The LeanLaunch Pad at Stanford – Class 4: Customer Hypotheses

The Stanford Lean LaunchPad class was an experiment in a new model of teaching startup entrepreneurship. This post is part four. Part one is here, two is here and three is here. Syllabus is here.

Week 4 of the class.
Last week the teams were testing their hypotheses about their Value Proposition (their company’s product or service.) This week they were testing who the customer, user, payer for the product will be (and discovering if they have a multi-sided business model, one with both buyers and sellers.) Many of them had heard the phrase “product/market fit” before, but now they were living it. And for some of the teams the halcyon days of “we’re taking this class so we can just build our great product and get credit for it” had come to a screeching halt. The news from customers was not good.

Let the real learning begin.

The Nine Teams Present
This week, our first team up was PersonalLibraries (the team that had software to help researchers manage, share and reference the thousands of papers in their personal libraries.) Going into the first four weeks their business model hypotheses looked like this:Last week we told them team: 1) see if the market size was really large enough to support a business, and 2) to find that out they were going to have to 
talk to more customers
 outside of Stanford. So during the past week, the team got feedback from >60 researchers from 
cold calls, in-person interviews, and a web survey.  (We were impressed when we found that they did the in-person interviews by hiring for $39 to set up test scenarios, gave the users specific tasks to accomplish with their minimum viable product, videotaped the customer interactions and summarized customer likes and dislikes.) The good news was that customers said that their minimum viable product (easily organizing research papers) was correct. The bad news was that users would play with their product on-line for a while and leave and never return.  Politely it was described as “poor customer retention” but in reality it was because the product was really hard to use.

But it was their market size survey that had the team (and us) even more concerned; last weeks “hot” market of biomed researchers looked like it was only $30m market, and the total available reference manager market was another $80M.The question was, even if they got the product right, were there enough customers to make it a business?

If you can’t see the slides above, click here.

For next week, they decided to improve the product by adding more tutorials, do a 2nd Customer Survey and begin to create demand for their product with AdWords Value Prop Testing and Landing Page A/B Testing.

The feedback from the teaching team was that customer feedback seems to be saying that this product is a “nice to have” versus “got to have.” Is the lack of excitement the MVP? Users?  Is this a hobby or a business?

Agora Cloud Services
The Agora team started the week wondering whether they were 1) a true marketplace for cloud computing, where they provide both matching and exchange capabilities for real-time trading. Or were they 2) an information exchange, providing matching services for cloud computing buyers and sellers, providing matching services.

They began with a set of questions:

  • What are our new hypothesized value propositions?
  • Which segments have we identified and which do we want to narrow in on?
  • Which value added services do public clouds want to attract customers for?
  • Is there a certain segment of buyers that continually makes purchasing decisions (as opposed to only once at the very beginning of a company).
  • How can we attract buyers to our channel before they make purchasing decisions?
  • Longer-term work/planning: what other experiments should we be constructing
  • Sales process: buyer/ user/ influencer etc.?  Demand generation?

The Agora team decided to formalize the customer discovery process by coming up with a set of Customer Discovery principles and questions that were as good as any I’ve seen.

They had 16 interviews with target customers (Zynga, Yahoo, VMware, Walmart,  Zeconder, etc.) as well as channel partners and cloud industry technology consultants.

Agora was in a classic two-sided market (having both buyers and sellers. The Business Model Canvas is a great way to diagram it out. Each side of a market has it’s own Value Proposition, Customer Segment and Revenue Model.) They learned that one their core customer hypothesis about their buyers, “startups would want to buy computing capacity on a “spot market” was wrong. Startups were actually happy with Amazon Web Services. The Agora team was beginning to believe that perhaps their ideal buyers are the companies that have to handle variable and unpredictable workloads.

If you can’t see the slides above, click here.

The Agora team left the week thinking that it was time for a Pivot: find cloud buyers and sellers who need to better predict demand.  Perhaps in market segment: medium-large companies that do 3D modeling and life sciences simulations

The feedback from the teaching team “great Pivot” and very clear Lessons Learned presentation. Keep at it.

(For the teaching team one of the most important ways to track the teams progress was through the weekly blogs we made each team keep. This of this as their on-line diary. They hated doing it, but for us it added a window into their thinking process, allowed us to monitor how much work they were doing, and more importantly let us course correct when needed.

BTW, If I was on the board of a startup with a first time CEO I might even consider asking for this in the first year as they went through Customer Discovery. Yes it takes time, but I bet it’s less than time than you would spend having coffee with an advisor each week.)

D.C. Veritas, was the team building a low cost, residential wind turbine that average homeowners could afford. From a slow start of customer interaction they made major progress in getting out for the building. This week they refined their target market by building a map of potential customers in the U.S. by modeling wind speed, energy costs, homeownership density and green energy incentives. The result was a density map of target customers. They then did face-to-face interviews with 20 customers and got data from 36 more who fit their archetype.  They also interviewed two companies – Solar City and Awea in the adjacent market (residential photovoltaic’s.)

If you can’t see the slide presentation above, click here.

The teaching team offered that unlike solar panels which work anywhere, they’ve narrowed down the geographic areas where their wind turbine was economical. We observed that their total available market was getting smaller daily. After the next week figuring out demand creation costs, they ought to see if the homeowners were still a viable target market for residential wind turbines.

