Original URL: http://www.theregister.co.uk/2006/03/13/shortfalls_of_enterprise_technology/

When 'enterprise' is self-defeating

Solutions for end-users - who would have thought?

By Philip Howard

Posted in Developer, 13th March 2006 10:17 GMT

Companies like to describe their products as being suitable for enterprise-wide deployment. However, this is by no means always a good thing. Indeed, an "enterprise" product, by its very nature, may be precisely the reverse of that.

Consider the nature of "enterprise" products. They have the performance to supports thousands, if not tens of thousands, of users, they can scale dramatically, can be distributed across the organisation, have failover and load balancing, and fault tolerance and disaster recovery facilities built-in. There is no question these are good things.

However, consider how enterprise products got to be that way: they are highly engineered and were typically designed to be enterprise-class products in the first place. As a result they tend to suffer from creeping elegance (that is, they have more and more features added to them, a large number of which are rarely, if ever, used), they are complex and cost a lot of money.

Now, we need to classify different types of enterprise products. There are some products that are designed for a limited class of users. For example, an enterprise data warehouse (EDW) caters to a specific subset of the user community. Similarly, ETL (extract, transform and load) tools are primarily designed for use within the IT department plus a few specialised business analysts. However, there are also enterprise products that are literally aimed at the entire enterprise: they might reasonably be used by very large numbers of people, in one way or another, throughout the organisation. For example, business intelligence has always held out the dream (if not the reality) of this possibility and the same applies to corporate performance management and to enterprise content management, among other areas. Arguably, it is these technologies, which are designed to reach out to the whole enterprise, literally, that are the true "enterprise" products.

The truth is that many (though not all) of these enterprise technologies have failed and are failing to actually reach those parts of the enterprise that they claim to address. This is partly because these solutions are too complex, it is partly because the products in question have not really been designed for end-users (such facilities often being bolt-ons rather than inherent design features), and it is partly because the products (being enterprise-class) are too expensive to deploy as widely as they might be.

Why do I mention all of this? Because I have recently come across counter-trend products in a number of areas. For example, in the content management space, Xythos (which has just announced a UK partnership with InTechnology, with which it will be addressing especially the academic and research space) talks about basic (rather than enterprise) content management, precisely because it is basic content management that most people need and not all the bells and whistles that get put into products for the power users. Both Microsoft and Oracle could be placed in a similar category when it comes to content management.

In the Business Intelligence space, the problem is somewhat different, but it is clear that much BI software is still shelfware. One company that is addressing this issue is Inflection Point, which has designed a BI tool specifically from the point of end-user in the first instance rather than after the event. I will be discussing this further in a subsequent article.

What is encouraging about these developments is that we may be seeing the emergence of solutions that are actually designed for end-users in the first instance rather than as an afterthought, and that’s what you really need if applications are truly to be used "enterprise-wide".

Copyright © 2006, IT-Analysis.com