What if your Fitbit Could Tell You When you are Burning out at Work?

We are all being asked to collaborate more, and that is a good thing isn’t it? Well apparently not, if it leads to collaborative overload, according to Babson College Professor of Global Business Rob Cross. In his 2016 HBR article on “Collaborative Overload”, he and his co-authors found that “Up to a third of value-added collaborations come from only 3% to 5% of employees.” In a follow-up article on “Collaboration without Burnout”, he identifies that “Much overload is driven by your desire to maintain a reputation as helpful”. In our own research work we find that this is a common trait for those we identify as “Influencers or key players” in enterprise social networks like Microsoft Yammer or Workplace by Facebook, or team collaboration platforms like Slack or Microsoft Teams. But how does one know when they are over-collaborating and therefore risking burnout?

A growing market is now emerging around workplace wellness, with the association of a healthy work-life balance and performance. Apps like Adrian Medhurst’s Benny Button look to scientifically assess staff wellness or ‘capacity’ that drives performance. Medhurst identifies burnout happening when high performance is being sustained at the expense of wellness. Wellness assessment however typically requires survey and assessment interventions. But what if we can come up with a way to provide early warning signals for workplace overload in real time?

Assessing Collaborative Overload in Real Time

At SWOOP we monitor digital interactions in close to real time, with an intent to provide guidance on how to collaborate more effectively. To date though, we have not considered the issue of collaborative overload.

We looked at data from a single organisation of over 6,000 actively interacting staff over an extended period of some 18 months. As a proxy for Wellness/Capacity we chose the ‘energy’ determined from the sentiment contained in messages posted. Our ‘energy’ measure is simply the strength of the sentiment expressed, whether positive or negative. We have noticed that the sentiment analysis algorithms can be overly harsh on the often short and sharp message exchanges in smaller familiar groups like teams. It was therefore safer to ignore the sentiment sign and just record the strength of sentiment, in what we are calling ‘energy’.

We also added the size of an individual’s network of staff whom they have developed an online connection with. Having a large network provides the individual with a larger capacity to influence, through their interactions.

We chose to use our diversity and reciprocity metrics as the independent performance measure, drawn from network science research. Diversity in this case, measures the breadth of exposure to multiple online groups and reciprocity is the proportion of connections that are reciprocated. Plotting Capacity against Performance we found:

We have simply used the average scores to create the 2x2 matrix. In the top right quadrant we can see staff that have built large networks that they engage with energetically and are rewarded with higher performance. In the bottom right quadrant we have staff who are matching this performance, but with less capacity i.e. potentially coasting, yet still performing. This begs the question as to whether those in the far top right are collaboratively overloaded? Are these the 3 – 5% identified by Rob Cross et al.?  

For those in the top left quadrant the staff are not being rewarded for their collaboration efforts. Does the lack of engagement from their efforts make them a flight risk? Or can some coaching on how to better direct their collaboration efforts move them to the higher performing right? The remaining quadrant identifies those that are potentially passive and under-performing. These are the staff that would benefit most from collaboration improvement advice e.g. SWOOP provides real time nudges and improvement hints to individuals exhibiting poor online collaboration behaviours. 

 A Fitbit for Collaborative Overload?  

 To better understand how real-time advice might impact on collaboratively overloaded staff, we chose one staff member from the far top right of the above plot to look at more closely on a month to month basis. 

Over the 18 month period this staff member had established a network of 525 members and posted 896 messages on the enterprise social network (ESN). Typically, messages form about 25% of all interactions (messages, likes, @ mentions and notifications) so the total interactions on the network over this period would be approaching 4,000, or roughly 10 interactions per working day. Given that the ESN is just one channel for interactions and doesn’t include email, chat, meetings, face to face, it’s likely that this individual is a  candidate for collaborative overload. 

The top plot below shows the capacity and performance scores over time. The bottom chart shows the differentials between capacity and performance on a month by month basis.The bars above the axis show the relative ‘excess’ of capacity over performance, or if you like ‘unrewarded capacity’. Below the axis is the more agreeable excess of performance over capacity, or ‘rewarded capacity’. 

 A number of insights can be drawn from the above graphic: 

• The first point to notice on the top graph is that capability and performance are broadly correlated (0.56) and in this case increasing.  
• The individual’s capacity is more volatile than their assessed performance. Perhaps this is simply reflective of the varying energy cycles we all experience throughout our work/life journeys. 
• This variability in capacity means that there will be periods where we feel we are working hard i.e. energetically engaging and interacting, but not appearing to be gaining the anticipated performance reward. Can this lack of reward be a sign of burnout? If it remains this way would this person become a flight risk? 
• A the other end of the scale we can see, fortunately in this case, many more periods where capacity is being rewarded, even when energy levels and network interactions have dropped off. Perhaps this is a delayed reward for networking efforts made in the previous period?  
• The cyclical nature of Capacity/Performance reward ratio might suggest that we all go through cycles where we feel our collaboration efforts are going unrewarded, and other times where we may seem to be cruising.  

While this research is a somewhat preliminary effort to address the critical and growing issue of collaborative overload, looking at interaction data from what we are actually doing in real time, can be a powerful ally for addressing potential overload. For example, much of the wise counsel that Cross makes in avoiding burnout requires a critical assessment of your network and the real value that your interactions are providing or receiving from others. However, when you are under the gun in an unrewarded capacity period, it is hard to step back and make the required reflections. If, however, you have visibility of the previous periods of overload, you can confidently predict that a new cycle of reward/capacity surplus is on the horizon, providing the space to reflect on your future network interactions. By implementing ‘burnout prevention’ strategies you can reduce the volatility in your capacity/performance work cycles, while increasing your performance into the future. 

At SWOOP we are looking forward to being able to provide these insights as part of our MSTeams product initially. We feel that the personal tab on your SWOOP for Teams product is the safe place to go to reflect on your own networking performance. 

We welcome your comments as well as any other ideas you would welcome having on your personal networking 'Fitbit'.