As James mentioned today and I alluded to, we got the chance to hear the pitch from open source messaging startup Zimbra yesterday. Formerly Liquid Systems, Zimbra’s the new home of Scott Dietzen, himself formerly of BEA. In a nutshell, it’s an open source, cross-platform messaging platform competitive with commercial products like Exchange and open source projects like Hula. The source will apparently be available in September and I’ll reserve final judgement until I can build it in house or get my hands on a live version, but having seen the demo I’m inclined to agree with Peter Yared that it’s one of the more impressive Ajax-style clients I’ve seen. More impressive than Gmail simply because of the high level of calendar integration. The license, for those that are curious, is an MPL derivative and thus is combinable with selected other licenses.
In an article by Bob McMillan, my fellow analyst Stacey Quandt is quoted as saying of Liquid Systems and other would-be Exchange challengers like Scalix, “It’s not clear that any newcomers will be successful there.” She does, however, go on to qualify that statement with the following, “There is going to be some low-hanging fruit of companies that don’t need the requirements of Exchange.” Given my feelings on the current state of collaboration (see here or here), it’s to be expected that I’m more aligned with the latter than the former.
I think there is indeed a significant market opportunity for messaging providers, but not necessarily for the reason that Stacey mentions. Sure, there are doubtless organizations out there wondering why they’re paying for some of the bells and whistles in Exchange (public folders, for one), but I think the opportunity for a vendor like Zimbra is multi-faceted.
- First, there’s the open source card. Anybody reasonably educated on the subject would probably (but not definitely) acknowledge that open source is not a panacea for development challenges, security vulnerabilities and other software related issues. But that’s reality; perceptions are what counts. And for many these days, open source is perceived to be intrinsically superior at producing secure code rapidly. Not for every customer, of course, but we’re listening to an increasing number of enterprises and governments – particularly abroad – prioritize open source.
- Second, there’s cost. Again, it’s a perception v reality thing, because while the perception often is that open source is free, the reality is that many if not most commercial entities pay for their open source products in the form of service and support contracts. Even allowing for the commercial pricing, however, open source products are often lower cost than their commercial counterparts.
- Third, there’s innovation (or lackthereof). Scheduling a meeting is just as difficult as it was a year ago, two years ago, five years ago. Handling timezones? Still not ideal. Tagging? Not available. SMS notification? Nope. Etc. Etc. The Hula folks said it best, I think; here’s how Nat Friedman described the goals for Hula when it was launched:
Our direction is distinct from other open source collaboration server projects in that we’re not trying to build every conceivable bit of functionality that someone might consider “collaboration” into the server. Instead, we are focused on building great calendar and mail functionality. The dominant collaboration solutions today (Exchange and Notes) are built on a pre-Internet design and are just no fun to use for real people who live on the web, who collaborate across organizational boundaries (or who don’t have organizational boundaries to worry about), who want light-weight tools and URLs for their meetings and their appointments on their cell phone and so on.
Is there room for new ideas in messaging? Seems pretty clearly yes to me.
- Fourth, there’s simplicity. Several months ago, I was talking to someone who was at Lotus a decade ago when Exchange took off, and their hindsight perspective on what fueled Exchange’s growth was that it was simple to deploy and use, versus Lotus which was more complex. I see similar opportunities here for some of the new entrants. The point here is not to try to be all things to all people – not to out-Notes Notes, in other words; it’s to provide a lighter weight messaging product designed on top of the advancements we’ve seen in communications infrastructure in the last few years (including the improvements in bandwidth which make hosted solutions realistic, thus prioritizing multi-tenancy). We’ve seen this simple v complex dichotomy play out countless times in the open source world: Linux was simpler than Windows, MySQL than DB2/Oracle, JBoss than WebSphere and so on. Lighter weight approaches won’t replace the more complex products, just fill a different need.
- Last, there’s the cross-platform issue. Are there shops that are now and will always be Windows only? Absolutely. A lot of them. For those folks, Exchange is a great option. But with IBM investing a $100M in the Linux desktop via their cross-platform Workplace product, and the resurgence of OS X as a platform of choice amongst the technical elite as well as a certain volume of home users, a single platform approach seems less attractive than it once did. As an example, Scoble claimed a while back that Outlook Web Access – the browser front end to Exchange – should be considered one of the best uses of Ajax in the industry. If you’re on IE, I’d agree. It’s very slick. But if at some point you consider another platform – such as Firefox, it’s dramatically less so (see inset pic). The Firefox version is fine and more or less functional (though lacking minor but irritating things like a “Check All” box or a preview window), but it’s not going to win any awards and it can’t be argued that the experience is degraded significantly from the IE version. Is there then an opportunity for vendors to deliver a client that renders successfully across a variety of platforms and browsers? Well, we just heard of a 60K seat enterprise deployment of Firefox the other day, and an article in this month’s Enterprise Open Source Journal (why no web content, guys?) claims that Boeing’s another firm standarizing on the upstart browser. So my answer is: yes.
Now does any of the above mean that Zimbra’s guaranteed success? Nope. I’m very partial to the direction that Hula is taking. But is the messaging market overall impenetrable, the province solely of the IBM, Microsoft, and Novells of the world? Well, a few years ago there were analysts saying that the operating system market would be a two horse race between NT and Unix, and that prediction seems to have missed a rather significant opportunity. Same with the relational database market. So I’m willing to accept the possibility that one or more of the messaging startups and/or projects might have some success.
The interesting thing is that our research around messaging is more than just academic for us: we’re actively looking for an Exchange replacement. About a year and a half ago when I was on Windows, Exchange was more than adequate for our basic needs. But RedMonk has become a multi-platform (Linux, Solaris and Windows) organization and Exchange is no longer the best solution for us. I get by with Evolution’s Exchange Connector, but it’s not ideal. Whichever of the open source, cross-platform messaging providers is able to get a hosted solution out the door first – either direct or through a reliable third party – is likely to get our business. All two accounts of it