest way possible? We are going to conduct EXPERIMENTS to INVALIDATE our HYPOTHESES about our business model using the least-expensive EXPERIMENT we can devise. There are multiple levels of experiments in Lean Startups. The one we speak of most often (the minimum viable product) is actually a culmination of several experimental results around customer problems, customer motivation and the value of a solution to their problems. The important thing to really pay attention to is knowing that every unconfirmed assumption requires the experimental “DO” step as the next move.

31 of 88

+0.3

+0.9

?

“OLD”

“NEW”

Experiment �on Value Chain

32 of 88

CHECK: Did the test invalidate the value hypothesis?

DATA

33 of 88

+0.3

+0.9

+ 0.1

+ 1.0

“OLD”

“NEW”

Data

+0.0

34 of 88

ACT: Move in direction of value creation (incl. MVPs)

ACTIONS (Incl. MVPs)

35 of 88

+0.3

+ 0.1

1.0

Action (MVP)

So suppose we ACT by installing a cardboard-and-glue cupholder in the vehicle that is always there to hold coffee when needed. �We don’t like to think of the MVPs as a standalone “product” as much as VALUE CHAIN MODIFICATION.

To illustrate the concept of a value chain (and to make my point about the definition of an MVP), suppose the MVP “product” is a cup-holder prototype, but the value-chain is the whole sequence of things that add value to the customer’s experience of putting a coffee cup somewhere while they drive. That doesn’t have to just be about your startup either. There are external parts of that value chain, like the smoothness of the ride, or even the condition of the road itself. Your MVP changes this value chain in some way and your task as an entrepreneur is to find out how. If you have this kind of perspective, it may occur to you that adding a bit of foam rubber at the bottom of the cup holder is doing an even better job on the value chain since it dampens high-frequency road noise that the wheels and suspension aren’t catching….

36 of 88

PLAN-DO-CHECK-ACT: “Classic” Lean cycle

INSIGHTS & ASSUMPTIONS

EXPERIMENTS

DATA

ACTIONS (Incl. MVPs)

37 of 88

BUILD - MEASURE - LEARN: Ries’ Lean cycle variant

BUILD

MEASURE

LEARN

(EXPERIMENTS)

(KEY�METRICS)

(INSIGHTS & ASSUMPTIONS)

MVPs

DATA

IDEAS

38 of 88

Example: Just-in-Time Inventory

What observation(s) spurred this?
How could Toyota have tested idea?
How did they know it was working?
What did they do about it?

39 of 88

Models, MVPs & Metrics

40 of 88

Key Partners

Key Activities

Key Resources

Value Propositions

Customer Relationships

Customer Segments

Channels

Cost Structure

Revenue Streams

Business Model Canvas

www.businessmodelgeneration.com

So, I’ve been extolling this one-page business plan, developed by Alex Osterwalder around 2008. In many ways, the Business Model Canvas “ate the world” in startup thinking. It absolutely dominates most startup coaches’ toolkits for communicating key business model assumptions. I happen to love it, BUT I would be the first to admit that it doesn’t QUITE reflected a “fully” Lean approach to validation -- at least not at the earliest stages of identifying & delivering customer value.

� I mentioned before that the Business Model Canvas tends to abbreviate things you would expect to see in an old-fashioned business plan, which saves a lot of wasted effort. IN MANY WAYS, THE BUSINESS MODEL CANVAS IS A PERFECT TOOL FOR PITCHING A BUSINESS MODEL TO AN INVESTOR BUT IS AN UNWIELDY TOOL FOR VALIDATING “FIT” BETWEEN A PRODUCT AND A MARKET. For instance, it’s really hard to conceive of good experiments for testing the validity of complex sub-models like “Customer Relationships”, “Key Activities” or “Key Partners”. And where are the learning outcomes of these experiments stored in this canvas? Ash Maurya went to work on this a few years ago. Ash is a Lean guy, like Eric, but he found that the Osterwalder canvas, while appropriately flexible and lightweight for an ever-changing business proposition, was structured as though it was describing a fully formed business model. Ash wanted the benefits of the Osterwalder Canvas, but more inherently structured to align with Lean experiment-&-validate processes of Plan-Do-Check-Act/Build-Measure-Learn. So he proposed the Lean Canvas….