Autonomow, the robot lawn mower, came in with a major Pivot. Instead of a robotic lawn mower, they were now going to focus on robotic weeding and drop mowing as a customer segment. (Once you use the Business Model Canvas to keep score of Customer Discovery a Pivot is easy to define. A Pivot is when you substantively change one or more of the Business Model Canvas boxes.)

Talking to customers convinced the team that the need for robotic weeding was high, there was a larger potential market (organic crop production is doubling every 4 years and accelerating,) and they could make organic produce more affordable (labor cost reduction of 100 to 1) – and could possibly change the organic farming industry!  And as engineers they believed weed versus crop recognition, while hard, was doable.

During the week the team drove the 160 miles round-trip to the Salinas Valley and had on-site interviews with two organic farms. They walked the fields with the farmers, hand-picked weeds with the laborers and got down into the details of the costs of brining in an organic crop.

They also talked by phone to organic farmers in Nebraska and the Santa Cruz mountains.

They acquired quantitative data by going through the 2008 Agricultural Census. Most importantly their model of the customer began to evolve.

If you can’t see the slide above, click here.

Our feedback: could they really build a robot to recognize and weeds and if so how will they kill the weeds without killing the crops?  And are farmers willing to take a risk on untested and radical ideas like robots replacing hand weeding?

The Week 4 Lecture: Customer Relationships
Our lecture this week covered Customer Relationships (a fancy phrase for how will your company create end user demand by getting, keeping and growing customers.) We pointed out that get, keep and grow customers are different for physical versus virtual channels. Then different again for direct and indirect channels. We offered some examples of what a sales funnel looked like. And we described the difference between creating demand for products that solve a problem versus those that fulfill a need.

If you can’t see the slide above, click here.


The biggest lesson for the students this week was the entire reason for the class – no business plan survives first contact with customers – as customers don’t behave as per theory. As smart as you are, there’s no way to predict that from inside your classroom, dorm room or cubicle. Some of the teams were coming to grips with it. Others would find reality crashing down harder a bit later.

Next week, Class 5 – each team tests its demand creation hypotheses. The web-based teams needed to have their site up and running and be driving demand to the site with real Search Engine Optimization and Marketing tests.

Listen to the post here: Download the Podcast here

The LeanLaunch Pad at Stanford – Class 3: Value Proposition Hypotheses

The Stanford Lean LaunchPad class was an experiment in a new model of teaching startup entrepreneurship. This post is part three. Part one is here, two is here. Syllabus is here.

Week 3 of the class and our teams in our Stanford Lean LaunchPad class were hard at work using Customer Development to get out of the classroom and test the first key hypotheses of their business model: The Value Proposition. (Value Proposition is a ten-dollar phrase describing a company’s product or service. It’s the “what are you building and selling?”)

The Nine Teams Present
This week, our first team up was PersonalLibraries (the team that made software to help researchers manage, share and reference the thousands of papers in their personal libraries.) To test its Value Proposition, the team had face-to-face interviews with 10 current users and non-users from biomedical, neuroscience, psychology and legal fields.

What was cool was they recorded their interviews and posted them as YouTube videos. They did an online survey of 200 existing users (~5% response rate). In addition, they demoed to the paper management research group at the Stanford Intellectual Property Exchange project (a joint project between the Stanford Law School and Computer Science department to help computers understand copyright and create a marketplace for content). They met with their mentors, and refined their messaging pitch by attending a media training workshop one of our mentors held.

If you can’t see the slides above, click here.

In interviewing biomed researchers, they found one unmet need: the ability to cite materials used in experiments. This is necessary so experiments can be accurately reproduced. This was such a pain point, one scientist left a lecture he was attending to find the team and hand them an example of what the citations looked like.

The team left the week excited and wondering – is there an opportunity here to create new value in a citation tool? What if we could help scientists also bulk order supplies for experiments? Could we help manufacturers, as well, to better predict demand for their products, or perhaps to more effectively connect with purchasers?

The feedback from the teaching team was a reminder to see if the users they were talking to constitute a large enough market and had budgets to pay for the software.

Agora Cloud Services
The Agora team (offering a cloud computing “unit” that Agora will buy from multiple cloud vendors and create a marketplace for trading) had 7 face-to-face interviews with target customers, and spoke to a potential channel partner as well as two cloud industry technology consultants.

They learned that their hypothesis that large companies would want to lower IT costs by selling their excess computing capacity on a “spot market” didn’t work in the financial services market because of security concerns.  However sellers in the Telecom industries were interested if there was some type of revenue split from selling their own excess capacity.

On the buyers’ side, their hypothesis that there were buyers who were interested in reduced cloud compute infrastructure cost turned out not to be a high priority for most companies. Finally, their assumption that increased procurement flexibility for buying cloud compute cycles would be important turned out to be just a “nice to have,” not a real pain. Most companies were buying Amazon Web Services and were looking for value-added services that simplified their cloud activities.

If you can’t see the slides above, click here.

The Agora team left the week thinking that the questions going forward were:

  • žHow do we get past Amazon as the default cloud computing service provider?
  • How viable is the telecom market as a potential seller of computing cycles?
  • We need to further validate buyer & seller value propositions
  • How do we access the buyers and sellers? What sort of sales structure and salesforce does it require?
  • Who is the main buyer(s) and what are their motivations?
  • Is a buying guide/matching service a superior value proposition to marketplace?

