The LeanLaunch Pad at Stanford – Class 5: Customer Relationship Hypotheses

The Stanford Lean LaunchPad class was an experiment in a new model of teaching startup entrepreneurship. This post is part five. Parts one through four are here, Syllabus is here. 

Week 5 of the class.
Last week the teams were testing their hypotheses about their Customers (who are the users, payers, buyers, etc.)  This week they were testing one of the most confusing sections of a company’s business model – Customer Relationships – the activities used to “Get, Keep and Grow” customers in a physical or virtual (web or mobile) channel. (Internet investor Dave McClure coined the acronym “AARRR,” to remember the parts of Customer Relationships on the web.)

Many of the students had heard phrases that fall under Customer Relationships before; “customer acquisition, SEO/SEM, public relations, Social Network, Advertising, Loyalty programs, cross-sell and up-sell” etc., but now they were actually trying to implement it. (If their team was a web or mobile app they actually had to buy Google or Facebook ads and create demand.)

For some of the teams their expectation was if they built the product customers will come. Filing into the classroom I could tell that for some reality had just come crashing down on them. Seeing the lack of customer interest for the first time is always depressing. (The goal of the class was to get them to understand that in a startup, that was the norm not the exception. And to teach them a methodology of what to do about it.) It was making some of the teams question other parts of their business model (did they have the right customer, did they have the right product features to meet customer needs, etc.)

The Nine Teams Present
The first team to present was D.C. Veritas, the team building a low cost, residential wind turbine. During the week they interviewed 7 more companies and consultants, developed case studies for 20 different cities in 5 states, and finalized the bill of materials for the wind turbine. But the big project for the week was testing and analyzing Customer Acquisition Costs.  The team put together their sales funnel and started testing demand.

The results were disappointing. The most optimistic estimates showed that the residential wind turbine market was less than $20m in year 5 and the costs to acquire the customers made this a money-losing business.

After regrouping the team decided that a major pivot was in order. Perhaps residential customers were the wrong target?  Maybe the wind turbine they were building was better suited to a different customer segment?  They had gotten feedback from consultants and industry experts that cities and utilities might be a more receptive audience. What if they redesigned the wind turbine to be embedded into street and highway light poles?  Then they could serve cities, lighting companies and utilities. Using the business model canvas, the changes to their business were obvious.

(BTW, our definition of a Pivot: it’s when you significantly modify one or more of the business model building blocks.)

Three more weeks to go.  Can the D.C. Veritas team discover whether there’s a real opportunity for their wind turbine in cities? The teaching team observed that the next few weeks are going to be interesting. Time to dig in and find out.

Our next team up was Autonomow, the robot lawn mower farm weeder. Last week they had pivoted from customers who needed large areas mowed, to organic farmers who needed lower costs for weeding. In this weeks foray into farm country they spoke to five farm implement dealers and interviewed yet another farmer. However, their primary focus was thinking through how they would “get” their initial customers. In talking to farmers and farm equipment dealers they learned the farm-specific places to create demand; trade shows like the World Ag Expo and magazines such as Vegetable Grower, Ag Source, Farm Equipment and Tractor House. The team then put together a specific budget for initial demand creation.

The teaching team suggested that was the research to date was great, but until they built a robot that could actual tell the difference between a weed and a plant, this would just be a paper exercise. They were engineers, certainly they could do better than that? The Autonomow team started thinking how they could prove that their paper business model was real.

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

Last week we asked the PersonalLibraries team: are there enough customers to make this a business?  So during the week they ran more hands-on user testing, A/B tests, landing page conversion tests, and bought Google Adwords.

The results were not impressive. The feedback they were getting was that the product was a “nice to have” but not a “hair-on-fire” product.

Our feedback was, that their data seemed to say that their current users don’t want to spend money and will incur infinite support and infinite cost. Our suggestion was, “run away from the academic researcher market as fast as possible.” We offered that the team  might want to expand their user research to think about new features and verticals (document management, law firms, lab managers with discretionary budget, etc.)

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

Agora Cloud Services
The Agora team ended last 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.  This week they answered the question by “punting.”  They decided they were going to start as a information services, move to brokering, then prediction and finally evolve into a true market.  They interviewed another 8 buyers/sellers/industry experts.

Their results on whether they could acquire with Google Adwords was a bit sobering. Their first effort didn’t get much traffic: 6 clicks out of ~2000 impressions.  Worse yet, each of these clicks cost about a $1.00.  Reason? They had been bidding on keywords that are too generic (e.g. cloud, ec2, Amazon Web Services, etc.)

Their ads of “Cloud Demand Prediction” hadn’t been catching the eyes of people searching for these keywords.  So they picked more specific keywords such as, (cloud comparison, best cloud providers, etc).  And they created ads with specific headlines, such as “Too many cloud providers?”, “Reduce your cloud spend”, etc).  They also increased their daily campaign budget to $20.00.  What they found was that the keywords that did have traffic volume are extremely expensive. Depending on the keyword, the first page bids were between $5.00 to $25.00 per click! Ouch.

The team concluded that AdWords may not be the best channel to create demand.

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

The Week 5 Lecture: Channel
Channels are how a company delivers its value proposition (i.e. its product or service) to its customers. There are two major channels – virtual (web/mobile) and physical channels – and the difference is dramatic. In one, physical goods move from a loading dock to a customer or a retail outlet. In another the product is offered and sold online. (If the product is itself bits, it may not only be sold online but is often also delivered or used on-line.)

Our lecture talked out how to choose the right sales channel, how the channel makes money, how they’re motivated, and the economics of a sales channel.

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


The lesson for the students this week was failure. What we wanted to teach them wasn’t how to fail fast – any idiot can do that. We wanted to teach them how to recognize failure, learn from it, and pivot.  It’s not about failing fast – it’s about learning faster. That’s the lesson at the heart of the search for a repeatable and scalable business model.

Now deep into the class most of the teams are starting to rethink their initial assumptions. Which teams will continue to Pivot?  Will any completely abandon their current business and pick a new one?

Stay tuned.

Next week – Class 6 – Distribution Channel Hypothesis

Listen to the post here: Download the Podcast here

3 Responses

  1. This is great stuff, but it leaves the impression that it only implies to beginning businesses. In fact, the majority of my time is spent doing approximately the same thing to much more developed businesses and even mature businesses. It is very easy for an organization to wonder down a wrong path and then wonder why they are in trouble. It may be hard to believe but this is true for the vast majority of all businesses. This methodology needs to be applied, and double checked almost continuously by all businesses, in my experience.

  2. Re: Peaya

    Conversion is low because nobody knows what a “reference manager” is. Sounds like a term coined by the founders, not one that is used among your target customers. Would recommend changing the value proposition on landing page so it’s clear and understandable to your target customers. Even something simple like a “productivity tool for busy students”. Focus on the desire outcome/job your targets are searching for. People aren’t searching for “reference managers”, we’re looking for ways to reduce our time managing all of our papers. Good luck

  3. […] Proposition Hypotheses The LeanLaunch Pad at Stanford – Class 4: Customer Hypotheses The LeanLaunch Pad at Stanford – Class 5: Customer Relationship Hypotheses The LeanLaunch Pad at Stanford – Class 6: Channel Hypotheses The LeanLaunch Pad at Stanford […]

Leave a Reply

%d bloggers like this: