Launching an enterprise-wide search application can be an overwhelming task, especially for large corporations with a multitude of sources and repositories and strict levels of confidentiality. b-i recently implemented such a project using Idol from Autonomy. Here is a high-level summary of the steps to think about when preparing, configuring, customising and deploying an enterprise search project.
1. Define the objectives and business benefits
Identify who are the stakeholders (ie who are searching), interviewing them, understanding the information they are looking for, which business processes is search part of (what they do with the information found), document current content responsibilities, identifying security policies , etc..
It is also important to understand the benefits that the target stakeholders are expecting.
2. Define a desired scope and inventory of repositories
Define the scope of the search in terms of business, business processes, repositories, applications, content, etc Create an inventory of all the required repositories
3. Identify the current search applications
Identify the current search applications and defining the various options about what to do with them once the new solution is deployed.
4. Define requirements
Define and agree which requirements are necessary to meet project objectives, which criteria are critical to making the solution decision. Decide the scope of the RFP .
5. Select a solution
Prepare the RFP to select the solution, deciding which vendors would receive the RFP, evaluating their responses, selecting the most appropriate vendor and signing the project contract. A Proof Of Concept with the short-listed vendors would be highly recommended.
6. Implement a relevant user experience
The way the user interface is design to present the results will be highly critical and needs appropriate attention. A faceted navigation is typically of great importance to enhance the search experience.
7. Develop and deploy the solution
Set-up the project and its team, construct the solution, test it and optimize it, document, train, and finally get it accepted by the user and IT community.
8. Monitor and improve
The importance of establishing and monitoring system performance parameters is crucial to continuously improving the user search experience. This also includes the regular reporting on search usage statistics and improving the effectiveness of the search process in terms of delivering relevant documents.