The feedback from the teaching team was a reminder that at times you may have a product in search of a solution.

D.C. VeritasD.C. Veritas, the team that was going to build a low cost, residential wind turbine that average homeowners could afford, wanted to provide a renewable source of energy at affordable price.  They started to work out what features a minimum viable product their value proposition would have and began to cost out the first version. The Wind Turbine Minimum Viable Product would have a: Functioning turbine, Internet feedback system, energy monitoring system and have easy customer installation.

The initial Bill of Material (BOM) of the Wind Turbine Hardware Costs looked like: Inverter (1000W): $500 (plug and play), Generator (1000W): $50-100, Turbine: ~$200, Output Measurement: ~$25, Wiring: $20 = Total Material Cost: ~$800-$850

The team also went to the whiteboard and attempted a first pass at who the archetypical customer(s) might be.

To get customer feedback the team posted its first energy survey here and received 27 responses. In their first attempt at face-to-face customer interviews to test their value proposition and problem hypothesis (would people be interested in a residential wind turbine), they interviewed 13 people at the local Farmer’s Market.

If you can’t see the slide presentation above, click here.

The teaching team offered that out of 13 people they interviewed only 3 were potential customers. Therefore the amount of hard customer data they had collected was quite low and they were making decisions on a very sparse data set. We suggested (with a (2×4) that were really going to have to step up the customer interactions with a greater sense of urgency.The teaching team offered that out of 13 people they interviewed only 3 were potential customers. Therefore the amount of hard customer data they had collected was quite low and they were making decisions on a very sparse data set. We suggested (with a 2×4) that were really going to have to step up the customer interactions with a greater sense of urgency.

The last team up was Autonomow, the robot lawn mower. They were in the middle of trying to answer the question of  “what problem are they solving?” They were no longer sure whether they were an autonomous mowing company or an agricultural weeding company.

They spoke to 6 people with large mowing needs (golf course, Stanford grounds keeper, etc.) They traveled to the Salinas Valley and Bakersfield and interviewed 6 farmers about weeding crops. What they found is that weeding is a hugeproblem in organic farming. It was incredibly labor intensive and some fields had to be hand-weeded multiple times per year.

They left the week realizing they had a decision to make – were they a  “Mowing or Weeding” company?

If you can’t see the slide above, click here.

Our feedback: could they really build a robot to recognize and kill weeds in the field?

The Week 3 Lecture: Customers
Our lecture this week covered Customers – what/who are they?  We pointed out the difference between a user, influencer, recommender, decision maker, economic buyer and saboteur. We also described the differences between customers in Business-to-business sales versus business-to-consumer sales.  We talked about multi-sided markets and offered that not only are there multiple customers, but each customer segment has their own value proposition and revenue model.

If you can’t see the slide above, click here.

Getting Out of the Building
Five other teams presented after these four. All of them had figured out the game was outside the building, with some were coming up to speed faster than others. A few of the teams ideas still looked pretty shaky as businesses. But the teaching team held our opinions to ourselves, as we’ve learned that you can’t write off any idea too early. Usually the interesting Pivots happens later. The finish line was a ways off. Time would tell where they would all end up.


Next week – Class 4 each team tests their Customer Segment hypotheses (who are their customers/users/decision makers, etc.) and report the results of face-to-face customer discovery. That will be really interesting.
Listen to this post here: Download the Podcast here

The LeanLaunch Pad at Stanford – Class 2: Business Model Hypotheses

Our new Stanford Lean LaunchPad class was an experiment in a new model of teaching startup entrepreneurship. This post is part two. Part one is here. Syllabus here.

By now the nine teams in our Stanford Lean LaunchPad Class were formed, In the four days between team formation and this class session we tasked them to:

  • Write down their initial hypotheses for the 9 components of their company’s business model (who are the customers? what’s the product? what distribution channel? etc.)
  • Come up with ways to test each of the 9 business model canvas hypotheses
  • Decide what constitutes a pass/fail signal for the test. At what point would you say that your hypotheses wasn’t even close to correct?
  • Consider if their business worth pursuing? (Give us an estimate of market size)
  • Start their team’s blog/wiki/journal to record their progress during for the class

The Nine Teams Present
Each week every team presented a 10 minute summary of what they had done and what they learned that week. As each team presented, the teaching team would ask questions and give suggestions (at times pointed ones) for things the students missed or might want to consider next week. (These presentations counted for 30% of their grade. We graded them on a scale of 1-5, posted our grades and comments to a shared Google doc, and had our Teaching Assistant aggregate the grades and feedback to pass on to the teams.)

Our first team up was Autonomow. Their business was a robot lawn mower. Off to a running start, they not only wrote down their initial business model hypotheses but they immediately got out of the building and began interviewing prospective customers to test their three most critical assumptions in any business:
Value PropositionCustomer Segment and Channel. Their hypotheses when they first left the campus were:

  • Value Proposition:  Labor costs in mowing and weeding applications are significant, and autonomous implementation would solve the problem.
  • Customer Segment: Owners/administrators of large green spaces (golf courses, universities, etc.) would buy an autonomous mower.  Organic farmers would buy if the Return On Investment (ROI) is less than 1 year.
  • Channel: Mowing and agricultural equipment dealers

All teams kept a blog – almost like a diary – to record everything they did. Reading the Autonomow blog for the first week, you could already see their first hypotheses starting to shift: “For mowing applications, we talked to the Stanford Ground Maintenance, Stanford Golf Course supervisor for grass maintenance, a Toro distributor, and an early adopter of an autonomous lawn mower. For weeding applications, we spoke with both small and large farms. In order from smallest (40 acres) to largest (8000+ acres):  Paloutzian Farms, Rainbow Orchards, Rincon Farms, REFCO Farms, White Farms, and Bolthouse Farms.”

“We got some very interesting feedback, and overall interest in both systems,” reported the team. “Both hypotheses (mowing and weeding) passed, but with some reservations (especially from those whose jobs they would replace!)  We also got good feedback from Toro with respect to another hypothesis – selling through distributor vs. selling direct to the consumer.”

The Autonomow team summarized their findings in their first 10 minute, weekly Lesson Learned presentation to the class.

Our feedback: be careful they didn’t make this a robotics science project and instead make sure they spent more time outside the building.

If you can’t see the slide deck above, click here.

Autonomow team members:
Jorge Heraud (MS Management, 2011) Business Unit Director, Agriculture, Trimble Navigation, Director of Engineering, Trimble Navigation, MS&E (Stanford), MSEE (Stanford), BSEE (PUCP, Peru)
Lee Redden (MSME Robotics, Jun 2011) Research in haptic devices, autonomous systems and surgical robots, BSME (U Nebraska at Lincoln), Family Farms in Nebraska
Joe Bingold (MBA, Jun 2011) Head of Product Development for Naval Nuclear Propulsion Plant Control Systems, US Navy, MSME (Naval PGS), BSEE (MIT), P.E. in Control Systems
Fred Ford (MSME, Mar 2011) Senior Eng for Mechanical Systems on Military Satellites, BS Aerospace Eng (U of Michigan)
Uwe Vogt (MBA, Jun 2011) Technical Director & Co-Owner, Sideo Germany (Sub. Vogt Holding), PhD Mechanical Engineering  (FAU, Germany), MS Engineering (ETH Zurich, Switzerland

The mentors who volunteered to help this team were Sven Strohbad, Ravi Belani and George Zachary.

Personal Libraries
Our next team up was Personal Libraries which proposed to help researchers manage, share and reference the thousands of papers in their personal libraries. “We increase a researcher’s productivity with a personal reference management system that eliminates tedious tasks associated with discovering, organizing and citing their industry readings,” wrote the team. What was unique about this team was that Xu Cui, a Stanford postdoc in Neuroscience, had built the product to use for his own research. By the time he joined the class, the product was being used in over a hundred research organizations including Stanford, Harvard, Pfizer, the National Institute of Health and Peking University. The problem is that the product was free for end users and few Research institutions purchased site licenses. The goal was to figure out whether this product could become a company.

The Personal Libraries core hypotheses were:

  • We solve enough pain for researchers to drive purchase
  • Dollar size of deals is sufficient to be profitable with direct sales strategy
  • The market is large enough for a scalable business

Our feedback was that “free” and “researchers in universities” was often the null set for a profitable business.

If you can’t see the slide above, click here.

Personal Libraries Team Members
Abhishek Bhattacharyya (MSEE, Jun 2011) creator of WT-Ecommerce, an open source engine, Ex-NEC engineer
Xu Cui (Ph.D, Jun 2007 Baylor) Stanford Researcher Neuroscience, postdoc, BS biology from Peking University
Mike Dorsey (MBA/MSE, Jun 2011) B.S. in computer science, environmental engineering and middle east studies from Stanford, Austin College and the American University in Cairo
Becky Nixon (MSE, Jun 2011) BA mathematics and psychology Tulane University Ex-Director, Scion Group,
Ian Tien (MBA, Jun 2011) MS in Computer Science from Cornell, Microsoft Office Engineering Manager for SharePoint, and former product manager for SkyDrive

The mentors who volunteered to help this team were Konstantin Guericke and Bryan Stolle.

The Week 2 Lecture: Value Proposition
Our working thesis was not one we shared with the class – we proposed to teach entrepreneurship the way you would teach artists – deep theory coupled with immersive hands-on experience.

Our lecture this week covered Value Proposition – what problem will the customer pay you to solve?  What is the product and service you were offering the customer to solve that problem.

If you can’t see the slide above, click here.

Feeling Good
Seven other teams presented after the first two (we’ll highlight a few more of them in the next posts.) About half way through the teaching team started looking at each other all with the same expression – we may be on to something here.


Next week – Class 3 each team tests their value proposition hypotheses (their product/service)  and reports the results of face-to-face customer discovery. Stay tuned
Listen to this post here: Download this Podcast here

A New Way to Teach Entrepreneurship – The Lean LaunchPad at Stanford: Class 1

For the past three months, we’ve run an experiment in teaching entrepreneurship.

In January, we introduced a new graduate course at Stanford called the Lean LaunchPad. It was designed to bring together many of the new approaches to building a successful startup – customer development, agile development, business model generation and pivots.

We thought it would be interesting to share the week-by-week progress of how the class actually turned out. This post is part one.

A New Way to Teach Entrepreneurship
As the students filed into the classroom, my entrepreneurial reality distortion field began to weaken. What if I was wrong? Could we even could find 40 Stanford graduate students interested in being guinea pigs for this new class? Would anyone even show up?  Even if they did, what if the assumption – that we had developed a better approach to teaching entrepreneurship – was simply mistaken?

We were positing that 20 years of teaching “how to write a business plan” might be obsolete. Startups, are not about executing a plan where the product, customers, channel are known. Startups are in fact only temporary organizations, organized to search–not execute–for a scalable and repeatable business model.

We were going to toss teaching the business plan aside and try to teach engineering students a completely new approach to start companies – one which combines customer development, agile development, business models and pivots. (The slides below and the syllabus here describe the details of the class.)

Get Out of the Building and test the Business Model
While we were going to teach theory and frameworks, these students were going to get a hands-on experience in how to start a new company. Over the quarter, teams of students would put the theory to work, using these tools to get out of the building and talk to customer/partners, etc. to get hard-earned information. (The purpose of getting out of the building is not to verify a financial model but to hypothesize and verify the entire business model. It’s a subtle shift but a big idea with tremendous changes in the end result.)

Team Autonomow: Weeding Robot Prototype on a Farm

We were going to teach entrepreneurship like you teach artists – combining theory – with intensive hands-on practice. And we were assuming that this approach would work for any type of startup – hardware, medical devices, etc. – not just web-based startups.

If we were right, we’d see the results in their final presentations – after 8 weeks of class the information/learning density in the those presentations should be really high. In fact they would be dramatically different than any other teaching method.

But we could be wrong.

While I had managed to persuade two great VC’s to teach the class with me (Jon Feiber and Ann Miura-ko), what if I was wasting their time? And worse, what if I was going to squander the time of my students?

I put on my best game face and watched the seats fill up in the classroom.

A few weeks before the Stanford class began, the teaching team went through their Rolodexes and invited entrepreneurs and VCs to volunteer as coaches/mentors for the class’s teams. (Privately I feared we might have more mentors than students.) An hour before this first class, we gathered these 30 impressive mentors to brief them and answer questions they might have after reading the mentor guide which outlined the course goals and mentor responsibilities.

As the official start time of the first class drew near, I began to wonder if we had the wrong classroom. The room had filled up with close to a 100 students who wanted to get in. When I realized they were all for our class, I could start to relax. OK, somehow we got them interested. Lets see if we can keep them. And better, lets see if we can teach them something new.

The First Class
The Lean LaunchPad class was scheduled to meet for three hours once a week. Given Stanford’s 10 week quarters, we planned for eight weeks of lecture and the last two weeks for team final presentations. Our time in class would be relatively straightforward. Every week, each team would give a 10-minute presentation summarizing the “lessons learned” from getting out of the building. When all the teams were finished the teaching team lectured on one of the 9 parts of the business model diagram. The first class was an introduction to the concepts of business model design and customer development.

The most interesting part of the class would happen outside the classroom when each team spent 50-80 hours a week testing their business model hypotheses by talking to customers and partners and (in the case of web-based businesses) building their product.

Selection, Mixer and Speed Dating
After the first class, our  teaching team met over pizza and read each of the 100 or so student applications. Two-thirds of the interested students were from the engineering school; the other third were from the business school. And the engineers were not just computer science majors, but in electrical, mechanical, aerospace, environmental, civil and chemical engineering. Some came to the class with an idea for a startup burning brightly in their heads.  Some of those applied as teams. Others came as individuals, most with no specific idea at all.

We wanted to make sure that every student who took the class had at a minimum declared a passion and commitment to startups. (We’ll see later that saying it isn’t the same as doing it.) We tried to weed out those that were unsure why they were there as well as those trying to build yet another fad of the week web site. We made clear that this class wasn’t an incubator. Our goal was to provide students with a methodology and set of tools that would last a lifetime – not to fund their first round. That night we posted the list of the students who were accepted into the class.

The next day, the teaching team held a mandatory “speed-dating” event with the newly formed teams. Each team gave each professor a three-minute elevator pitch for their idea, and we let them know if it was good enough for the class. A few we thought were non-starters were sold by teams passionate enough to convince us to let them go forward with their ideas. (The irony is that one of the key tenets of this class is that startups end up as profitable companies only after they learn, discover, iterate and Pivot past their initial idea.) I enjoyed hearing the religious zeal of some of these early pitches.

The Teams
By the beginning of second session the students had become nine teams with an amazing array of business ideas. Here is a brief summary of each.

Agora isan affordable “one-stop shop” for cloud computing needs. Intended for cloud infrastructure service providers, enterprises with spare capacity in their private clouds, startups, companies doing image and video processing, and others. Agora’s selling points are its ability to reduce users’ IT infrastructure cost and enhance revenue for service providers.

Autonomow is an autonomous large-scale mowing intended to be a money-saving tool for use on athletic fields, golf courses, municipal parks, and along highways and waterways. The product would leverage GPS and laser-based technologies and could be used on existing mower or farm equipment or built into new units.

BlinkTraffic will empower mobile users in developing markets (Jakarta, Sao Paolo, Delhi, etc.) to make informed travel decisions by providing them with real-time traffic conditions. By aggregating user-generated speed and location data, Blink will provide instantaneously generated traffic-enabled maps, optimal routing, estimated time-to-arrival and predictive itinerary services to personal and corporate users.

D.C. Veritas is making a low cost, residential wind turbine. The goal is to sell a renewable source of energy at an affordable price for backyard installation. The key assumptions are: offering not just a product, but a complete service (installation, rebates, and financing when necessary,) reduce the manufacturing cost of current wind turbines, provide home owners with a cool and sustainable symbol (achieving “Prius” status.)

