As an aside, does anyone else feel like folksonomy is kind of a funny word for the ones and zeros crowd? It seems a little too patchwork pants, veggie burritos, and Sam Smith, doesn’t it? Who knows…maybe Thomas Vander Wal is a big Leftover Salmon fan.

Righto….back to tags and performance management.

One of the neato features of Trampoline’s SONAR application was the use of NLP to derive themes (tags) from indexed emails or other corporate content. Put simply, if I create a lot of corporate content about “Shrek” my profile is tagged with “Shrek” (and probably “animation”, “pink-slip”, and “single”). Point being is that this approach gets around the issue of people tagging themselves with “database administration” or “useful” when in reality, they are “java developers” or “lawyers”. (I kid. Really. I love lawyers. Each and every one of the 6 million of you who live in DC.)

I called this phenomenon “alter-ego-projecting” in my previous post, and would guess it is a major drawback of something like Facebook when brought within the walls of a company. Especially a knowledge-worker or professional services firm. Let’s face it. For corporations, the theme-thyself only model of Facebook will fall on its face tomorrow, in the same way competency management did yesterday. On my Facebook profile, my favorite quote is the Helen Keller oldie but goody “life is a daring adventure, or nothing at all.” In reality, I am afraid of heights and fear the worst every time there is in-flight turbulence.

So now take this concept and apply it to performance management…

If you work for a largish organization, it is probably fair to assume that there is some sort of goal-setting and subsequent evaluation process. In an attempt to provide some equality and career growth guidance, the HR folks have divined standard metrics, categories, and other means by which to develop goals, and subsequently evaluate your performance against them. There is a usually a mix of qualitative and quantitative data that provides your superiors input in a resulting HR or managerial action. You could be promoted. You could be counseled. You could be moved to a different part of the organization. You could be given a bonus. You could be rapped upon your knuckles with a ruler. (Maybe that last one only happens to me…)

The qualitative part of this process has always sort of bothered me. In a matrix organization, where you are moved from project to project, and advancement hinges upon a combination of project and functional leadership input, I feel that project reviews and other qualitative input are somewhat lacking in utility. Either people do not have the time to write a review, or an evaluator has little time to read the review. Also, if pressed to present evidence of performance, most people will more than likely only present the most positive reviews. To get around some of these issues, I propose the following performance management system concept:

Develop “performance tag/theme clouds” for people based on indexing and NLP of qualitative reviews. Present this information to evaluators as a means to quickly understand and assess a somewhat large amount of performance related data. Define tag clouds based on corporate content from formal project reviews, self-assessments, and an ad-hoc review email repository. The first two sources are probably self explanatory, but the third…maybe not so much. Simply put, the ad-hoc email repository would be a “place” (like an mailbox) where both the employee and evaluator could send performance related emails. More than likely any negative feedback would be sent to the supervisor anyway, and positive would be sent to both. Forward those notes along to the “performance mailbox” and have the content added to the tag cloud. Also, allow for anyone in the organization to anonymously send performance-related emails to the mailbox. (Whoah…anyone? Yes. Anyone.). Maybe even allow for both employees and evaluators to add manual tags to the emails as they are submitted/reviewed.

As this tag cloud develops, both an employee and a supervisor will see emerging performance themes. If tags like “competent” and “innovative” are assigned a greater font weight, perhaps this employees deserves a raise. Conversely, if tags like “unsatisfied” and “moron” rise to the top, maybe it’s time to polish up the resume. You could even quantify the delta between this year’s tag cloud and last year to show progress over time. Finally, the evaluator can drill down to the specific emails or reviews that produce themes of interest (like moron) to see if there was real substance behind the attribute.

The benefits of this approach is that it gives employees and supervisors a quick and easy way to augment the performance assessment process. Instead of trying to wade through reams of text, both evaluators and employees can see key performance themes and better focus the performance management process. Also, the decision making process is made better by leveraging content from a greater variety of sources.

So what do you think? Is the performance management world ready for the Clifford Ball approach?