Writer, consultant, investor and entrepreneur David Siegel is no stranger to the Internet. The author of Creating Killer Web Sites and Futurize Your Enterprise designed his first web site in 1993 — and he’s been strategizing about the Internet’s future ever since. His most recent book, the just-published Pull: The Power of the Semantic Web to Transform Your Business, explores how the overhaul of our information infrastructure that the semantic web heralds ultimately will lead to a performance economy, where the winners will be those companies that understand how to leverage the transition from a world where information is pushed to one where customers pull it on demand.
SemanticWeb caught up with Siegel to discuss his vision and his latest work.
SemanticWeb: You’ve been involved in the Internet since its early days. When you think back to the beginning and the first web sites with which you were involved, would you ever have imagined the need for semantic web standards to emerge, and for the evolution from a web of documents to a web of data?
Siegel: I did. I knew Tim Bray and Charles Goldfarb back in the mid-nineties, and when XML came out I realized that this thing could go one of two ways: Either everyone would build his own internal XML “standards,” and we’d have to translate from one to another, and that would just be another way to build a schema, or we’d start to share XML formats and then the web would become the database and the search engines could do all the work. I said this in 1997. I talked about it at a conference at Stanford, and again at Tim Bray’s XML conference. People didn’t understand it. I told them I was concerned, because the web had a chance to really be this connected thing that took us in a big new direction — the open web — rather than using XML to recreate the databases inside the silos we had already built using internal schema languages. Unfortunately, things didn’t go my way. When I wrote my 1999 book, Futurize Your Enterprise, I tacked on about 12 extra chapters at the end, showing how different the world would be if everyone in an industry would adopt the same XML schema. But then came the tech wreck, the false start of the semantic web (after Tim Berners Lee’s 2001 Scientific American article), and Web 2.0, so it never had much chance to get any traction. About the same time, Bill Joy was talking about something similar. A few of us were, but it just wasn’t the right time.
Even though I started as an “HTML terrorist” and crusaded hard for a visual web — a crusade I clearly won, along with a handful of other pioneering web designers — I also wanted semantic connections underneath and have been waiting patiently until now to try to get it going for real. Finally, it’s nice to see some of these concepts being taken up by a new generation of people who can see the same vision. That’s why I wrote Pull – to inspire more people to join us and get this thing going.
SemanticWeb: Can you briefly explain to our readers what ‘pull’ is, why it matters to business transformation (and indeed why you think business must be transformed), and why you can’t realize that transformation if you don’t have the semantic web?
Siegel: I think it’s clear that all businesses will have to make the transition from pushing information, products, and services to customers to pulling them. Look at media — they are tumbling down the face of this waterfall right now. I don’t want to watch CNN all day to see what’s happening over and over again, I want to pull the news to me according to my tastes and availability. If I’m on my computer and something I think is interesting happens, I want my computer to tell me, and if I’m in my car, I want something audible, whenever the time is right. We want what we want, how we want it, when we want it — not when it’s convenient for the producer to send it down the pipe.
This will invert most business models and change the direction of the supply chain. I think this is going to be a permanent, one-time shift, one that will serve well beyond this century. Many books and management consultants talk about being “customer focused,” but customers still don’t have the tools to pull products and services to them. That’s what we’re building with the semantic web — we’re making our data, metadata, products and services pullable.
SemanticWeb: So is that transformation going to be cost-efficient and seamless or difficult and investment-intensive for today’s cash-constrained enterprise?
Siegel: Great question. I think you can compare it to the web we use today. Think of where everyone was back in 1992: No one knew what a server was, a web site, even a domain name. Companies had to acquire all these new assets, skills, and capabilities to take advantage of the web, and there were huge benefits. I don’t think any company wants to go back to the days before the web, when you had to send out mass mailings and go to a ton of trade shows to find new customers. So there will be new costs — and new opportunities for growth and profit. Conversely, I think it will be expensive for companies not to participate, since they will lose out on the opportunities ahead.
SemanticWeb: You mention in your book something called the Semantic Web Acid Test (that is, it’s semantic if the terms are unambiguous, in a royalty-free format, governed by non-profit organizations, in a form every software can understand). These days, though, there are web services out there-â€“especially for consumers — that consider themselves semantic services but don’t necessarily subscribe to semantic web standards and/or data formats. Do you A) consider such efforts to be a distortion of the real ends of a semantic web, and B) believe that businesses that want the power of pull must themselves abide by open semantic web standards and formats?
Siegel: I think we’re in a serious transitional phase, where putting the word “semantic” in your business plan is practically a requirement, yet few companies are doing much good with it. Similar to what I said about XML. As Dan Connolly says, “The key term in ‘semantic web’ is ‘web’” and I really think he’s right. So the Semantic Web Acid Test helps us know where we’re going eventually. But, in the meantime, I’d rather see companies focus on concepts than keywords. Once people are focusing on concepts, I can see the day when they will hook to an overall topic backbone and that will bring it all together. With Twitter and microblogging and news and video, things are going toward keywords. But the next phase will be to consolidate the keywords around concepts, and then link the concepts to build webs of ideas and content that are related.
It will have to happen organically in most cases (and by fiat in government-related areas). We just don’t have enough momentum at the moment to get everyone using common standards. I want people to be aware they exist and try not to reinvent the wheel as much as possible.
SemanticWeb: You also talk about the concept of linked dataâ€”are you seeing businesses more interested in how adding their own semantically-formatted information to the world of linked data can ultimately help them, or is there still some uncertainty around how open they want to be with their proprietary information? And if so, how does or might that impact the pull movement?
Siegel: Woah, these are good hard questions! Greg Boutin talks about why it’s not getting going that quickly. In my book, I tried to show a few different examples to illustrate that each industry has its own issues. I believe the world of XBRL financial reports is rapidly becoming the best example of linked data, although there’s no RDF and it’s not a graph. It’s a tree based on taxonomies. But it works, and it’s a lot better than the free text and spreadsheets we had before. Then in the area of book metadata I show how it’s much more loosely connected, with many different formats, no RDF, and still you can find book metadata and start to make the connections. I also mention GoodRelations, a product-descriptor graph that I think has tremendous promise and may be the first large ontology to be commercialized and widely used. Drug companies are starting to add their research to the Science Commons and put their data on the open web.
All of these examples are signs of life. It’s not going to be all at once, and it’s not going to be the same in every domain, but the early adopters are already seeing results that everyone else will want, so it’s coming.
SemanticWeb: A year ago you wrote on your blog that you wanted to be Apple’s next CEOâ€”now that Steve’s back I’m guessing that ambition is on the back-burner, but I think our readers would be interested in a quick look at your ideas about how and why this so-very proprietary company has to move into the pull world to be enabled by the semantic, social, linked Web — and whether you see any such transformation underway there in the year since you wrote that.
Siegel: I think every single product Microsoft makes, with the possible exception of its browser, will be completely irrelevant and useless in ten years. If you look back ten years, you can see almost everything they sell today in a previous form, so I’m saying the next ten years will be very different for Microsoft. And I say the same is true for Apple. They’ve got the push model down cold — they push songs to people, and you have to store all these songs on your single device, back them up, and if you want your music on another device you have to somehow get it from one device to another. In the same way, Apple has these two application platforms — Mac and iPhone, and they won’t survive. Who wants to download apps when you can just go to any web page and start using the latest version of the app right there? What developer wants to develop for ‘n’ different platforms, when you can deploy one application on a web site and serve everyone?
Apple’s approach is full of limits, and that has served the company well. But Apple, like every other company, will need to make the shift from Push to Pull, and that should change not only the company’s product line but its culture as well. It’s a bit late, but the recent acquisition of LaLa is an indication that their model is broken. There are no surprises in the upcoming tablet announcement — it’s just a bigger iPhone. But look how the company does with digital services. Look at DigitalMe — it’s behind everyone on the curve. It’s embarassing how bad it is.
Within five years, Apple will have to stop designing beautiful hardware and start designing seriously beautiful data centers. They have a long way to go in online services, and they know it.
Here’s a great example. The first thing they should do is buy ZumoDrive and incorporate that into everything, so that cloud-based storage is built into all their products. But they can’t, because their products come with memory, and they make so much money when they announce the next generation that has double the storage of the last. So they are caught in a trap where they go for the short-term profits at the expense of the long view. In the long term, I don’t care how many times they double the local storage of their devices — the web has an infinite amount of storage capacity, and that is going to win.
Look, Apple is an insanely great company. It’s living proof that leadership counts. Vision counts. The current vision is good for another five years. After that, this company will be dragged into the pull era, and I think it will need a leader who can go with the flow. I respect Steve for all his accomplishments, and certainly there’s a lot of milk left in the cows he’s built. But they are now sacred cows, and that will have to change. If Apple is going to turn the corner, it’s going to have to slaughter those cows in 5 to 10 years. I may not be the guy to lead the company where it needs to go, but I hope they find the right person. They have great management and great leadership, but the corporate culture isn’t ready for the shock of becoming a pull company, not yet at least.
SemanticWeb: Taking off from that, what industries or businesses are or will be the leaders in enabling the pull transformation you envision?
Siegel: I think XBRL is a great example of doing something simple that turns out to be quite hard and yet very rewarding. There will be much more built on top of it. If I had an investment fund, I’d be starting companies in the financial space to build (mostly tools and services) on the success of XBRL. I think GoodRelations has a great chance to kick some ass in the product space, which is near and dear to my heart. I’d love to be on their board. I think Semanti.net is very promising, but I don’t see them breaking through any time soon. I think there are some potentially exciting developments in real-estate listings and all the loans and asset-tracking tools you need to buy and own a home (see chapter 9 of my book). And there are many more potential areas for breakthroughs I talk about in my book and my new blog.
SemanticWeb: So right now are you investing in any semantic web companies or related vendors you think can help businesses get to ‘pull’?
Siegel: I’m not really. I’d love to be. I’m hoping to raise an early-stage fund to build some companies I think will be winners and to help forward-looking entrepreneurs bring their visions to reality. I’m a shareholder in GiantChair, which is a service-model tool for managing book catalog data, and that’s already part of it, but it’s a small stake and that’s about all I have at the moment.
I guess if I had a fund I would focus on transitional strategies as well as long-term truly semantic web companies, but I do have a tendency to be a bit too far ahead and a bit more optimistic than the reality of market adoption, so I would have to pace myself. We really are at the very beginning. In about three years, I think we’ll start to see things really pick up. I hope to be at the center of that storm, as I was with web design back in 1994-99.