Iterative entrepreneurship

In "Getting to Plan B" John Mullins (London Business School) and Randy Komisar (Kleiner Perkins Caufield & Byers) argue that entrepreneurial success is often an iterative process. Since the initial (Plan A) business model rarely leads to success, it is the capacity to evolve toward a Plan B (and C) that constitutes the test of entrepreneurial savvy.

	Title:	Getting to Plan B
	Author:	John Mullins and Randy Komisar
	Pages:	224pages
	Publisher:	Harvard Business Press
	Price:	$29.95

Getting to Plan B: Breaking through to a Better Business Model argues that the business models of successful ventures are typically not the initial business models. Accordingly, it is the ability to test and modify the initial model, the Plan A, that puts entrepreneurs on the path to success. While success typically occurs only after several iterations of the business model, this does not mean that Plan A is useless; it means that Plan A should be treated as a modifiable starting-point, a base camp rather than the summit. The authors offer two interesting quotes from American generals on the subject of  initial planning and plan modification: “Plans are useless but planning is indispensable “ (Eisenhower); “No plan ever survives its first encounter with the enemy” (McArthur).

The authors define a business model as “the pattern of activity – cash flowing into and out of your business for various purposes and the timing thereof - that dictates whether or not you run out of cash and whether or not you deliver attractive returns to your investors.”  In their vision, moving toward a better business model builds upon four blocks: analogs, antilogs, leaps of faith and dashboards. Five financial elements determine the cash activity of the venture: revenue model, gross margin model, operating model, working capital model and investment model. Iterative entrepreneurs will use the building blocks to build, test and modify the five financial models.

The first two chapters of the book are devoted to the four building blocks of the process while the ensuing five are devoted to the five financial models. In each chapter, the authors offer two or three mini-case studies of firms as evidence for their entrepreneurial arguments.

Analogs and antilogs
A key element in the process is to learn from others, from successful and not-so-successful ventures. Man is mimetic and so are successful business models. The authors call analogs those predecessors that are worth mimicking in some way. They call antilogs predecessors with whom one begs to differ. So analogs are sources of imitation while antilogs are sources of differentiation.

The authors use Apple’s iPod  as an example of how analogs and antilogs can contribute to a business model. Among the iPod’s analogs were Sony’s Walkman and Napster. The Walkman proved that people wanted to be able to listen to music on portable devices. Napster proved that downloading music was at least as attractive as going to the store and that users were more interested in single songs than entire albums. Three antilogs helped refine the model: MusicNet, Pressplay, Diamond Media Systems. MusicNet was a subscription service but users could only download to a computer. Pressplay allowed burning to a CD but its library was restricted. Diamond Media Systems had marketed an MP3 player but the interface made song search and organization difficult. The analogs and antilogs set Apple on the path to a portable user-friendly device that enabled users to download single songs from an extensive library.

Leaps of faith and dashboards
The examples of analogs and antilogs will not suffice to answer all questions regarding the success of the business model. Sometimes there will be no evidence to prove or refute some of the crucial choices within the model – the authors call those choices leaps of faith. These leaps of faith must be translated into hypotheses which can be tested via measurement, quickly and inexpensively. Recording these leaps, the hypotheses and the test results is what the authors call dashboarding.

Whereas established businesses often reel under data overload, entrepreneurial ventures often suffer from underload.  Dashboarding helps overcome underload. Small tests can provide the data that allows entrepreneurs to transform leaps of faith into justified steps or to move onto a tweaked model. Think of dashboarding then as the entrepreneur’s version of experimental science, where a premium is placed on the simplicity of the experiment.

GlobalGiving is an example of a venture that successfully applied dashboarding principles.  This venture was the brainchild of Dennis Whittle and Mari Kuraishi. While at the World Bank, these two created the Development Workplace, a two-day forum for bottom-up poverty alleviation projects. They subsequently created a venture, Development Space, in partnership with a foundation (Calver Foundation) based on the leap of faith that the Internet could be used to generate donations for global development projects. Three hypotheses needed to be tested: that attractive projects would come to the marketplace, that a sufficient number of donors want to fund these projects, and that the marketplace would be financially sustainable.

By creating a dashboard around these leaps of faith, Whittle and Kuraishi realized, before it was too late, that a selection of projects would contribute to trustworthiness, that donors could be aggregated through corporate partnerships and that wealthy individuals and foundations would provide the bulk of donations. Dashboarding thus led them to their Plan B in which they worked with an existing organization, Ashoka, to find development projects. They collaborated with companies like HP for access to their employee giving programs. Lastly, in their Plan B, Whittle and Kuraishi created their own foundation to target high net worth individuals and foundations.

Revenue model
The authors then move on to the cash flow determinants and beginning with the revenue model, emphasize the point that the Plan A revenue model does not always meet expectations and that capacity for evolution is crucial. Elements of the model that might need to be adjusted are: who will buy; what they will buy; in what quantities and with what timing will they buy; at what price and with what payment methods will they buy.

The authors look at three ventures that saw evolution in their revenue model. One of them is Silverglide Surgical Technologies which had developed a non-stick technology applicable to medical instruments.  This venture’s initial revenue model had plastic surgeons purchasing a surgical probe that offered advance non-stick technology. But plastic surgeons were not that accustomed to probes. Silverglide had to look elsewhere and in their Plan B came to focus on neurosurgeons and applied their non-stick technology to the forceps that such surgeons made heavy use of.  After five years, the venture’s cash flow turned positive and shortly before its tenth birthday it was sold to a large diversified medical technology company.

Gross margin model
To be successful, ventures must also generate positive gross margins (defined as revenue – cost of goods sold).  As one of their cases, the authors present eBay, the exception that proves the rule that Plan A does not succeed.  eBay owes its financial success in large part to extraordinarily high margins. In 1995, Pierre Omidyar who was really only trying to create a perfect market launched AuctionWeb. Because AuctionWeb was not involved in any physical handling of goods, gross margin was in the 80% range. This contrasted with its older competitor, and antilog, Onsale which took possession of the auctioned goods. Onsale needed warehouses and personnel to store and process the goods.  eBay (AuctionWeb was renamed in 1997) had chosen the right gross margin model and Onsale the wrong one. When eBay went public in 1997 its gross margin was 88%, while Amazon’s, for example, was 22%.

Operating model
Operating costs are all the day-to-day costs that are incurred in addition to cost of goods sold. The operating model will specify  what level of operating costs will be incurred, what operating costs can be reduced without creating strategic damage, what costs might be increased to provide strategic benefit. Among the cases centered on operating expenses, the authors present ZoomSystems.  ZoomSystems is an automated, low operating-expense form of retail.  In its Plan B version it amounts to a smart vending machine for small but high-value products ($10-200). Located in high-traffic locations such as airports and hotels, the machine offers a touchscreen pad, a virtual salesperson, enabling the customer to find out about the products.  Given that the machines typically required only around thirty square feet of space, ZoomSystem’s rent expenses were much lower than a traditional retailer. Furthermore, it chose to reduce the risk associated with a fixed rent by offering the property owner a share of the revenues.  By 2008, ZoomSystem’s approximately 800 machines were delivering $3000-10000 of sales per square foot per month as compared to the American mall average of $33.

Working capital model
The working capital model is key to a new business’s sustainability but also forms the foundation of some of today’s most interesting business models.  Whereas the previous models dealt with income statement items, the working capital model focuses on the management of balance sheet items. With working capital defined as current assets (e;g; inventory, receviables, cash) minus current liabilities (e;. payables), successful working  models will feature low or negative non-cash working capital. One of the benefits of negative non-cash working capital is that entrepreneurs will need less access to equity capital, be it from venture capitalists or bankers.

The authors present Costco as one example of a venture that has benefited from an astute working capital model. Founded in 1983, this American warehouse club retailer, a competitor of Walmart’s Sam’s Club, relies on its low inventory and receivables and higher payables to generate negative working capital.  To increase inventory turns Costco carried only 4,000 items, much easier to manage than a supermarket’s offering of 40,000 items. By not allowing credit card payments, it kept receivables to a bare minimum of 3 days of sales. In contrast, its payables amounted to 27 days. This working capital model enabled Costco to achieve negative noncash working capital, equivalent to seven days of sales. Ten years later it had maintained control of the assets side while gaining four more days of payables. The authors summarize the benefits of this model by pointing out that for each store these eleven days amounted to about $3.6 million or enough to build a supplementary store.

Investment model
The investment model is about determining how much cash will be needed to start the business and to make the transitions to improved plans. The investment model must navigate between Scylla and Charbydis. Raising too much money leads to sloppiness and less than optimal focus on testing of hypotheses. Raising too little money means that the transition to Plan B or C might prove financially impossible.

Another investment model issue is source of financing: family and friends, or banks or venture capitalists. Since banks will usually only lend against collateral or cash-flow, they are rarely an option. The trouble with venture capitalists is that they are not in the business for the long run and that most exits take the form of the sale of the start-up to another, larger company rather than the form of an initial public offering. Google and Amazon are the exception rather than the rule in this regard.

One of the two examples presented by the authors is Skype. The founders, Niklas Zennstrom and Janus Friis, were devoted to applications of peer-to-peer (P2P) technology. They used the proceeds from the sale of their Plan A (KaZaA, a file-exchanging platform) to develop P2P telephony. They spent a year developing the service which they offered for free initially, with the goal of building a huge base of peers. Their Plan B amounted to signing up a critical mass of users and worrying about revenue later. Once the users had signed on, they moved to Plan C, generating revenue by allowing cut-rate calls to traditional phones and through services for a fee. They raised $19 million in venture capital financing and by August 2005, 2 million of the 40 million users started availing themselves of the new for-fee services. In September 2005 eBay purchased Skype for $2.6 billion. Zennstrom and Friis had gone the route of proving major leaps of faith (Plan B) before raising money to implement Plan C – iterative entrepreneurship at work.