JointBuy is an online platform that allows buyers to purchase products or services at a cheaper price by giving sellers opportunities to sell them in bulk. Unlike Groupon which offers one product deal per day chosen based on the customer’s location. JointBuy allows buyers to start a new deal on any available product and share the idea with others through existing social networking sites. It also allows sellers to place bids according to the size of the deal.  

MammOptics is developing an instrument that can be used for noninvasive breast cancer screening. It uses optical spectroscopy to analyze the physiological content of cells and report back abnormalities. It will be an improvement over mammography by detecting abnormal cells in an early stage, is radiation-free, and is 2-5 times less expensive than mammographs. We will sell the product directly to hospitals and private doctors.

Personal Libraries is a personal reference management system streamlinig the processes for discovering, organizing and citing researchers’ industry readings. The idea came from seeing the difficulty biomed researchers have had in citing the materials used in experiments. The Personal Libraries business model is built on the belief that researchers are overloaded with wasted energy and inefficiency and would welcome a product that eliminates the tedious tasks associated with their work.

PowerBlocks makes a line of modular lighting. Imagine a floor lamp split into a few components (the base, a mid-section, the top light piece). What would you do if wanted to make that lamp taller or shorter? Or change the top light from a torch-style to an LED-lamp? Or add a power plug in the middle? Or a USB port? Or a speaker? “PowerBlocks” modular lighting is “floor-lamp meets Legos” but much more high-end. Customers can choose components to create the exact product that fit their needs. is an ad-supported, web-based comment platform for daily news content. Real-time conversations and dynamic curation of news stories empowers people to expand their social networks and personal expertise about topics important to them. This addresses three problems vexing the news industry: inadequate online community engagement, poor topical search capacity on news sites, and scarcity of targeted online advertising niches.

While I was happy with how the class began, the million dollar question was still on the table – is teaching entrepreneurship with business model design and customer development better than having the students write business plans? Would we have to wait 8 more weeks until their final presentation to tell? Would we signs of success early?  Or was the business model/customer development framework just smoke, mirror and B.S.?

The Adventure Begins
We’re going to follow the adventures of a few of the teams week by week as they progressed through the class, (and we’ll share the teams weekly “lessons learned,” as well as our class lecture slides.

The goal for the teams for next week were:

  • Write down their hypotheses for each of the 9 parts of the business model.
  • Come up with ways to test:
    • what are each of the 9 business model hypotheses?
    • is their business worth pursuing (market size)
  • Come up with what constitutes a pass/fail signal for the test (e.g. at what point would you say that your hypotheses wasn’t even close to correct)?
  • Start their blog/wiki/journal for the class

Next Post: The Business Model and Customer Discovery Hypotheses – Class 2
Listen to this post here: Download the Podcast here

The Lean LaunchPad – Teaching Entrepreneurship as a Management Science

I’ve introduced a new class at Stanford to teach engineers, scientists and other professionals how startups really get built.

They are going to get out of the building, build a company and get orders in ten weeks.

Jon Feiber of Mohr Davidow Ventures and Ann Miura-Ko of Floodgate are co-teaching the class with me (and Alexander Osterwalder is a guest lecturer.) We have two great teaching assistants, plus we’ve rounded up a team of 25 mentors (VC’s and entrepreneurs) to help coach the teams.

Why Teach This Class?
Business schools teach aspiring executives a variety of courses around the execution of known business models, (accounting, organizational behavior, managerial skills, marketing, operations, etc.)

In contrast, startups search for a business model. (Or more accurately, startups are a temporary organization designed to search for a scalable and repeatable businessmodel.)  There are few courses which teach aspiring entrepreneurs the skills (business models, customer and agile development, design thinking, etc.) to optimize this search.

Many entrepreneurship courses focus on teaching students “how to write a business plan.” Others emphasize how to build a product. If you’ve read any of my previous posts, you know I believe that:  1) a product is just a part of a startup, but understanding customers, channel, pricing, etc. are what make it a business,
2) business plans are fine for large companies where there is an existing market, existing product and existing customers. In a startup none of these are known.

Therefore we developed a class to teach students how to think about all the parts of building a business, not just the product.

What’s Different About the Class?
This Stanford class will introduce management tools for entrepreneurs.  We’ll build the class around the business model / customer development / agile development solution stack.

Students will start by mapping their assumptions (their business model) and then each week test these hypotheses with customers and partners outside in the field (customer development) and use an iterative and incremental development methodology (agile development) to build the product.

The goal is to get students out of the building to test each of the 9 parts of their business model, understand which of their assumptions were wrong, and figure out what they need to do fix it. Their objective is to get users, orders, customers, etc. (and if a web-based product, a minimum feature set,) all delivered in 10 weeks.  Our objective is to get them using the tools that help startups to test their hypotheses and make adjustments when they learn that their original assumptions about their business are wrong.  We want them to experience faulty assumptions not as a crisis, but as a learning event called a pivot —an opportunity to change the business model.

How’s the Class Organized?
During the first week of class, students form teams (optimally 4 people in a team but we’re flexible.) Their company can focus in any area– software, hardware, medical device or a service of any kind.

