We know from painful experience that enterprise content management (ECM) software often fails to deliver on search capabilities, at least in the direct cause-and-effect way that vendors promise. But as long you start with the appropriate building blocks, an ECM system can become a powerful search tool.
The first step is to broaden your perspective to encompass more than just the ECM repository. Although the search capabilities that come with ECM applications are optimized to return results from within their own stack, you undoubtedly have a lot of content that resides elsewhere: in your finance and human resources applications, on shared drives and individual PCs, etc. This information -- be it structured or unstructured -- is no less important than the information in the repository, so you should include it in your thinking.
One important wrinkle has to do with the cloud. If your organization is storing or sharing documents, images or other data in the cloud, then you also have to account for that -- meaning that you have to learn how to get your search engine to work outside your firewall as well as within it. Depending upon the service(s) you use, this can be easy or hard to do, and it's often something that is done in a later phase of work, if initial scope is an issue. The point is to make a decision about it early rather than let a decision "happen" to you as the situation evolves.
Take the time to tag
One of the most efficient ways to search across all these siloes is to search the metadata associated with each bit of information. Metadata, of course, can include things like the date the information was created or modified, who created or modified it, which department that person works in and which project the information relates to. How does this metadata get associated? The system can apply much of it by using the data in the system log. But people also must apply metadata, and that means figuring out how to make the process as painless as possible by using the likes of drop-down menus and field autocompletion.
Searching the metadata is faster than performing full-content search because the amount of information to be searched is smaller. But because different people use different terms to describe the same things -- and because computers innately view XYZ Corp. and XYZ Corporation as different values -- it's important that the vocabularies used be reconciled and synthesized. Often, this is accomplished by convening multiple cross-departmental meetings, and using an electronic thesaurus to facilitate the automatic "translation" of terms across content stores.
There are tradeoffs, of course. For instance, having meetings and negotiating language can take a long time, and even then there's no guarantee that certain expressions won't be missed. In fact, a case can be made for using full-text search and content analytics to identify most-commonly-used phrases and incorporating them into the discussion to ensure completeness.
Another important tradeoff has to do with aligning the security levels across your information siloes so people can find the content they need even if it's in an unfamiliar repository. From a practical perspective, this means that a single person probably is authorized to dig deeper in one stack vs. another, and thus may not be allowed to see that the information they need exists no matter how essential it is to their work. So you need to balance access with security to further business objectives.
At the end of the day, your goal must be to enable all the information you have so that it behaves as if it lived in one big box, able to be found by anyone who needs it, when they need it, subject to appropriate controls, in a straightforward manner. It can be done -- but only with careful forethought and a sensible plan, rather than blind faith in the ECM system's capabilities.
ECM helps unite information silos
Improving content accessibility in SharePoint
How to make digital assets easier to find