Friday, January 7, 2011

Launching Tech Ventures: Part I, Course Overview

This is the first of four posts about Launching Technology Ventures (LTV), a new MBA elective course I'm developing at Harvard Business School to explore lean startup management practices. Part II will describe LTV's class sessions. Part III will list a set of tools and techniques that I think any MBA working in a tech venture should master. Part IV lists recommended readings for the course.

[Addendum, Apr. 10, 2011: if you are interested in lean startup concepts, you should also check out the LTV course blog. Students write blog posts instead of taking an exam. They've done terrific work.]

Most startups fail — usually due to lack of customer demand, not product development problems. These new ventures burn through their capital, wasting money on engineering and marketing before discovering they have built a product no one wants. Startups are more likely to succeed when they rapidly and iteratively test assumptions about a new venture’s business model based on customer feedback, then quickly refine promising concepts and ruthlessly cull the flops. New ventures that follow this approach are lean startups. “Lean” invokes the image of bootstrapping entrepreneurs, sustained by ramen noodles and a dream. Some lean startups fit that image. However, the term “lean” is derived from Toyota’s management philosophy. Toyota uses short production cycles to reduce inventory and eliminate waste. Lean startups similarly rely on short product development cycles to eliminate waste and gain rapid market feedback.

Lean startup practices are being pursued by firms in Silicon Valley and beyond. These practices have gained special traction in the information technology sector, where rapid “build/measure/learn” cycles are facilitated by the availability of open source engineering tools and Internet marketing channels. However, lean practices can also be applied by startups in other sectors and by large corporations launching new products.       

Launching Technology Ventures uses case studies to examine lean startup practices. LTV focuses on the integration of marketing and engineering functions and emphasizes implementation rather than strategy formulation issues. The course does not examine financing options or the composition of founding teams. LTV draws heavily on the ideas of Eric Ries, Steve Blank, Marty Cagan and other practitioners. Ries coined the term “lean startup” when he connected ideas from lean manufacturing and agile software development to Blank’s customer development process.  

Course Structure and Key Concepts

LTV is organized into two modules that explore execution challenges before and after a startup achieves product-market fit,  i.e., a match between its product solution and market needs.

The first module about challenges prior to achieving product-market fit covers the following core concepts:
  • The importance of well-structured experiments to confirm or disprove hypotheses about uncertain business model elements, thereby securing what Ries calls validated learning.
  • The benefits of what Ries calls a minimum viable product (MVP), i.e., the smallest set of product features and business initiatives needed to secure the next round of validated learning. MVPs can be counter-intuitive for managers and entrepreneurs, who often see fully-featured products as being better. However, building more features than necessary risks wasting time on functionality no one wants. It also compromises experimental designs, because it can be difficult to determine why a customer rejected a new version that incorporates many simultaneous changes.
  • The value of rapidly iterating the MVP based on customer feedback obtained through interviews, focus groups, usability tests, customer support interactions, etc.
  • Pivoting, i.e., changing a startup’s business model based on validated learning. When they pivot, startups retain some elements of their prior model to avoid waste. Lean startups may make many small pivots or a few big ones.
  • The need for metrics to gauge whether business model hypotheses have been validated, e.g., viral coefficients, customer retention rates.
  • The benefits of bootstrapping and avoiding big investments in marketing and infrastructure until business model hypotheses are validated.
In the first module, in addition to covering these core concepts, we’ll consider several issues of practical concern to marketing and product managers in early stage companies, for example:

  • When should a startup outsource engineering work?
  • How do firms design products for virality?
  • In a B2B context, how can an unknown startup with an unproven product identify potential early adopters and craft an effective sales pitch for them?
  • When does a “do-it-yourself” approach to public relations make sense, and how long should startups wait before they aggressively invest in PR?

The second module addresses challenges when scaling a business after achieving product-market fit. Topics include:
  • Approaches to customer conversion funnel analysis and optimization, e.g., usability labs, A/B testing.
  • Metrics for scaling startups, e.g., Net Promoter Score, Lifetime Value of a Customer.
  • Challenges in scaling a direct sales force; processes for prioritizing sales leads.
  • Tradeoffs for startups relying on business development partnerships with large companies.
  • Why, when and how startups should introduce formal product management processes, e.g., project prioritization and tracking systems, product roadmaps.


Boundaries

In addition to the issues described above, LTV will consider whether lean startup principles apply in three special contexts:
  • Platform-Based Businesses. "Do not scale until you validate your business model" is a core principle for lean startups. But does this make sense for platform-based businesses that harness strong network effects, such as Facebook, YouTube, and Twitter? All these firms scaled aggressively before they had proven business models; they subsequently relied on ecosystem partners to experiment with ways to monetize their big platforms.
  • Science-Based Businesses. For a class of capital- and/or science-intensive ventures—e.g., biotech, clean tech—the rapid iteration that is central for lean startups may not be a practical option. For such products, development and/or deployment inherently takes a long time due to fundamental uncertainty about engineering approaches or delays in deploying production capacity.
  • Large Corporations. Big companies have deep pockets and ample resources; that is their principal advantage over nimble startups. So, why should big companies constrain themselves unnaturally by running lean?
I'm eager to get feedback on the concepts covered in my course. What seems off target or missing?