The class meets ten times, once a week for three hours. In those three hours we’ll do two things.  First, we’’ll lecture on one of the 9 building blocks of a business model (see diagram below, taken from Business Model Generation).  Secondly, each student team will present “lessons learned” from their team’s experience getting out of the building learning, testing, iterating and/or pivoting their business model.

They’ll share with the class answers to these questions:

  1. What did you initially think?
  2. So what did you do?
  3. Then what did you learn?
  4. What are you going to do next?

At the course’s end, each team will present their entire business model and highlight what they learned, their most important pivots and conclusions.

We’re going to be teaching it for the first time in January.  Below is the class syllabus.


Class 1  is here.  Follow along!

Engineering 245
This course provides real world, hands-on learning on what it’s like to actually start a high-tech company. This class is not about how to write a business plan. It’s not an exercise on how smart you are in a classroom, or how well you use the research library. The end result is not a PowerPoint slide deck for a VC presentation. Instead you will be getting your hands dirty talking to customers, partners, competitors, as you encounter the chaos and uncertainty of how a startup actually works.  You’ll work in teams learning how to turn a great idea into a great company. You’ll learn how to use a business model to brainstorm each part of a company and customer development to get out of the classroom to see whether anyone other than you would want/use your product. Finally, you’ll see how agile development can help you rapidly iterate your product to build something customers will use and buy.  Each week will be new adventure as you test each part of your business model and then share the hard earned knowledge with the rest of the class. Working with your team you will encounter issues on how to build and work with a team and we will help you understand how to build and manage the startup team.

Besides the instructors and TA’s, each team will be assigned two mentors (an experienced entrepreneur and/or VC) to provide assistance and support.

Suggested Projects: While your first instinct may be a web-based startup we suggest that you consider a subject in which you are a domain expert, such as your graduate research. In all cases, you should choose something for which you have passion, enthusiasm, and hopefully some expertise.  Teams that select a web-based product will have to build the site for the class.


Pre-reading For 1st Class:  Read pages 1-51 of Osterwalder’s Business Model Generation.

Class 1    Jan 4th Intro/Business Model/Customer Development
Class Lecture/Out of the Building Assignment:
What’s a business model? What are the 9 parts of a business model?  What are hypotheses? What is the Minimum Feature Set? What experiments are needed to run to test business model hypotheses?   What is market size? How to determine whether a business model is worth doing?

Deliverable: Set up teams by Thursday, Jan 6 (a mixer will be hosted on Wednesday to help finalize teams).  Submit your project for approval to the teaching team.


Deliverable for January 11th:

  • Write down hypotheses for each of the 9 parts of the business model.
  • Come up with ways to test:
    • is a business worth pursuing (market size)
    • each of the hypotheses
    • Come up with what constitutes a pass/fail signal for the test (e.g. at what point would you say that your hypotheses wasn’t even close to correct)?
    • Start your blog/wiki/journal

Jan 6th 5-7pm Speed Dating  (Meet in Thornton 110)

Get quick feedback on your initial team business concept from the teaching team.

Class 2            Jan 11th Testing Value Proposition
Class Lecture/Out of the Building Assignment:
What is your product or service? How does it differ from an idea? Why will people want it? Who’s the competition and how does your customer view these competitive offerings? Where’s the market? What’s the minimum feature set?  What’s the Market Type?  What was your inspiration or impetus?  What assumptions drove you to this?  What unique insight do you have into the market dynamics or into a technological shift that makes this a fresh opportunity?


  • Get out of the building and talk to 10-15 customers face-to-face
  • Follow-up with Survey Monkey (or similar service) to get more data


  • Business Model Generation, pp. 161-168 and 226-231
  • Four Steps to the Epiphany, pp. 30-42, 65-72 and 219-223
  • The Blue Ocean Strategy pages 3-22

Deliverable for Jan 18th:

  • Find a name for your team.
  • What were your value proposition hypotheses?
  • What did you discover from customers?
  • Submit interview notes, present results in class.
  • Update your blog/wiki/journal

Class 3            Jan 18th Testing Customers/users
Class Lecture/Out of the Building Assignment:
Who’s the customer? User? Payer?  How are they different? How can you reach them? How is a business customer different from a consumer?


  • Get out of the building and talk to 10-15 customers face-to-face
  • Follow-up with Survey Monkey (or similar service) to get more data


Deliverable for Jan 25th:

  • What were your hypotheses about who your users and customers were? Did you learn anything different?
  • Submit interview notes, present results in class. Did anything change about Value Proposition?
  • What are your hypotheses around customer acquisition costs?  Can you articulate the direct benefits (economic or other) that are apparent?
  • If your customer is part of a company, who is the decision maker, how large is the budget they have authority over, what are they spending it on today, and how are they individually evaluated within that organization, and how will this buying decision be made?
  • What resonates with customers?
  • For web startups, start coding the product. Setup your Google or Amazon cloud infrastructure.
  • Update your blog/wiki/journal

Class 4            Jan 25th Testing Demand Creation
Class Lecture/Out of the Building Assignment:
How do you create end user demand? How does it differ on the web versus other channels?   Evangelism vs. existing need or category? General Marketing, Sales Funnel, etc


  • If you’re building a web site:
    • Small portion of your site should be operational on the web
    • Small portion of your site should be operational on the web
  • Actually engage in “search engine marketing” (SEM)spend $20 as a team to test customer acquisition cost
    • Ask your users to take action, such as signing up for a newsletter
    • use Google Analytics to measure the success of your campaign
    • change messaging on site during the week to get costs lower, team that gets lowest delta costs wins.
    • If you’re assuming virality of your product, you will need to show viral propagation of your product and the improvement of your viral coefficient over several experiments.
  • If non-web,
    • build demand creation budget and forecast.
    • Get real costs from suppliers.


