Posts from SXSW Interactive 2014 – day 2

What: Workplace distractions: A new focus on focus
Who: Ben Waber (Sociometrics), Janet Pogue (architecture), Gloria Mark (informatics), Rachel Silver (moderator)

We are all pretty aware of the problems that task switching and context switching can cause. It’s one of the identified Lean “wastes” of knowledge work, and we who go to war on this know all too well how hard it is to get to the bottom of the problem. In this panel, we heard from people involved with understanding the effects of – and possible remedies to – work flow interruptions.

Gloria Mark kicked the whole thing off with an interesting talk, where she pointed out that we are monochronics in a polychromic world. Gloria’s research demonstrates that people spend on average three minutes in a focused state, on one separate task, before switching or being interrupted. Going up one level in the work break down structure, she notes that people on average switch between separate projects every 10,5 minutes. She also showed that the come back time, as in the time it takes to get back to the productive state on the original task after the interruption, is on average 23 minutes and 15 seconds. In Lean we would call that relearning, a waste for sure.

We also know that people get stressed by these interruptions, and surprisingly Gloria claims that people with many interruptions actually work faster when they do get their focus time, than people in an interruption free environment. And here I must add that in knowledge work, faster is rarely better. Quality of work and quality of thought matters. This is an important aspect of this topic, and I felt it deserved some attention. Or at least a mention. There is a fundamental difference between output and outcome. Productivity is a high risk metric just waiting to be gamed.

Interestingly, people are just as likely to self interrupt as being externally interrupted. Gloria wonders if we might be conditioned to be interrupted, but leaves the question unanswered. I was a tad surprised that no one in the panel brought up Mihaly Csikszentmihalyi (Flow: The Psychology of Optimal Experience) and his ideas at this point. I think that could have added another layer, or depth, rather, to the discussion. And I think Csikszentmihalyi proves quite well in his work that no, we are not conditioned towards interruption. Under the “right” circumstances we don’t self interrupt at all.

Next, Pogue talked about how the physical workplace impacts performance. Apparently, Pogue and team designed the new Facebook HQ, which looks a lot like any given Google or IDEO. I have always felt that the office design of the high tech super novas feels more like branding than anything else. I apologize for this blatant display of cynicism.

And surprisingly, the Hawthorne effect was never mentioned.

Next up Waber, who pointed out that we have become quite apt at using behavioral data to study and understand our customers, but we haven’t until just recently started to use the same behavioral data to look at work. Waber and team have what they call a sociometric badge, which collects data on various aspects of the work day. It can track who talks to who, speed of speech, changes in tone of voice and it can measure stress.

Waber taked about a case study: A company where workers are paid based on throughput. So if you are interrupted you are paid less. So what happens when people have a problem? In their research Waber and team can see who the workers turn to for answers. He demonstrated that there were four people (in a rather large company) that everyone turned to with their questions. Interestingly, these experts are statistically average when measured on throughput. But they make others look good and help people out, at their own detriment it seems. Great for the company, but bad for the experts. So when this became clear, the company started to incentivize these connections and rewarded that behavior with some sort of bonus. It is of course a good thing that people who add value are rewarded properly, yet here there were a few aspects that I found lacking.

First, why didn’t they experiment with the way people are compensated? Is being paid based on throughput really the best way to go?

Second: I’d venture a guess here and propose that the experts helped their team players for very good reasons. Because it feels good helping people out. And they felt important and valued. We might call this meeting people’s needs, and we might call it intrinsic motivation. I wonder if adding extrinsic motivation in the shape of bonuses to the mix risk doing more harm than good?

Third: These informal experts constitute a risk for the organisation. I would have paid them a high base salary for being available for the other team players. Answering questions, helping them out, coaching and teaching. This is knowledge sharing. It’s a good thing. It’s the way out of the experts trap, and I like to refer to this as staff liquidity.

Fourth: If you as a leader need sociometrics to figure out that people are stressed you’re really not doing your job right. The whole concept of sociometrics feels a bit like Tayloristic time and motion study, and that creeps me out a bit.

And then I saw a lot of other stuff, drank lot’s of beer and then it rained a lot.  It was a good day in Austin.



Leave a comment

Your email address will not be published. Required fields are marked *