Carl Weinschenk spoke with Nicolas Vandenberghe, founder and CEO, Iterating.com.
Weinschenk: Is corporate use of wikis growing?
Vandenberghe: There definitely is more and more corporate use. There is fast-growing adoption in the corporate world. It's a big thing right now.
Weinschenk: Is there a difference between consumer and corporate wikis?
Vandenberghe: There definitely is a difference. There are a lot of solutions to choose from. We cover open source, commercial and hosted software, which is now called software as a service. Whether the user is a corporation, an organization or someone on their own, they should consider all the options at this stage. All three can be the right answer.
Weinschenk: Your company uses wikis internally, you offer visitors information in a wiki, and the content you offer is about wikis. Bottom line, you guys know your wikis. What direction have you gone in?
Vandenberghe: We are based in New York and our engineering is in Romania. We started with Wikka Wiki. It's extremely easy to install. So you use it for a small groups like us, 10 minutes and you are ready to go. Then it gets interesting. The wiki became more important to us and we decided Wikka Wiki was not good enough. A lot of features were missing, like advanced formatting of tables. A lot of the time, there was no response. So we migrated to MediaWiki.
Weinschenk: What additional features does it offer you?
Vandenberghe: Reliability, breadth of features and scalability. When it comes to consumers, those are not concerns. I used a wiki to coordinate an article I was writing. The number-one thing I needed was ease of use. I went through a hosted service.
Weinschenk: What features are important for corporate use?
Vandenberghe: Corporate users want to easily use many features, such as being able to put text in bold, control how to size tables, how to customize the formatting. Another thing that will become useful for companies is the ability to reorganize things in categories and make them searchable. You are going to need these things if, for instance, you are a large bank and are going to use a wiki to manage knowledge. The categorization and search functions are important. On the consumer side, the keys are ease of use, ease of use, and ease of use.
Weinschenk: So what is the trade-off for corporations to get the advanced functionality in their wiki, and what does the organization get in return?
Vandenberghe: On MediaWiki, you have to add obscure syntax and, if you do it wrong, you have to come back and redo it. They say you don't need to know HTML. It's true, but you have to be pretty comfortable with computers. You get the ability to write powerful pages and get scalability. It's more powerful and it's more scalable. You go to Wikipedia [which uses MediaWiki] and see that the pages look as beautiful as HTML pages. You can format in tables, add paragraphs, and add summaries to the page with links. If you are writing an article that will be read by a million people, the investment is worth it. If you are organizing a pool party, it's not worth it.
Weinschenk: What are the trade-offs on the operations side in deciding which way to go?
Vandenberghe: The number-one trade-off is whether you have the resources in-house to do it versus how critical and important you expect the wiki to be. If you are using the wiki to coordinate a few things, you probably want to go with a hosted service. A lot of the solutions are free or very cheap. If you expect the wiki to become a vital tool in which everyone reports everything, then you may want to spend more. For example, we switched browsers. We reported everything, and that's very critical to us. If someone fixed a problem, others know to look there. The third issue is confidentiality and robustness. Obviously, consumer services pay less attention in terms of backup. So if it is enterprise-critical, think twice about using a hosted or consumer service.
Weinschenk: Is that it?
Vandenberghe: The next question is whether to use open source or commercial software. The leading product in terms of quality is an open source product, which is MediaWiki. It's just the best solution in the market right now. Another issue, once you've decided which package to go with, is integrating with existing applications. Let's say a company is on Lotus Notes. If they install MediaWiki, you have the issue of integration with existing systems [and] I think IBM is coming up with a wiki solution within Lotus Notes. If I am a Microsoft shop, I would want to go with SharePoint and use whatever wiki capabilities are available.
Weinschenk: Should a wiki be moderated?
Vandenberghe: When we started, we felt we needed complex rules. Once we built the system, users just looked out for themselves. What I find works is keeping it very open and accessible. It works in many cases, in more cases than you'd expect. People are not as bad as you think and it's so easy to fix things that go wrong. That's a general rule, there are exceptions. There are some new features restricting access for corporations. There are a few things that they don't want to let everyone do. What we are finding is that we often are better off letting people have freedom. It tends to work out. Related to that is that it is the nature of wikis that something bad is going to be corrected quickly.
Weinschenk: What will be next with wikis?
Vandenberghe: The emergence of strikis, which are structured wikis. People say they think wikis are great, but the problem is that they are based on a page. Suppose people want to use a wiki to talk about a piece of art. A page is interesting, but there are a lot of different categories of information: the painter, the gallery, where it was painted, the owner, the price that was paid, what type of art. This information is structured. The reason it is structured is because you want to search. So, say you want to find all the art between $5,000 and $10,000. You need the data to be structured to do that. People are working on structured wikis. We built one ourselves. People have said we've solved the problem and want us to release it as an open source product.