When it comes to timing, Joe Kraus knows when to hold them--and surely when to fold them.
As co-founder of Excite, one of the earliest Internet search companies, Kraus belongs to that class of tech entrepreneurs that flew close to the sun in the late 1990s. He also ranks among the lucky few who escaped the Internet meltdown.
Nowadays, Kraus is applying lessons learned from the go-go days to his latest venture, JotSpot, a San Francisco-based company that builds applications around online collaborative authoring systems. With wikis beginning to find their place in the business world, Kraus believes the technology could lead to another Internet shift in the way corporations work.
How far the wiki phenomenon will go remains an open question. Critics believe the hype has far outstripped the potential--let alone the performance. But Kraus believes wikis can function as lightweight alternatives to heavy-duty CRM (customer relationship management) applications where corporate developers use them as tools to build customized enterprise applications. CNET News.com recently talked with Kraus to find out about the future of wiki technology.
Q: Considering the mania around wikis, can you explain why we shouldn't think this is going to become the second coming of the dot-com? The resemblance seems striking.
Kraus: Wikis in general?
Kraus: I think all technologies go through the typical hype cycle. There will be a point at which wikis are perceived as good for absolutely nothing and then they will find their natural rhythm and become a staple in the way that people manage and collaborate.
Some new, new technology never replaces the thing that happened before. It finds its own place in the middle ground.
In the corporate setting as well?
Kraus: Absolutely. If you look at any technology--from Bluetooth to blogs--they have gone through this cycle of monstrous hype where there's an absolute crash before they then find their natural place.
Is that true about blogs? They haven't hit the crash phase. It's only the last year-and-a-half that they have really become hyped to ridiculous levels.
Kraus: That's true. I think they're probably still on the growth part of the curve, but certainly technologies like Bluetooth went through this. It's only now that we really are starting to see interesting applications of Bluetooth that probably make sense as opposed to what we were talking about before.
So, wikis are not a fad?
Kraus: I don't believe so. When I started to look at this market two years ago, what was interesting was that it felt like the Internet in 1993. Back then, you had this technology that was in the hands of tens of thousands of companies--where the bottom ranks of the organizations, the engineers, were using it and the people at the top had no idea. To me, that is indicative of a technology uptake trend.
OK, so how do you make money from this?
Kraus: Well, JotSpot is a subscription service. There's a hosted service that we provide, and there is something we're calling the JotBox, which is an appliance-based version that larger enterprises tend to buy when they want to host it behind the firewall. But we will remotely manage that appliance into the software update.
And you'll charge using a cost-per-user model?
Kraus: Originally, JotSpot was playing around with a cost-per-user model for its wiki, but I think there are problems with that model.
What's the biggest one?
Kraus: It punishes the exact act you want to reward, which is sharing the wiki. It's a collaborative tool, and so the more people who collaborate, the more it costs. I don't think that's correct. You want to tie it to how much value you get from the service, which I think applies more closely to how many pages are in the wiki.
What gives you confidence that user behavior is going to shift in a big way toward the use of wiki-like products such as yours?
Kraus: The adoption of new technologies is always difficult because you've got existing habits, and the hardest thing to change is an existing habit.
What are the habits today that potentially compete against wikis?
Kraus: You have e-mail, you've got shared folders--the truth is that it's going to take time. Wikis don't replace e-mail. It never happens. Some new, new technology never replaces the thing that happened before. It finds its own place in the middle ground.
There's generally a broad sense that you get too much e-mail. E-mail is a great communication vehicle, but it's a terrible vehicle for storing conclusions or decisions. The challenge for any company in a space like ours, or any company providing wiki technologies, has been how to tie into those existing habits. How do you make publishing to the wiki as easy as writing an e-mail? This takes time, just as any
new habit takes time. I remember back in 1994 with search. Search wasn't a habit; it took five years to become a habit.
It took awhile for desktop publishing to take off. It entered the corporate market through the back door. So with wikis, you may have a long period before it becomes widely accepted. Given that, should you expect that some other obvious players will get into the business?
Kraus: Sure. I think any exciting category attracts competition--otherwise it wouldn't be exciting. When we started at Excite nobody gave a crap about search. Suddenly a lot more people started caring, and then Microsoft started caring. If some really big players think that this is important to the future of their company, that would be an indication that we're onto something interesting.
How long will wikis take to really catch on with the majority of computer users?
Kraus: Let me use the desktop publishing example. Twenty years ago, typewriters were cheap, widely available, and everybody knew how to use them. But they had crappy output. At the high end, you had professionals doing expensive work, with expensive machines and beautiful output.
When DTP (desktop publishing) came along, typesetters thought it was a joke. But it appealed to the average consumer who could use typewriters. And over 20 years, it just completely abolished typesetting and typewriters. Well, I think that's kind of a good example here.
In other words, this is part of a broader technology shift that takes place over time, rather than through a sudden embrace?
Kraus: Whenever you're trying to create a category, it takes years to really take hold. Early on, you asked why this isn't the second coming of the bubble. I think part of the reason is that, if you look at the way companies--including JotSpot--are scaling and building in this environment, they're not doing it the same way we did (during the late 1990s). Back then, it was, "OK, the market's here right now. So let's blow the company up to hundreds of people." It's just not that way now. It's much more measured and realistic.
When we started Excite, we had to buy expensive Sun servers and disk arrays and all sort of stuff. Hardware costs nothing these days.
What's the right analogy? Search?
Kraus: They didn't ask for search in 1994, either. When you put people in front of a search box in '94 and said, "Search for something," they said, "What do I search for?" and I said, "Anything you want." (They would say), "I don't really know what I want," and then eventually they would type their name and they would judge you whether or not you provided good information on them. But they at least could categorize you then and knew what to do, and it took five years for that to become a habit. I think the same is true here. Most people don't know they want to publish an application.
One of the key things about wikis is that they evolve. They don't just start. They don't begin with the end in mind. They just begin, and you can change them over time. You can change the links; you can change the structure they involve. Users are authors. The difference between a user, a reader and a writer is very small. Also, they integrate really well.
Is it easier to run a company on a limited investment than it was, say, when you started Excite?
Kraus: Oh yeah. It took $3 million to get Excite from concept to release, it took $100,000 to get JotSpot from concept to release.
Why is that?
Kraus: Three reasons. When we started Excite, we had to buy expensive Sun servers and disk arrays and all sorts of stuff. Hardware costs nothing these days. Second, you don't pay for compilers, app servers, Web servers--any of that anymore. You use Linux, you use Tomcat, you use Apache--I mean, the infrastructure software is free, essentially. And the third reason is that start-ups have access to offshore labor in a way that they didn't have in the early 1990s. IBM had access to offshore labor in the early 1990s but Excite as a start-up didn't.