Sell direct or embed? (click here)

Tom Evslin, a former entrepreneur and senior officer at AT&T and MSFT posted an excellent piece that is highly topical for young companies which are exploring alternative revenue strategies. He begins his post by highlighting the news that Dash Navigation has laid off 65% of its workforce (venture backed by Sequoia and Kleiner) and is now looking to license its software to GPS manufacturers. In essence, switching from an end-user application to an infrastructure provider to its former competitors.

During the dark days of Apple, Steve Jobs was roundly criticized by analysts who noted that the world moved beyond vertical integration (where you make your own chips, OS, and applications) and towards horizontal integration (where you partner with various folk to produce a functionally similar box as the next guy). The rabble spouting a tyranny of dead ideas pointed to the success of Gateway, Compaq and the upstart Dell saying that only by licensing its OS would Apple prosper. We will never know if Apple could have been the 'next' Microsoft, but I suspect that the path of tight chip, OS and application integration is what gave us the iPhone and iPod.

Tom does a great job focusing on examples of license vs manufacture in the hardware business. This strategic quandary of being a licensed infrastructure element for someone else's solution (OEM), or a direct to end-user business, is amplified in a down environment where the necessary capital to build a sustainable revenue stream is in short supply....and expensive.

This key strategic element is today being faced by many software/internet companies. Back in the halcyon days of the internet, Google crossed the chasm when it moved from licensing its search, as an infrastructure element, to folk like Yahoo, to building its own brand by going direct to end-users (and advertisers). Adopting a business model that rapidly weaned its dependency from license revenue streams and pioneering search based advertising was a key success factor. Inktomi, an independent search provider, was acquired by Yahoo for $235mm as a response to Google going direct and threatening Yahoo's franchise. They never made such a transition. Phoenix Technologies is taking tentative steps down the same road with its HyperSpace solutions.

I suspect that in an era of ever declining cost of IP development, more standard API's, and multiple workarounds to IP patent protection, the long-term value of being an 'arms' supplier has been deeply compromised. Building sustained shareholder value depends more and more on building a broad customer franchise and reducing dependencies on a small group of customers, where you are ultimately at their mercy. Building such companies, in a capital efficient manner is the challenge today facing venture investors and software/internet entrepreneurs.