by David Provost
Inevitably, questions about whether and/or why the SW is so slow to be adopted arise and I think it's worthwhile to respond. First, let's take a step back for an extremely brief review of the World Wide Web's (WWW) evolution. After being released in 1991, the Web made its first steps into e-commerce around 1995 - four years later. (Note that at the time, any hint of commercialism was blasphemy to the people that had populated the Web to that point.) In the late 1990's the world experienced a colossal explosion of innovation and Web adoption. Millions were being poured into Web ventures (anyone remember webvan.com? They IPO'd November, 1999 and raised $375 million. The company went out of business in mid-2001.)
The buzz and the vast sums certainly attracted attention and adoption, but at its core, the WWW was a combination of relatively simple technologies like HTML, TCP/IP, HTTP, lightweight browsers, and modem technology that had been around for decades. These latter factors were central to making it easy to go online and pursue your interests, whether they might be chatting, reading, shopping, or seemingly aimless "surfing." In any case, it took about ten years to go from launch to boom to bust and seems to have left the impression that success online is defined by getting big and doing it fast.
The SW is considerably more complex than the WWW. As an extension of the WWW, it's considerably more powerful, but it ain't easy stuff to learn. Further, at least for the moment, there aren't any visible applications that are as compelling as what people were exposed to when the WWW began to take off (and I don't consider development tools to be mainstream applications.) So considering its complexity, "invisibility", and the fact that the initial SW constructs were released about a few years after the WWW, I think we're doing pretty well. Note that W3C issued its XML 1.0 recommendation Feb 10, 1998.
In fact, given the number of technologies (read: standards) required to make the SW really perform and how recently some important and fundamental recommendations have been issued (OWL: Jan '04; RDF: Feb '04; SPARQL: Jan '08), I'd actually say the SW is moving along quite nicely.
Getting back to reality, I have to admit to wishing that the SW emulated the adoption curve of something like Facebook and that we were all making millions - daily! But the fact is that the SW may be around for far longer than Facebook and it may reach far greater heights. I've been at this since late 2003 and given the level of interest I'm seeing from a growing number of major players, I think I'm going to stick around for a while longer.
I try to be a regular attendee at the monthly SW gatherings in Cambridge and I recently gave a brief presentation to this group. In case you're not familiar with the audience, that's a pretty novel idea - I'm purely on the business side of things and haven't written a line of code since FORTRAN was big. In the meantime, the audience couldn't be more deeply involved in SWT and W3C personnel are almost always present. I pressed on.I was motivated by the fact that I tend to be preoccupied with wanting to know who's going to buy SWT, exactly what problem do they want to solve, how much will they buy, and when will they buy it? Questions like these lead directly to wanting to understand the process of innovation, adopting rates and patterns, market entry points, and a whole lot more. With this in mind, I figured this would be a good audience to test my thinking with and maybe get some good ideas in return. I was warmly received and surprised by the active and constructive conversation my comments evoked. Here's a recap:
This first slide is a general view of the evolution of IT within large scale enterprises. I've found it very helpful in imaging the flexibility, span, and availability that modern computing environments are striving for.
This next slide really puts adoption curves in perspective. I'm often asked about the adoption rate of SWT and I respond that I think it's doing pretty well. The slide below points out that it took thirty years for the typewriter to achieve a significant portion of the market, which means it can be hard to know when you've won the game (or reached a tipping point in product/market evolution.) In the case of the ice industry (when people had ice boxes in their homes), as the refrigerator began to make serious inroads into this market, the ice industry was racking up record profits, so you'd think they'd be well financed to either evolve or defend their position. As we all know, no one uses iceboxes anymore. Jim Utterback's books and papers on innovation are a major source for this slide.
The final slide is the most interesting. Among social scientists that research the process of innovation it's referred to as the "Roberts Familiarity Matrix" after Ed Roberts at MIT Sloan. It's always reminded me that while people can learn and adapt, it helps to communicate in terms that are familiar to them. Note that the squares with asterisks indicate where the most successful companies and technologies are found.
Here's what the headers mean:
Technologies or Services Embodied in the Product: Refers to the components that are intrinsic to the overall product.
Newness of a Technology: The degree to which that technology has not been commonly employed.
Familiarity with a Technology: The degree to which knowledge of the technology exists but not necessarily in the context being considered.
Market Factors: The environment, situation, or setting to which a technology is applied.
Newness of a Market: The degree to which a technology has not been applied to a given environment, situation, or market.
Familiarity with a Market: The degree to a potential application of the technology is understood, but not necessarily through direct experience.
Base: Common knowledge and common application of the technology (e.g., part of the everyday course of business).
In English, this means that technology vendors approach the Base square as they combine their technical knowledge with increasingly clear knowledge of how a market might use their product. Sort of like "I know hammers and all the technology behind them. Here, use it to write on a blackboard." vs. "I know hammers and all the technology behind them. Here, use it to drive nails and build houses." I see this as the theoretical underpinning to the need for domain knowledge that all successful vendors strive for.
Conversely, customers approach the Base square as they combine their deep market knowledge with increasingly clear knowledge of how to use a technology to their advantage. Sort of like "We bought this CNC lathe because we needed something that would take us to the grocery store - it's no good for that." vs. "We bought this CNC lathe because we needed to churn out a lot of precisely made, high strength steel parts and now our production schedule is booked for the next five years. Can we buy another?"
This means that vendors will make lots of money when they know exactly who will use their products and how. Customers will make lots of money when they know exactly what products will solve their problems and how to use them. Nirvana is mutual convergence on the Base square.
In my career, I've spent about six years carrying a quota as a salesman. Some of the time I was ranked among the top producers and some of the time I wasn't. Add another three years of sales experience for my work leading startup ventures (because in tech ventures, you either make the product or you sell the product) and the sum is a body of valuable, real-world, dollars and cents experience. This experience taught me quite a lot, including the fact that any salesperson worth their salt will tell you to find people that speak your “language”. Sure, there may be an educational process required, but that’s where the money’s made.