41 of 88

Problem

Solution

Key metrics

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams

Lean Model Canvas

The Lean Canvas is structured just like the Business Model Canvas, and it owes its fundamental utility model (its value proposition) to Alex Ostewalder’s earlier work. However, the Lean Canvas’ categories of assumptions are “stripped down” to the simplest elements of customer value, unlike the BMC which points to several multi-step value-chain assumptions (such as “Customer Relationships”, “Key Activities”, “Key Partners”). Notice also that the Lean Canvas insists on putting KEY METRICS (i.e., the things that are important to measure & monitor) front and centre, WHICH IS FUNDAMENTAL TO LEAN PHILOSOPHY (transparency in the value chain) and also, as Ries has noted, one of the main modes of startup failures (measurement of the wrong things). Let’s take a quick tour….

42 of 88

CHANNELS

EXPERIMENT

FACT

RESULTS

GUESS

MVP

=

OFFER + DELIV. VALUE

43 of 88

Problem:

What is the primary PAIN that customers want to alleviate?

What is the SCOPE of this problem for these customers?

Solution

Key metrics

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams

Lean Model Canvas

GUESS

FACT

CHANNELS

EXPERIMENT

44 of 88

Problem

Solution

Key metrics

Unique value proposition

Unfair advantage

Customer Segments:

Who is trying to solve this pain and HOW DO YOU IDENTIFY THEM?

Channels

Cost Structure

Revenue Streams

Lean Model Canvas

GUESS

CHANNELS

EXPERIMENT

FACT

45 of 88

Problem

Solution

Key metrics

Unique value proposition

Unfair advantage

Customer Segments

Channels:

HOW DO YOU REACH these people who are trying to solve their pain?

Cost Structure

Revenue Streams

Lean Model Canvas

GUESS

CHANNELS

EXPERIMENT

FACT

46 of 88

Problem

Solution

Key metrics

Unique value proposition:

What OUTCOMES are the customers really looking for in addressing their pain?

What balance between Time-Cost-Quality (pick two) represents the greatest value to customers?

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams

Lean Model Canvas

GUESS

CHANNELS

EXPERIMENT

FACT

47 of 88

Problem

Solution

Key metrics

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams:

What is the value proposition WORTH to customers?�In what WAYS are they willing to PAY?

Lean Model Canvas

GUESS

CHANNELS

EXPERIMENT

FACT

48 of 88

Problem

Solution:

Are you ABLE to deliver a value proposition that customers will pay you to obtain?

Key metrics

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams

Lean Model Canvas

MVP

=

OFFER + DELIV. VALUE

49 of 88

Problem

Solution:

Key metrics

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure:

Is your value-creating solution SUSTAINABLE?

Is it SCALABLE?

Revenue Streams

Lean Model Canvas

GUESS

CHANNELS

EXPERIMENT

FACT

50 of 88

Problem

Solution

Key metrics

Unfair Advantage:

Is it HARD to replicate the value-proposition through other means? Are the OBSTACLES for would-be competing value providers HARD to overcome?

Unique value Proposition

Customer Segments

Channels

Cost Structure

Revenue Streams

Lean Model Canvas

GUESS

CHANNELS

EXPERIMENT

FACT

51 of 88

Problem

Key metrics:

What things must be quantitatively MEASURED & MONITORED to prove your assumptions were (and continue to be) correct?

Solution:

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams

Lean Model Canvas

RESULTS

52 of 88

Cust. Interviews

Surveys

Landing Pages

Split Test

Multi-Factor Tests�All Offers...

All MVPs...

Simulated

“Glued to X”

Concierge

Wizard of Oz

“Cardboard”

Alpha 0.1

Release 1.0

Soln. Interview

Simple Offer

Demos

Teaser Packages

Smoke Test

Commit to Buy

EXPERIMENTS

OFFERS

MVPs

53 of 88

Example: McDonald’s Restaurants

Problem:

Solution:

Unique value proposition:

Unfair advantage:

Customer Segments:

Channels:

Cost Structure:

Revenue Streams:

Key Metrics:

Fast, affordable comfort food, available almost anywhere

Children in back seats of minivans

Parents with children to feed

Highway travellers

Budget constrained teens

Standardized menu & prep and centralized supply chain

Hurried, habit focused diners can’t get predictable meals on the go

Legally protected brand, trademarks and unique menu items

Cost of goods sold, Cost of acquiring customers, Customer Retention

Franchise Fees, Sales

Retail Margins, Advertising, Overheads

54 of 88

Problem

Solution

Key metrics

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams

Lean Model Canvas

PRODUCT ASSUMPTIONS

MARKET ASSUMPTIONS

55 of 88

Problem

Solution

Key metrics

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams

Lean Model Canvas

PRODUCT-VALUE MODEL

LEARN

CUSTOMER MODEL

BUSINESS PERFORMANCE MODEL

When I do sub-divide the model, I often think of the minimum level of validation that tends to correlate with specific levels of investment due diligence. I tend to see recurrences of the following pattern:

LEARNING: which is simple the things you MUST observe to validate any part of the model’s evolving validity. Startups that do not capture learning might get pitch prize money, but will inevitably fail to land any other external funding.
PRODUCT-VALUE: which is often enough for the very earliest stage of seed funding and typically sufficient for certain kinds of product-oriented pitching contests.
CUSTOMER MODEL: which, when added to the PRODUCT-VALUE is often enough for “light” angel investing. Getting to this level of validation (LEARNING + PROUCT VALUE + CUSTOMER MODEL) probably should be the bare minimum for attracting any level of “disciplined” capital.
BUSINESS PERFORMANCE MODEL: which when coupled with PRODUCT VALUE and CUSTOMER MODEL, is enough to attract “full” seed round angel capital.
FULLY VALIDATED: This is, in an ideal world, the bare minimum requirement for disciplined growth capital (Venture, Private Equity, Strategic Investment). The proven growth metrics required to gain each flavour of growth capital varies widely.

56 of 88

MVPs

“The smallest thing you can build that delivers customer value (and as a bonus captures some of that value back i.e. gets you paid).”

-- Ash Maurya

“The easiest available goods or services transaction that measurably tests a customer value-chain assumption in your business model.” �

-- Me

“The version of a new product which allows a team to collect the maximum amount of validated learning about customers with the least effort."

-- Eric Ries

We can’t discuss Lean Startups without talking about Minimum Viable Products (MVPs). We’ll breeze through this today and delve in deeply in Workshop Session #2. The MVP is the single most widely misunderstood concept in current startup conversation (Lean or otherwise). I think the FIRST issue is confusing the MVP with “a viable product”. A MINIMUM Viable Product CAN be a standalone product (i.e. a sustainable and scalable solution to a problem), but it DOES NOT HAVE TO BE A PRODUCT AT ALL in order to fulfil its function of A) validating the value proposition of solving a customer’s problem and B) delivering some portion of that value.. �

I don’t think the healthy debate of MVP definitions is done, but there is a consensus on these two points at least:� 1) We would all agree that the goal of an MVP is some learning outcome.

2) We would all agree that the point of MINIMUM is the ability to speed up our learning and ITERATIVELY OPTIMIZE CUSTOMER VALUE�

My current MVP definition is just a bit more specific about learning outcomes than Eric’s or Ash’s and is probably a bit more flexible about “payment”. Like Ash, I strongly agree that payment should be part of the MVP cycle, not because it necessarily validates the “price” of a solution (although it often does that as a bonus), but because payment is the ultimate data quality-control measure against customers lying to you to spare your feelings. Forcing the exchange of value is the most reliable way I can think of to ensure that your customers are leading you to value-creation and not towards false-confidence.

My current MVP definition owes an additional debt to Prof. Clayton Christiansen, who clarified the role of products [“People hire services and products to get specific jobs done in their lives.”]). The key here is that getting the job done (i.e., the RESULT) is what customers pay for -- that is to say they pay for alterations to the parts of the value chain that affect them. Whether or not the customer’s job is getting done by the product or service we may have in mind is often completely immaterial to validating the value proposition, the channel, the customer segment, problem and solution spaces and even the revenue stream (in some cases). The implication is that merely solving problems, either with or without a product, allows you to validate the least certain parts of a business model.

