Skip to content
March 23, 2005 / jnolen

JotSpot tackles enterprise deployments

Well, don't I feel ill-informed. It turns out that JotSpot actually did do something interesting at PCForum yesterday. They released the JotBox! (Cute name.) I found the announcement through this eWeek story. How despressingly old-school.

Jot is no doubt responding to the same pressures that SocialText is feeling from enterprise customers. Oh, wait. I suppose that might have been apparent from the press release subtitle, "Company Responds to Enterprise Customer Demands with "JotBox" Appliance." I should really read the entire press release before I start to spout off.

The press release also quotes Joe Kraus saying, “While most of our customers prefer the JotSpot hosted service, we understand that large businesses want total control over their corporate data.” I suppose that may be the case now that most customer would prefer a hosted solution, but I am confident that it will not be the case for long. My guess is that the vast majority of businesses larger than 15-20 people are not going to be comfortable with a hosted service like this. It’s a combination of security-paranoia and the unwillingness to rely on an external partner for critical business infrastructure. Even if it might be more secure and more reliable in actuality, enterprises aren’t yet comfortable with it.

Salesforce.com is the canonical counter-example of a critical piece of infrastructure which enterprises are comfortable hosting externally. But I truly think this is the exception, rather than the rule. I think the Salesforce.com ASP model succeeded primarily because the users were already accessing the information from off-site — so institutions did not perceive a greatly increased security risk. Also, the sales-lead database, while important, is not as critical as the corporate documentation which might end up on the wiki.

As I mentioned yesterday, I firmly believe that both JotSpot and SocialText are going to have release a software-only version of their products. Enterprises are going to demand the flexibility to run the software on the hardware and OS that they choose. They’re going to want control over securing and patching the OS and underlying software. They’re going to want to configure the wiki itself for scalability and redundancy.

For example, some organizations standardize around a particular distribution of linux. They install the same applications on each box, often off the same cloned drive each time. And they keep all their boxes are the same version and patch level, especially if they’ve automated it well. So asking an operations department to maintain 200 identical RedHat boxes and one customized Debian box for a third party app seems like a hard sell.

However, productizing software is much harder than slapping the code in a dedicated appliance — a known environment. You have to worry about differing operating environments, upgrades and downgrades, and users-as-sysadmins. You have to write more code. You have to write more documentation. You have to deal with more support requests. Altassian’s Confluence does all of these things very well. I don’t know what percentage of their time they spend on this type of development, but I imagine that it is significant.

Aside: In the case on inside-the-firewall software, open source code becomes a critical issue. Using a hosted service, source code doesn’t do you a lot of good. You can’t customize it. You can’t deploy bug fixes. You can’t extend it. You’re entirely at the mercy of the company doing the hosting. But installed software, even if it is installed in an appliance, changes everything. All of the arguments for open-source become relevant again. If the enterprise versions of these products enjoy the adoption that I predict, then I hope that Jot and Socialtext take the opportunity to reevaluate their stance on allowing access to the source code, as does Confluence. I know it would become a much bigger factor in my purchasing decision, if I were looking at installed software instead of a hosted application.

I think that the wiki vendors are going to be forced to offer installable software. If it were me, I would be starting this effort now, and would probably expect to see some announcements in late summer of early fall. Well, I’ve placed my bet. Let’s see what happens.

One Comment

  1. Ludovic Dubost / Mar 23 2005 10:04 am

    Hi Jonathan,
    Nice take on the JotBox.. What seems odd to me, is that Jot is saying “large businesses want total control over their corporate data”.
    Large businesses want also to control the software that they run and the environement on which they run it.
    As you say, source access is another important thing that large customers need when they wish to customize the software.
    For XWiki, we want customer to have complete freedom on how to run the software. Hosted on xwiki.com or on their own environement. If they want to run xwiki in a portal server (many large corporations have or are deploying JSR 168 Portals), they should be able to do it. If they want to start hosted and then migrate their data to their own installation, it should also be possible.
    One thing that I believe is important is the freedom to choose your provider. Currently it seems difficult to provide standards which allow to migrate 2nd generation wiki data from one wiki to another. So the only way to provide independence from providers for such a product is the freedom to *not* have a commercial relationship with the wiki vendor.
    I’m not saying here that customer don’t want or don’t need commecial relationships and vendors. I do believe that they actually need it.
    What I’m saying is that the time clients don’t have the choice to go away is behind us. Clients should and will stay because there IS value provided, no because there WAS value provided or promised.
    I believe a lot that XWiki will get and keep it’s customer because it allows them to walk away any time.

Comments are closed.

%d bloggers like this: