Social performance management : is Frankenstein coming to HR ?


Summary : we know that yearly performance reviews have limits and that it’s important to show employees recognition more often. It’s also obvious that a large part of what employees do for their employer, both qualitative and quantitative, is overlooked because of lack of tangible indicators. That’s the promise of “social performance management” : using, in the flow, customers and colleagues feedbacks about someone’s work to enrich the review. But, not even mentioning such systems may face legal issues in many European countries, we may wonder if, because of their nature and their logical imperfections due to their youthfulness, they’re not a wrong good idea. That’s a very interesting matter for HR anyway because, even if such tools represent an actual risk, it’s impossible to overlook the questions they raise about how performance is reviewed.

What’s social performance management ? To make it short, let’s say that it’s about using what’s done and said on a social network to improve performance reviews. Let’s start with a few words on my beliefs regarding this matter.

– people should be appraised according to their mission, not to the tool they use to accomplish it.

– what people do on social platforms way be helpful to assess the qualitative side of their work (helpfulness, receptiveness…) but that’s a very sensitive matter that need to be clarified with all stakeholders, especially unions etc. in some countries.

Before, in organizations equipped with social networks or similar tools, such things were carried out by guesswork. Managers used to rely on what they saw, on their impressions and do what they wanted with such informations. But the good point, compared to the past, is that they had tangible informations to explain their decisions (“you did not help this colleague while you had the solution”…). In some ways it was “less worse” than before because the informations that could be found on these platforms made a part of the appraisal a little bit less subjective.

Now we’re being offer to automate a part of this job with specialized software. What do these softwares do ? In the current state of the art they don’t try to find if you’re trying to help others or not, if you try to solve problems or not. They just have a look at how people comment and rate your own participation, if they “like” what you do or say. Example : Rypple and Salesforce.

What is it going to change ? It will be possible to widen the scope of the appraisal, take more elements into account that used to be left aside before. So, the one who’s helpful, the one who’s appreciated will be rewarded. Or, at least, his manager will have a quantified view of this part of the appraisal…and will be free to make the most of it…ir not. This mechanics are also very interesting because in service and support activities they’ll also make it possible to track customer’s and partner’s feelings about an employee, provided something was said on platform that’s tracked.

We all know that the traditional yearly review is becoming less and less relevant. This concept of a 360° review based on actual work interactions makes it possible not only to have in indicator that updates in the flow but also to take into accounts things that were very hard to seen and track. So we can expect that employees who are really engaged will get the recognition and reward they deserve. And even more.

Yes but…

Dealing with such matters is like walking on eggshells. Of course there is some risk, as there’s risk in anything. This is not the point. What’s important is the nature of the risk. When it comes to performance reviews, imperfections in the systems harm the global balace of the group, are a threat to trust. And from a pragmatic standpoint, behind the human risk lies the legal one…which is a major one.

Let’s be clear. Performance review systems must be know and accepted by everybody and their impartiality guaranteed. In other words, any error is prohibited. Better not implement a new system (even if promising and potentially powerful) than implement time bomb that instead of rewarding those who deserve the most will make the groupe explode.

As of today, such tools rely on two mechanics to do their job.

The first is the abillity, for any user, to give badges, props, likes or I don’t know what to any other. Of course no one can cheat, no arrangement will happen between employees and there’s no chance to see the search of effectiveness turn into a popularity contest. No…no chance to see such things happening. It won’t even come to people’s minds to do so…

The second is identifying “positive” and “negative” words. A kind of “sentiment analysis”. I’m a strong believer in sentiment analysis but there’s a drawback. It it’s used to monitor how people feel about any event, the way they perceive a product etc… we can accept a margin of error. Possible misunderstandings on slight differences or irony is not harmful in this case : when the analysis relies on hundreds of thousands occurrences, the global trend will be reliable.

But when it’s about measuring someone relying on only ten or less occurrences, the right to make mistakes is nonexistent. Do you thinks that things like : “great job !”, “could have done better” could be misunderstood ? Without information on the context, humans can misunderstand such words. Knowing the context, some misunderstandings will stay. So, when the one in charge of understanding what people say to and about others is a machine, in the current state of technology I prefer to pass.

Of course some will say that the data collected by the system will have only a little weight in the global appraisal, that managers will be cautious, that… But the germ of potential problems would be introduced in the system.

Some may also thing that it will be easy to game the system and cheat. But it seems that these systems have built-in functionalities to prevent such issues. Wait and see.

In fact this is one of the limits of bringing what works on the public web in the workplace and of early adoption.

In the other hand that’s a topic HR departments should watch carefully. Not to prevent it (in fact from a legal standpoint I’m very curious to see how it will work in most of European countries…) but because, like it  or not, performance measurement methods need improvement. More relevance, more qualitative elements, more recognition and instant feedback… In this context, social performance management systems may be game changers…once they are more mature. Meanwhile, it’s high time to think about reinventing the current model, maybe with more elbow grease than technology.

Anyway, one day the technology will be reliable and no one will be able to afford overlooking the performance measurement concern. But technology will be useless until organizations have changed their mindset. So better work on changing practices first instead of playing the sorcerer’s apprentices with technology.

Conclusion is we should beware of the “new is beautiful” attitude and, because it’s new, promising, fun and social, throw oneself one what does not look more than the Frankenstein of HR systems. Except if HR people decide to turn themselves in sorcerer’s apprentice…but I don’t think they will.