In about 80% of the business models I engage with every year, the “product” is mostly important for validating sustainable and scalable cost structures, unfair advantages and revenue streams (sometimes). That is to say, the product is important mainly to the sustainability and scalability of your business model, but not for validating the customer’s behaviours and value-drivers, which just happen to be the riskiest and most uncertain assumptions in these business models. I would argue that most startup can easily validate half of a business canvas with no product at all! Think of how many opportunities there are to rapidly test a business concept at nearly zero cost -- people who want to be serial entrepreneurs are crazy if they don’t gravitate to Lean.

57 of 88

BUILD

MEASURE

LEARN

(EXPERIMENTS)

(METRICS)

(INSIGHTS & ASSUMPTIONS)

MVPs

DATA

IDEAS

58 of 88

Product vs. MVP design

No value

Value

This is the classic “dual-track” illustration of “classic” product development, vs. iteration of a minimum-viable product. Say your customer wanted these jobs done (in order of priority):

travel from A to B
do it fast
do it with minimum physical effort
do it comfortably and safely

The high-risk, un-Lean approach would be to design a vehicle with all the features meeting the full specification. A Wheel to Travel, Four Wheels and a suspension to enable higher speed, an engine to do the “work” and a cabin with windows for comfort and safety. They would invest huge sums of money to build the product that they HOPE meets every declared customer need. Even more dangerous, until every single feature has been integrated, there will be zero potential of creating customer value (i.e. getting any of the customer’s jobs done).

The low-risk Lean approach would be to start with the Customer’s highest priority (travelling from A to B), and working on that problem with the easiest, lowest-cost solution available, like a skateboard. If the customer pays for it and finds it acceptable (an improvement over walking) then you might go after a more lavish design that not only gets from A to B, but goes faster, like a bike. Now the customer is happier and still getting value. More importantly, the risk taken to deliver customer value is much lower and the iterative value-creation cycle has already happened twice. By the time we get to the design of a scooter, or a sophisticated car, we will have validated every value proposition for the customer with a proper exchange of value at each step and we’ll have taken a much safer and certain path to really understanding our customer’s behavioural responses to the product at each stage of product capability. If the Lean customer had rejected the bicycle because getting there fast wasn’t REALLY as important as getting there at all, we’d be out $50 in prototyping expenses rather than $200k spent building an automobile prototype and we would have “failed” within a couple of days rather than six months or more. Better yet, we could have used insights from earlier MVPs to inform downstream MVPs (e.g., half of scooter drivers love to have the wind on their face, which told us to design automobile windows that roll up and down).

59 of 88

+0.3

+0.9

+ MVP

+1.0

Even standalone MVPs affect existing value chains

“OLD”

“NEW”

To illustrate the concept of a value chain (and to make my point about the definition of an MVP), suppose your startup wants to solve the problem of where to put coffee when driving. The MVP “product” is a prototype cup-holder installed in the car, but the value-chain is the whole sequence of things that add value to the customer’s experience of putting a coffee cup somewhere while they drive. That doesn’t have to just be about your startup either. There are external parts of that value chain, like the care and attention of the detailing team, the interior layout chosen by the auto manufacturer, the glare from the windows or the smoothness of the ride, or the mug in the cup holder, not to mention the quality of the coffee itself. Your MVP changes this value chain in some way and your task as an entrepreneur is to find out how.

60 of 88

Problem

Solution

Key metrics

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams

Lean Model Canvas

FACT

GUESS

61 of 88

Problem

Solution

Key metrics

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams

Lean Model Canvas

FACT

OFFER

DELIVERED VALUE

GUESS

62 of 88

Problem

Solution

Key metrics

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams

Lean Model Canvas

MVP

=

OFFER + DELIV. VALUE

63 of 88

Problem

Solution

Key metrics

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams

Lean Model Canvas

MVP

=

OFFER + DELIV. VALUE

RESULTS

64 of 88

Problem

Solution

Key metrics

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams

Lean Model Canvas

MVP

=

OFFER + DELIV. VALUE

RESULTS

CHANNELS

EXPERIMENT

CHANNELS

EXPERIMENT

GUESS

FACT

GUESS

65 of 88

Cust. Interviews

Surveys

Landing Pages

Split Test

Multi-Factor Tests�All Offers...

All MVPs...

Simulated

“Glued to X”

Concierge

Wizard of Oz

“Cardboard”

Alpha 0.1

Release 1.0

Soln. Interview

Simple Offer

Demos

Teaser Packages

Smoke Test

Commit to Buy

EXPERIMENTS

OFFERS

MVPs

66 of 88

Cust. Interviews

Surveys

Landing Pages

Split Test

Multi-Factor Tests�All Offers...

All MVPs...

Simulated

“Glued to X”

Concierge

Wizard of Oz

“Cardboard”

Alpha 0.1

Release 1.0

Soln. Interview

Simple Offer

Demos

Teaser Packages

Smoke Test

Commit to Buy

EXPERIMENTS

OFFERS

MVPs

67 of 88

Problem

Solution

Key metrics

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams

Lean Model Canvas

The family dog sheds hair on the sofa that is really hard to clean up

Enjoy a nice, pristine sofa with minimal effort

Over-busy professional dog lovers with little free time

Fee-for service

OFFER?

DELIVERED�VALUE?

68 of 88

Key Metrics

Measurable (NUMBERS)�
Actionable (DRIVES BEHAVIOUR)�
Exists in the model’s value chain �(PERTAINS TO VALUE)

We talked about measuring important things. This topic could be a whole week-long seminar. Measurement & Monitoring is the main thing that entrepreneurs get wrong. If the “metrics” you are monitoring do not drive your behaviours, they are not key metrics, but “vanity metrics”.

“We got 10,000 likes on Facebook,” is useless to most startups. “Two per cent of our Facebook likes clicked through our landing page and committed to look at our solution [i.e. a conversion rate],” IS actionable to most startups.

“Five of 10 customers love our solution” is probably not a key metric “Five of 10 customers who’ve used our product intend to buy from us again,” probably is.

“We made offers to 1,000 customers” is not a key metric. “We spent $10 for each customer we acquired,” most definitely is a key metric.

69 of 88

Churn

Conversion Rates

Measurable

Actionable

Valuable

referrals

web traffic

events hosted

client engagements

#trained

revenues

patents filed

associations joined

#startups

#customers

attendance

media coverage

“likes” & RTs

CAGR

# queries

#downloads

Winning Pitches

brand�recognition

CoCA

COGs

LTV

valuation

Viral Coeff.

awards

margins

Investor interest

Here’s the thing. Key Metrics will change depending on the “state of uncertainty” in your business model. Until you’ve generated revenues for four months, CAGR is a figment of your imagination, but later becomes important. Likewise, the # of customers for a product that does not depend on referral marketing has no real meaning to customer value.

At the very earliest stages of starting up, I DO like the exercise of imagining the metrics you think MIGHT be important to an ASSUMED business model. 1) Jot them all down 2) Classify each as measurable, actionable and/or relevant to customer value (use a value-chain/value-stream diagram to assess each metric). 3) Score each metric across each of the three categories (the highest scoring metrics are the current metrics you are aiming to collect).

Finally, just because a metric isn’t a Key Metric doesn’t mean that it has absolutely no utility whatsoever, it just means that you do not want to waste time trying to drive changes to that metric because that metric ultimately does not drive customer value. If a metric is not fundamental to your business model at this moment in time, keep it off the canvas. Aim to have no more key metrics than a reasonable person can keep in their head at all times (the upper limit would be 7, and for most people it’s more like 5 and for me it’s probably 3 or 4). Use a shareable dashboard/whiteboard/blackboard to constantly update those key metrics and share them with the entire team of co-founders at all times … this is the bedrock of your startup’s situational awareness.

70 of 88

A Toyota vanity metric or key metric?

