How do most portals fail?
One example of a failure is when you build a portal and nobody comes. Most times, it's because you haven't embedded any application, or data source, that users need on a daily basis. We call that the failed portal, a Teflon portal.
Then there's a more horrifying, or disturbing prospect, for someone buying a portal. You select a portal. It works. People use it. The overall enterprise selects a different portal that also works.
Suddenly, the question put to you is, "Why do we have this portal over here? Why did you waste our time?" You weren't wasting time. You were doing your job. You're stuck as the champion of the portal that hasn't been selected as the standard. And that situation keeps popping up for people.What about Microsoft?
Going forward, Microsoft is also going to be in that group of large stack vendors because they can afford to give away SharePoint in a lot of scenarios; they are charging a very low entry point for it. In my opinion, they are behind the curve from a functionality perspective, in a lot of regards. Their road map, over the next year, will get them to where a lot of people already are. But, being Microsoft, that doesn't mean they aren't going to sell a heck of a lot of portal software. Which vendors are most likely to have an impact in the portal market?
First off, the market is still in a state of pretty dynamic flux. But vendors who have had the most impact on the market over the last year are the large stack vendors. For a long time these vendors, like IBM, Oracle, Sun, and BEA, hadn't quite focused. IBM at one point had three competing portal products. Over the last year, they've focused. IBM brought all their portal products under the WebSphere brand.
You're talking about vendors who have incredibly large footprints, from a client perspective. I mean, Oracle can give you their portal for free (if you have their application server). Also, some of these vendors are highly credible middleware vendors, and when it comes to portal success, the real integration relies on middleware.
Now, from a technology perspective, someone like an Epicentric is having a lot of impact. The question is: "Will they be around?" And that question has to be asked.What advice do you give to people shopping for portals?
You need to talk to an independent third party. And that does not mean a third-party systems integrator. A lot of those have relationships with specific portal vendors, even if they say that they work with anyone. A lot of people ask, "Should I select the systems integrator or the portal vendor first?" I tell them to select a portal first. Most vendors out there will -- not all -- but most will fudge the truth. They will tell you they can do something, and you sign the contract. Then they say, "OK, we'll have it in three months." Where do PeopleSoft and SAP figure into the equation?
Both of them have unveiled strategies over the last year which are, at the very least, flirting with middleware. SAP has their own application server now, their cross-apps strategy. PeopleSoft has their application infrastructure strategy. My viewpoint on it is that their core markets aren't growing as fast as they used to be. They see this as an opportunity for profit.
In some cases, what they are proposing is credible -- not only credible but [with] some potential benefit. But they aren't at the same level as the IBMs and Oracles of the world. Right now, they are at the application layer. They are still running on top of other people's application servers. And it would really, in my opinion, be foolish for them not to do that. They know that.How should I define a portal?
The Gartner definition (which my colleague Gene Phifer formulated) of a portal is a Web interface that provides access to, and interaction with, applications, information and business processes for targeted audiences in a highly personalized manner. Now that's a pretty wide definition. But it's a lot better than a lot of what is out there. Is it better to have a portal than not?
Yes. We're not saying everyone should buy one. We do think there are benefits to a portal when done well. This is legitimate technology. Portals can provide benefits. They are just very difficult to quantify from a strict dollars-and-cents perspective, in many cases.