Operation successful - patient dead

Over the last five years I've been dealing with the implementation of knowledge-based systems to large enterprises, as a project manager. Part of my experience on these projects has been as follows: everything has been delivered as agreed, acceptance criteria have been met 100 % but the system struggles to find its everyday users. With only a slight exaggeration, one could call this the "operation successful, patient dead" scenario.

Business intelligence departments in larger companies are perceived as a company within a company. They produce reports, analyses, data objects etc and deliver them to other departments in the company (ie. business users), who are able to reach their goals thanks to this information. The BI department therefore has its customers, produces

"information" products for them, and satisfies their need to make decisions based on facts, not just gut feeling. Speaking of information products, the idea of an information marketplace, where the customer goes "shopping" for info-goods comes to mind.

The knowledge system of a BI department should be just such a marketplace. It should be a web application accessible to business where the BI department displays its products, services and the business user consumes and "pays" for it with feedback (either by sheer viewpage counts or by commenting or asking a question).

Why is it so rare then, for knowledge systems built by, or for, enterprise BI departments to become such a marketplace? Why does the usefulness of these systems for business remain questionable or marginal? And why is it so difficult for knowledge-focused initiatives to become a legitimate item on a tight BI budget?

There are many reasons. However, there's always a single common denominator: the project does not take into account the end users of the knowledge system. On average, the project is commissioned and evaluated by people from BI or another technical department. The knowledge system per se contains mainly BI governance methodology documents and guidelines (instead of info-goods such as up-to-date list of reports, business terms etc.) and primarily reflects and meets all the architectural, technological and security standards (instead of focusing on usability and user experience).

The problem is that when you roll out such a knowledge system, it is (maybe) understandable by BI experts, it is possibly 100 % technically compliant with enterprise architecture, but it is never even slightly compliant with the way people work. It is not the transparent marketplace where shoppers easily navigate and quickly find what they seek. On the contrary - they are difficult to understand and the majority of content is unrelated to a customer' s everyday needs. After the first visit to such a system, a business user draws the conclusion "this is not for me" and s/he never returns; the system becomes obsolete before it has even been rolled out properly.

In the next few articles I want to show how such a situation arises even before the beginning of the knowledge system's implementation and offer ideas about how to reach people with a more useful result. It will address the following topics :