The big news in Seattle just before the holidays was Amazon's plan to lease up to 1.6 million square feet for a new corporate headquarters near South Lake Union. (For you non-local types and newbies like me, geographically, that would be near here.) As a frequent and happy customer, I am pleased by Amazon's success.
Like many of you, I no longer haunt bookstores. Rather, I read book and movie reviews online or in the New York Times, then order the interesting ones via the web. Undoubtedly, part of this is because I lack the time that I had when I was a poor graduate student years ago. I also miss the serendipity of wandering the physical store, but all of this is more than offset by the convenience of electronic ordering.
Lessons from Dot.coms
As many of us shopped online for holiday gifts, I also found myself thinking about the rise of effective electronic retailing from the Strum und Drang that was the dot.com boom and bust. Amazon has been successful because it recognized the critical elements of early (and current) electronic retailing:
- High value, small items that can be shipped rapidly and at low cost (books, videos, music and electronics)
- Inventory and logistics for supply chain management
- Larger selection than commonly available at most physical merchants (e.g., most chain stores)
- Consumer participation via reviews and ratings
Conversely, Pets.com was an archetype of the dot.com fallacies. Remember the relentless advertising to build brand recognition via the sock puppet and the "Because Pets Can't Drive?" slogan? By the way, they did succeed. The sock puppet was a successful brand, and you can still watch the commercials on YouTube. Today, the pets.com domain name is used by PetSmart.
Like many web startups, Pets.com also accepted losses on each order to build volume. This was the get big fast rationale of the frenzied 1990s. Alas, there are some old-fashioned economic truths; profit and loss statements do matter. Finally, dog food is heavy (hence, expensive to ship), widely available in local stores, and has (relatively) few brands.
Web 2.0 Futures
We are now in the midst of the web 2.0 frenzy, with echoes of the irrational exuberance from the 1990s. As we collectively identify new interaction models and understand social acceptance, we would do well to remember the past. What are the economic value propositions from social networks, complex web mashups and services? What balance between privacy and social sharing is acceptable?
As an example, consider the recent brouhaha over Facebook's Beacon service, which publishes member product purchases, and changes to the sharing policies for Google Reader. The GMSV blog has a great summary (Developers are from Mars, users are from Venus) of these social expectation issues.
I absolutely believe there are powerful value propositions, just as there are for electronic retailing, but we must separate the hype from the reality. Unlike many of the web 1.0 companies, which delivered physical products, social networking and cloud computing are more about services, with concomitant advantages and disadvantages.
Because the product is intangible, delivery is inexpensive, if not free – just a consumer broadband connection. However, the differential value proposition is also different. What services are valued? What interactions are desired and accepted? What price points are viable?
The answers to these questions determine the requisite IT infrastructure (data centers, broadband footprint) and software capabilities. The game is afoot.
If you are interested in the infrastructure behind web services and cloud computing, you will enjoy the High Scalability web site. It pulls back the magic curtain – just a bit – for a peek at the experiences (good and bad) building scalable sites. Because infrastructure provides competitive advantage, the lore is often closely guarded and a few of the observations are dated. However, it is a fascinating and useful read. I recommend it.