“In epistemology, and often in common use, the prefix meta- is used to mean about (its own category). For example, metadata are data about data (who has produced them, when, what format the data are in and so on). In a database, metadata are also data about data stored in a data dictionary and describe information (data) about database tables such as the table name, table owner, details about columns, – essentially describing the table.” – Wikipedia
How meta is your data? I could add another layer of abstraction and ask, how meta is your metadata? It may seem like a strange question, but I ask in the spirit of getting “meta”: that is, “meta” in the contemporary usage of the term as an act of a discipline reflecting on itself. In this case, I suggest your organization gets self-reflective and meta about its content.
When you look around your organization, does your data (including structured data and all other business content in un- and semi-structured formats) seem indicative of your business, culture, and goals? I would posit the level of organization and self-awareness inherent in an organization’s content is a good measure and reflection of the clarity of purpose in the organization. When the content repositories are a mess, the data is inconsistent, and findability is a chore, chances are the organization has not made the effort to get meta.
Getting meta involves knowing your company inside and out, including those ugly dust bunnies of content hiding under the couch.
From a content perspective, knowing your company starts with a knowledge audit. You could also call this a content audit, information audit, or data audit, but I prefer knowledge audit because it includes a broader range of information. After all, this isn’t just about recording what data is in a database, this is also about knowing the purpose of the data, its location, how it is used, who uses it, and its provenance.
Knowledge audits often start with an information management technology project, such as installing a content management system or updating an Intranet. What initially starts off as a project with a limited scope quickly spirals into a host of sub-activities, including knowing what information input will go into the tool, what information will be created in the tool, and what information output is produced. To do the project well, the organization must conduct a knowledge audit to avoid information duplication or ignoring important business information in other systems.
Part of getting meta is viewing the content through an organizational lens. What is the data? Where did it come from? Was it produced internally or externally? What system or systems does it live in? What is the data used for? By whom? In what processes? How is it stored and how long is it retained? These questions are just a start for a full knowledge audit.
But, you may say, that sounds like a lot of work. Sigh. Yes, it is a lot of work, but I’ll return to my hypothesis about clarity of vision. Meandering, ever-changing, and unfocused organizational goals may take some time to surface, but they will. When they do, the results can be very painful, very expensive, and, in the worst case scenario, could actually bring the business to an end. The time spent auditing and reflecting on organizational knowledge is worth it.
Where do you start? The best leads for content profiling will come from people rather than attempting to conduct a blind and impartial audit directly in unknown technology systems. Identify the users directly involved in the creation or use of data and they will happily tell you about what’s important and what’s not. They will even more happily tell you their grievances about the system of record and the quality of the data. Listen to them. What may sound like a lot of griping is a tell-tale sign there’s an opportunity to organize information more effectively and get more mileage out of it.
Employees know the content best. While each person may have a limited view of the organizational content, assembling the interviews will give a holistic data portrait. Not only that, the picture will include details such as information duplication, process bottlenecks, security risks, and repetitive, inefficient processes.
Speaking of processes, I’ve never met one I liked. There’s no process which can’t be improved in some way, and knowledge audits reveal the pain points.
During your interviews, be sure to ask about the details in information processes. You will likely find inefficiencies, duplicated efforts, and a patchwork of data leaks and misuse. Finding opportunities to improve information processes is another factor in getting meta. If you were performing a task in your personal life that was inefficient and unproductive, wouldn’t you want to find a better way? A knowledge audit is your chance to expose inefficiencies and fix them.
When content is no good, there’s a real tendency to blame the technology. While it’s true the system may have been designed poorly or have other flaws, get meta. Why is the system so bad? Was it always bad or has your organization outgrown it? Is what is bad fixable? Is it time to bite the budgetary bullet and go through the effort of finding a replacement?
Remember, the technology giving you the headache could potentially be the technology that could solve other issues exposed by the knowledge audit. Similarly, the flow of knowledge may be bottlenecked due to technical constraints which could be fixed.
No, Seriously…Get Meta
The results of the knowledge audit will tell you just how much you know yourself as an organization. You’ll find content with unknown origins. You’ll find content with unknown importance. You’ll find content so old you can’t bring yourself to trash but don’t know what to do with if you retain.
A serious analysis of this content allows you to get meta. Once you know the current state lay of the information landscape, you can assess whether the content reflects the purpose and mission of the organization. If it does, you may already have good information processes in existence. Chances are, your content will look like a disorganized mess. What do you do?
Getting meta gets literal. Do a deep dive into the content itself using text analytics tools to perform information extraction. For example, take search logs (from multiple search engines or tools, if available) and perform term clustering techniques on them to group similar concepts into larger categories. You’ll be amazed at what people are searching for. Again, this is getting meta. What are users searching for, and, more importantly, does that content exist and are they finding it?
You could also perform a function-by-function or process-by-process information extraction exercise on content to identify concepts and themes which could be then managed in a taxonomy management system and applied back to content as metadata. This is really getting meta: look at your content to find what topics are important and apply those topics back to categorize content so it can be found, routed, and accounted for within the organization’s goal activities.
Probably the most important aspect of getting meta is to make it visible and reinforce it wherever possible. Self-reflection serves no purpose if it’s locked in a diary and no actions are taken on the insights.
The meta metadata should be seen in Intranet navigation, in search facets, in content management system tagging, and anywhere else that concepts reinforce the organizational culture and mission.
Now go get meta!