Number of likes on facebook
Number of J. D. Power awards
Percentage of repeat customers
Number of quality complaints
Number of celebrity endorsements

71 of 88

Validating the Model

72 of 88

Best Practice: F.O.C.U.S.

Find early adopters ... customers who are trying to solve problem
Offer testing: do they want your help?
Currency testing: would they pay you to help them?
Utility testing: can you actually solve their problem?
Scaling: can you solve the problem over and over again without going broke?

Justin Wilcox: Founder of Customer Development Labs and creator of the F.O.C.U.S. method for product-market fit validation

So I’m going to go all fan-boy right here. Justin Wilcox’s F.O.C.U.S. steps are the easiest and most accessible framework for initial-stage startup behaviour that I have seen so far. As an added bonus, I happen to think that they fit beautifully into a Lean framework. He’s provided a lot of open overview material on YouTube to promote his in-depth methodologies. Regardless of what tools and methods you use inside of a Lean Startup framework, you should commit the FOCUS idea to your arsenal as it is supremely powerful in practice. Why do I love the FOCUS framework?

It is the fastest path to validating customer value (which is the riskiest assumption in your business model)
It does an excellent job of disentangling the idea of customer value from the “classic” product development cycle
It is really, really simple and captures the essence of Lean Startup behaviour without the “special language” Lean enthusiasts tend to promote (i.e., it makes it much easier to communicate your Lean approach with customers, suppliers, co-founders and investors).
Justin Wilcox has been an adamant proponent of Geoffrey Moore’s excellent customer behaviour theories, and has kept front-and-centre the one key behavioural trait that identifies which customers startups need to approach.

73 of 88

Problem

Solution

Key metrics

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams

Lean Model Canvas

1.

2.

3.

4.

5.

74 of 88

1.

2.

3.

4.

5.

75 of 88

$

FIND EARLY ADOPTERS

OFFER HELP

TRANSACTION

UTILITY

SCALING

And here we are at the end of a rapid 10,000 foot view of a Lean Startup approach to validating a new business model. We’ve gained a sense of what priorities inform Lean, what differentiates a startup from a business, the kinds of fundamental questions we’re trying to answer about our business model and one particularly helpful sequence for getting those questions answered. In simple, plain English we’ve arrived at a method for:

engaging people trying to solve problems (and failing to do so satisfactorily)
offering to help them accomplish the outcome they want
Setting up a transaction to confirm they value this outcome
Providing the value with an iterative MVP loop … taking the value-realized feedback of the customer and improving our solution until satisfied that it could repeatibly solve similar customer’s similar problem.
Getting the improved solution in the hands of more customers starting to test our assumptions about doing so sustainably and scalably while zeroing in on the key pieces of actionable information that will let us know if we are continuing on the right path.

76 of 88

$

FACE-EXERCISE VIDEO SERIES

Now there is an unspoken ART to all of this MVP iteration. The practiced entrepreneur will be audacious in using “light-weight” and often manually implemented MVPs to validate problem, customer segment, value proposition and revenue, but will ALSO have a scalable (often automated) PRODUCT ROADMAP in mind (something that could deliver the desired outcome, but can be delivered to many customers in a profitable way). Knowing when to “pull a customer’s face into a smile” and when to “tape it into a smiling position” and when to “hire a physiotherapist to design face-toning exercises” will all depend on the level of value-driven customer behaviour you have validated (i.e., how much risk you have taken off the table by testing [and proving] customer responses to your ability to solve their problems). If you have already collected $5,000 dollars to tape up 200 customer’s faces for $40, it might be a safe bet to pay $20,000 to contract an exercise program that will help 20,000 faces accomplish the same results at $4 apiece. On the other hand, you wouldn’t even contemplate spending the $20,000 if you had completely failed to collect 5 cents to temporarily pull even one customer’s face into an attractive smile.

