Friday, March 11, 2011

Startup Stacks

Startup Stacks

Building startups quickly and effectively depends on using prebuilt operational components, just like rapid web development uses technology stacks

Download this whitepaper at the Windcastle Venture Consulting site.

Capital requirements for startups (at least in the technology industry) have been falling steadily. One major reason for this is the existence of stacks: technology stacks, strategy stacks and marketing stacks. Stacks are a set of preexisting components that can be snapped together, quickly reconfigured and dynamically repurposed depending on the particular challenges of the moment.

An example of a technology stack is the LAMP stack (Linux, Apache, MySQL, PHP or Perl) commonly used for web applications. These components provide a prebuilt and commonly used operating system, HTTP server, database and scripting language, respectively. Together, these and similar stacks for Windows and OSX operating systems form the basis for many web applications.

Know where the money is going

At the bottom level, the strategy stack is so noncontroversial and universal that it’s almost invisible. This first layer is the double entry accounting method in use since the Middle Ages. It’s applicable to any type or size of business—although modern GAAP has a number of features better suited to atoms than bits—and is widely understood. The figures and metrics generated by this system provide an invaluable resource to the founders and managers. I’m not going to talk about good accounting here, except to say that if it’s done well, it gets a lot of trouble out of the way, freeing up efforts elsewhere in the startup—so get a copy of Quickbooks or Peachtree and someone who knows how to use them, and learn how to read a financial statement yourself so you can tell what’s going on with the business.

Wasted time is worse than wasted money

Lean Thinking provides a strong, middle layer. Adapted from Toyota’s production management model, lean thinking focuses on reducing wasted time, work or money. Again, this is a suite of tools, methodologies and mindsets adopted from traditional manufacturing, but eminently suited for the technology startup.

The need for lean stems from the scarcity of capital and the high discount rate for any capital that is available. In the context of DCF/NPV valuations: For a normal, stable business with full access to open debt and equity markets, a dollar delivered in a year might be worth spending $.90-.95 today. For a privately financed startup in an industry with a 30-50% annual failure rate, next year’s dollar is worth perhaps $.50 today. The non-zero chance of failure results in a situation where it is more important in many ways to save time than to save money.

At the same time, Lean Thinking and Lean Management are extremely congenial to startup culture, which has made a strength out of mobility, flexibility and bootstrapping. Every startup in history has started without a history, meaning that it’s got a lot less baggage to lose. It can pivot faster, and run faster in a new direction than a company with more resources (which are also more firmly dedicated toward minding the existing business). YCombinator’s funding model, Eric Ries’ Minimum Viable Product concept, prototyping, virtual company models and agile programming methodologies are all built on top of a lean mindset.

Discover the Market Fit

The choices for the next component are somewhat more broadly varied. Steven Gary Blank’s Customer Development model seems to be quite popular at the moment for consumer, mobile and web apps in particular. A few years ago, Geoffrey Moore’s Crossing the Chasm was the similarly big hit for enterprise marketing and sales technologies (among others).

The basic issue of entering a market with a new product or service is this: how do I efficiently find enough of a customer base for my product for my company to survive and grow? Finding the answer to this question often consumes as much time, cash and effort as actually doing the coding for a software project, and involves finding the right product/market fit. Both target customers and products go through a number of rapid iterations until a match is found—some products will never actually exist (vaporware, Escher products), and some will never find a real buyer (phantom customers, zero marginal value markets). Some products do poorly because they’ve found the wrong prototypical customer (nonrepeatable sales) or because the market is obscure (unfindable sales) and cannot be accurately targeted.

This is actually quite rare. With the possible exception of defense/aerospace products, if there’s one findable customer, there will almost always be many findable customers. A more common problem is that the first customer is so idiosyncratic in their requirements that an overspecialized product emerges. That means that a sale to your mom doesn’t count. A sale to the world’s largest widget maker doesn’t count. Your prototypical sale should be to a complete stranger, or the eighth largest widget maker—someone who doesn’t have a motive to make you feel good, or more market power than anyone else is exactly who you want as a first customer. Get the exceptional customers as you get everyone else.

Maps and Models

At the top of the stack, it’s important to have a strategic view of the startup. Not a fussy, detailed one, but one that lays out the important assumptions, capabilities, attitudes and moving parts that go into your idea of how things are going to come out. This enables you to test your own thinking cheaply, and to have a feeling that you’ve got options in any situation.

A lot of good ideas are only good while they’re in your head, and getting them outside of your head in some form or other is going to let you find out which are really viable, which are viable only with optimism, and which ones shouldn’t really be tried at all.

Third, it gives you the ability to see where problems may arise, and what might be done about them as they do. This gives you a full range of responses and a greater degree of robustness for the business than you would otherwise have.

The Chicago Entrepreneurship faculty have come up with a lot of useful tools in this area: Steven Kaplan’s OUTSIDE/IMPACTS model for investable ideas, Scott Meadow’s Unit Model, and Jim Schrager’s New Venture Strategy provide excellent tools for building strategy maps.

The concept of the “business model” as commonly discussed in Web 2.0 circles is a less attractive variant. Too often, business models as described are actually elaborate work-arounds that arrive at the solutions of “somebody else is going to pay” or “somebody else is going to do the work”, when they should really be saying, “I’m going to make something that my potential customers will value enough to pay cash for. Here’s how.”

The “Here’s how” component.

This one is up to you. There are a lot of books and experts that can tell you better ways to and how. They can’t really tell you how to in the first place—otherwise, why are you involved at all?

No comments: