What if I told you that there was a magical integration piece that could pull together information management, collaboration, big data analytics and knowledge management — would you believe me?
I read a book recently which made me think about the people, processes and technology it would take. And that magical integration piece's name? Search.
Except of course it's not "just search." There is a lot to enterprise search, and for this discussion we'll include the related disciplines of text/content analytics. The book that got me thinking about this was Martin White's excellent "Enterprise Search: Enhancing Business Performance."
The Basic Theory
Businesses keep adding new repositories and systems in an attempt to manage the ever increasing amounts of unprocessed data, processed information and processed knowledge. Search technologies will only grow in importance as a result to assist in finding the content needed to get work done.
The value created from a powerful and easy to implement search strategy will of course depend on the context of the industry, the size and type of organization and whether or not the workforce is made up predominantly of knowledge workers or process workers. Note the term "search strategy," not search engine or search application in the previous sentence. Chances are, your organization probably has multiple repositories and search tools already — and despite what the vendors might say, there really is no one technology product that will solve all your problems.
Let's investigate a search strategy through the lens of people, process and technology, but in reverse order.
Search engine technology overlaps considerably with what's used in text analytics (a.k.a. content analytics). According to Wikipedia, text analytics is "roughly equivalent" to text mining, defined as: "the process of deriving high-quality information from text. High-quality information is typically derived through the devising of patterns and trends through means such as statistical pattern learning."
White takes this definition further in his book to include concepts such as computational linguistics, natural language processing and entity extraction, which are used to identify relationships and trends within the corpus of text. By deriving relationships and trends, text analytics plays a similar role in delivering actionable insights from unstructured information as the complex business intelligence analytics tools do with big data.
Think about this in terms of using collaboration platforms to enable knowledge management. Imagine running a text analytics engine against all the text in blog entries, micro-blogs, news feed posts, etc and the potential intelligence you might gain from this new "meta-information."
When developing your more conventional search strategy, consider the number of repositories you have and whether each has search capabilities and indexes built in. Consider this when deciding whether a centralized enterprise search platform, which crawls and indexes all your content sources, or a federated search will meet your needs.
Federated search passes queries to existing search engines, which run the query against their indexes and send the results back through a single interface. A warning though: White cautions about the potential of confusing recall, precision and relevance of results when pulling results from multiple sources that may calculate these values in different ways.
That last point brings us to the business processes which underpin your information management efforts. Federated search results may be much easier to manage if all of your repositories have consistent, standardized metadata. But in a world where it's easier to invest in a shiny new technology than put in the effort to create metadata schemas, taxonomies, ontologies, and the standards, policy and procedures that go with them, content is often lacking in the attributes of "findability."
Again, this where search strategy comes in, as the element that can pull various technologies and business processes together. A large organization may have considerable strengths in these areas in some business units, and considerable deficiencies in others.
Technology can help enable our business processes. So to return to our social collaboration for KM example or KM for big data example: imagine a single search portal that returns pertinent results from your CRM, ERP, social platform, document management system, data lake, data warehouses, etc. Some of you may be lucky enough to already have this in place.
One common theme in White's writing over the years is that when it comes to search, most organizations underfund and undervalue the human aspect. Businesses pay good money for new search platforms, but then fail to invest in the people who run it, manage the presentation of the results, tweak the functionality, etc.
White suggests a set of roles required for a high performing search team in his book, so I won't repeat them here. But understand that the investment in people doesn't stop with a search team. Do your CIOs, CKOs or CDOs have the necessary people under them? Even worse, and seemingly far more prevalent, have you invested in the organizational learning needed for a KM strategy, and on the tactical level, in training end users?
You may provide compliance training because regulators say you have to, but how much information management training do you provide your employees? Do they have the skills and training to use all the technology tools they have access to? Rather than handing out a policy document, have they been trained what to put in their OneDrive, what to put in the hundreds or thousands of SharePoint Team Sites, what to discuss in the social collaboration tool, what metadata to use in the DMS, or what constitutes a "record"?
Employee learning and development often lands pretty low in business's priority lists. And — surprise — lack of training will negatively impact the use of sophisticated search tools.
So no, there are no technology silver bullets. But an enterprise search strategy — which includes the people, process and technology lenses and text analytics — can go far to support and enhance knowledge management and in turn, business value.