Notice something here? The first 4-steps of business model validation, including the MVP, really don’t speak to “product development” so much as they address the alignment of customers’ desired outcomes to a value proposition on which you can deliver. “Product Development” is actually a mechanism for making the delivery of that value proposition a repeatable and scalable proposition (the 5th step). The ART is to use the MVP cycle to bridge between a valued outcome and a sustainable/scalable outcome (i.e., has a validated cost-structure, unfair advantage and key metrics [measurable means to monitor/ensure delivery of customers’ desired outcome remains valuable and sustainable]). THE MVP GETS YOU TO A POINT WHERE TWO-THIRDS OF YOUR ENTIRE BUSINESS MODEL IS VALIDATED [at a small scale, at least]. The next task is to move toward a “sustainable MVP”, which is of course, a fully realized “product”. If you’ve nailed this progression correctly, your “product” will already have a satisfied customer base that gets measurable value from its use, and whose cost-structure and business-model advantages ensure it can accomplish growth in a marketplace of early adopter customers.

77 of 88

78 of 88

79 of 88

80 of 88

81 of 88

82 of 88

$

Problems
Customer Segments

Value Prop
Channels

Revenue Streams

Solution

Key Metrics
Cost Structure
Unfair Advantage

We have nine flavours of business-model assumptions strongly-associated with each of these steps (and weakly associated with other steps, but that’s for an advanced workshop).

The caveat here is that there is an ART to identifying and validating a business model, even when doing so one step at a time. We have to anticipate the next step a little bit to ensure that the path we’re on won’t force us to a dead end unnecessarily . Yes, our entrepreneur could physically pull on customers’ cheeks to correct a flawed smile, but we know that it will not be sustainable to provide that particular mode of solution for many customers. But it is perfectly valid to offer to pull on customers’ cheeks if we have confidence that a scalable solution is possible (e.g. face muscle toning exercises).

Most entrepreneurs DO have some product in mind right out of the gate and will want to work toward building it as soon as possible. THAT’S WRONG by the way, but it is PERFECTLY HUMAN and therefore almost unavoidable. Lean tells us to at least slow down and get the customer engaged in that building process by first helping us to test the riskiest of our many assumptions about his/her VALUE-CHAIN. The art is not so much the ability to anticipate the eventual product, but to control our own ego enough to learn, from customers, what the product is really supposed to be and what job it is supposed to perform.

Last, Lean obeys the rule that “the customer is always right”. It can take a lot of change to get our product or solution aligned to customer value. We can end up in places radically different from what we anticipated. Sometimes, it’s just not possible for a startup to get alignment between the help you want to give and the problem the customer needs to solve. Lean allows us to identify these “failures” rapidly, and at minimal cost, allowing entrepreneurs to look for the next opportunity to solve a problem in the world with their good spirits and their wallets intact. That is to say, Lean entrepreneurs never really “fail” … they just use lowest-risk/highest-certainty methods to rapidly learn which opportunities are real and which opportunities are not.

83 of 88

5-minute research-to-startup model

Problem

What problem could your research address?

Solution

Unique value proposition

Unfair advantage

Customer Segments

Who is already trying to solve this problem?

Channels

Cost Structure

Revenue Streams

Key Metrics:

84 of 88

5-minute research-to-startup model

Problem

Solution

Unique value proposition

What outcome is this person hoping to accomplish?

Unfair advantage

Customer Segments

Channels

How can you find more people like this person?

Cost Structure

Revenue Streams

Key Metrics:

85 of 88

5-minute research-to-startup model

Problem

Solution

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams��What would they give to accomplish their aims?

Key Metrics:

86 of 88

5-minute research-to-startup model

Problem

Solution

What is the fastest way you could accomplish their aims for them?

Unique value proposition

Unfair advantage

Customer Segments

Channels

Cost Structure

Revenue Streams

Key Metrics:

87 of 88

5-minute research-to-startup model

Problem

Solution

Unique value proposition

Unfair advantage

How will you prevent someone else from stealing your business model?

Customer Segments

Channels

Cost Structure�Can you accomplish the desired outcome with indicated revenue and not go broke over time?

Revenue Streams

Key Metrics:

What will you measure to know you are right?

88 of 88

Key Takeaways

Address highest-risk assumptions FIRST

Customer-before-product
Product-market fit before Scaling

Have a “gameplan” for your startup’s progression from assumptions to knowledge