Watch: Mark Pincus, “Quick and Frequent Product Testing and Assessment”,

Deliverable for Feb 1st :

  • Submit interview notes, present results in class.
  • Did anything change about Value Proposition or Customers/Users or Channel?
  • Present and explain your marketing campaign. What worked best and why?
  • Update your blog/wiki/journal

Class 5            Feb 1st Testing Channel
Class Lecture/Out of the Building Assignment:
What’s a channel?  Direct channels, indirect channels, OEM. Multi-sided markets.  B-to-B versus B-to-C channels and sales (business to business versus business to consumer)

Action: If you’re building a web site, get the site up and running, including minimal feature.

  • For non-web products, get out of the building talk to 10-15 channel partners.

Read: Four Steps to the Epiphany, pp. 50-51, 91-94, 226-227, 256, 267

Deliverable for Feb 8th:

  • For web teams:
    • Get a working web site and analytics up and running. Track where your visitors are coming from (marketing campaign, search engine, etc) and how their behavior differs. What were your hypotheses about your web site results?
    • Submit web data or customer interview notes, present results in class.
    • Did anything change about Value Proposition or Customers/Users?
    • What is your assumed customer lifetime value?  Are there any proxy companies that would suggest that this is a reasonable number?
    • For non-web teams:
      • Interview 10-15 people in your channel (salesmen, OEM’s, etc.).
      • Did anything change about Value Proposition or Customers/Users?
      • What is your customer lifetime value?  Channel incentives – does your product or proposition extend or replace existing revenue for the channel?
      • What is the “cost” of your channel, and it’s efficiency vs. your selling price.
      • Everyone: Update your blog/wiki/journal.
        • What kind of initial feedback did you receive from your users?
        • What are the entry barriers?

Class 6            Feb 8th Testing Revenue Model
Class Lecture/Out of the Building Assignment:
What’s a revenue model? What types of revenue streams are there? How does it differ on the web versus other channels?

Action: What’s your revenue model?

  • How will you package your product into various offerings if you have more than one?
  • How will you price the offerings?
  • What are the key financials metrics for your business model?
  • Test pricing in front of 100 customers on the web, 10-15 customers non-web.
  • What are the risks involved?
  • What are your competitors doing?

Read: John Mullins & Randy Komisar, Getting to Plan B (2009) pages 133-156

Deliverable for Feb 15th :

  • Assemble an income statement for the your business model. Lifetime value calculation for customers.
  • Submit interview notes, present results in class.
  • Did anything change about Value Proposition or Customers/Users, Channel, Demand Creation, Revenue Model?
  • Update your blog/wiki/journal

Class 7            Feb 15th Testing Partners
Class Lecture/Out of the Building Assignment:
Who are partners?  Strategic alliances, competition, joint ventures, buyer supplier, licensees.

Action: What partners will you need?

  • Why do you need them and what are risks?
  • Why will they partner with you?
  • What’s the cost of the partnership?
  • Talk to actual partners.
  • What are the benefits for an exclusive partnership?

Deliverable for Feb 22nd

  • Assemble an income statement for the your business model.
  • Submit interview notes, present results in class.
  • Did anything change about Value Proposition or Customers/Users, Channel, Demand Creation?
  • What are the incentives and impediments for the partners?
  • Update your blog/wiki/journal

Class 8            Feb 22nd Testing Key Resources & Cost Structure
Class Lecture/Out of the Building Assignment:
What resources do you need to build this business?  How many people? What kind? Any hardware or software you need to buy? Any IP you need to license?  How much money do you need to raise?  When?  Why? Importance of cash flows? When do you get paid vs. when do you pay others?

Action: What’s your expense model?

  • What are the key financials metrics for costs in your business model?
  • Costs vs. ramp vs. product iteration?
  • Access to resources. What is the best place for your business?
  • Where is your cash flow break-even point?

Deliverable for March 1st

  • Assemble a resources assumptions spreadsheet.  Include people, hardware, software, prototypes, financing, etc.
  • When will you need these resources?
  • Roll up all the costs from partners, resources and activities in a spreadsheet by time.
  • Submit interview notes, present results in class.
  • Did anything change about Value Proposition or Customers/Users, Channel, Demand Creation/Partners?
  • Update your blog/wiki/journal

Guest: Alexander Osterwalder

For March 1st or 8th

Class 9            March 1st Team Presentations of Lessons Learned (1st half of the class)

Deliverable: Each team will present a 30 minute “Lessons Learned” presentation about their business.

Class 10            March 8th Team Presentations of Lessons Learned (2nd half of the class)

Deliverable: Each team will present a 30 minute “Lessons Learned” presentation about their business.

March 11th 1-4pm Demo Day at VC Firm (Location TBD)

Show off your product to the public and real VC’s.  Set up a booth, put up posters, run demos, etc.  Food and refreshments provided.

Class 1  is here.  Follow along!

Mentor List (as of Dec 3rd 2010)

Class 1  is here.  Follow along!
Listen to the post here: Download the Podcast here