Andreas Weigend | Social Data Revolution | Fall 2014
School of Information | University of California at Berkeley | INFO 290A-03


Audio: weigend_ischool2014_5.mp3 (dropbox)
Transcript: weigend_ischool2014_5.docx


How Can We Better Understand Employees?

People analytics is both an old and new phenomenon. It allows researchers to understand the behaviors that make people effective, happy, healthy, creative, more successful. It is becoming more evident that using data to build strong organizations can be better than relying on instinct. Companies using data behind business decisions perform 5% better than other companies (People Analytics).
Before we relied on data such as demographics, experience, education, test scores and peer evaluations to understand employees. Now, through people analytics we have begun to study face to face communication and quantify it. Doing so can provide valuable information on relationship/ group dynamics that one can’t garner from individual’s data.

How Can We Measure How People Interact

Over time, there has been a shift from qualitative to more quantitative forms of interaction measurement. This began with mere observations of the world. Surveys, technology, and quantitative strategies were then developed for measuring human interaction and determining what makes people happy. Now, hard data is everywhere and much easier to gather and measure. Detailed data on behavior can be collected from email records, web browsing behavior, IMing, and all other internet technology. This reveals personal information, such as an individual’s daily activity and the range of networks he or she has interacted with on the web. Data on the physical world is also expanding rapidly due to wearable sensing technology (data on interaction and speaking patterns, motion, location, etc.). We can now understand behavior on a huge scale with this combination of data from both real and virtual worlds.
Wearable sensing technology allows us to gather data that is objective, continuous and at a high resolution. Because the data gathered is quantified we can compare it and formulate ideas based off of these comparisons. Wearable technology allows one to measure:
  • who do you meet throughout the day? in what context? for how long? how often? was it scheduled?
  • how did the two interact: was it chit-chat, informal, sit down?
  • who talks more? who interrupts whom?
  • what is the conversation dynamic: frequent back and forth, question and answer?
  • how are they physically presented? who is leaning in or out? how much do he people move? who are they facing?
  • what is their speech pattern? what do they emphasize? is there variation? what is the speed? what is the tone? volume? intonation?

What Drives Performance?

By studying mathematical models Sociometrics have found that the key performance indicator is cohesion. Cohesion studies two nodes that one person is connected to and the probability that those two nodes are connected. The higher the probability, the more cohesion exists within that environment which translates into a closer knit community within the workplace. A tight knit workplace community has been linked to higher happiness levels in staff which leads to better performance. This is due to to the fact that because there is a strong community, workers can find people to vent to or help solve their problems. Many interactions between workers involve “job training” where one gives advice or general information on a problem.

How Can We Encourage Higher Performance?

We can encourage higher performance in a number of ways, the first can be a change in the break structure. Rather than making people take breaks at different times, if you enforce the same break time for everyone people will be more likely to mingle and this can lead to a stronger group dynamic. You can also change the spatial structure of the office or break room. For example, Sociometrics found that if you add an espresso machine to the break room (which let’s say is a 500 dollar investment) this could increase cohesion by 3% which leads to a 2.5% increase in sales. They also discovered that replacing 4 seat tables with 12 seat tables led to an increase in both cohesion and performance.

We can also encourage higher performance by changing the reward structure. For example rather than focusing on individual performance it might be more beneficial to judge performance based on teams. For example one person might not be a stellar worker but they are knowledgeable and help others work better. Feedback systems that are real time and more organized can also encourage higher performance. By making feedback systems real time workers can compare their data and study it over time. Organizing this data in different ways, like for example comparing store to store or division to division can also be very helpful.

How Is The Balance Of Power Shifting?

Data-gathering technology has changed over time so that data collection has become a win-win situation; the participant uses a device that is beneficial in one or multiple ways (and perhaps aesthetically pleasing, easy to transport, not too bulky as to become a nuisance, etc.) and researchers gain informative salient data in return. Oftentimes, the participant may be oblivious to such data collection. However, in the past, incentives to participate in studies were sparse, unless individuals had a keen interest in advancing data research and knowledge.

Is Anonymization Possible? Is It Important?

Big data = big brother? It is crucial that the fine line between data collection and invasion of privacy is not crossed. Gathering data can pose a major threat to privacy—whether from cell phones, web browsing histories, etc. There is a high potential abuse of data yet low awareness of this issue because individuals often don’t realize the sheer quantity and power of the data they create and make accessible. In the workplace, it is legal to watch employees with CCTV, log keystrokes, take screenshots of employees using company computers, and read employee emails. In order to maximize employee happiness, productivity, and success, there must be trust and transparency within organizations between employees and employers.

What is the Future of People Analytics?

In past years, information technology has had a large impact on work atmosphere and organizational design. Information technology has allowed for companies to acknowledge new innovative ways to work and collaborate in the work environment. For example, global collaborations of professionals or organizations are made possible with technology like the Internet, email, telephony, real-time video, etc. In the future, people analytics can create a community of organizations all over the world by exchanging data; data exchange can occur between monster corporations and local small town shops alike. Technology will continue to advance and thus continue to shape the work environment and organization, hopefully in a manner that will increase both happiness and productivity of individuals.


Questions for the Class?

If you had all the data in the workplace how could you improve productivity and happiness?

I'd want to allow people the freedom to bring up concerns -- either anonymously or openly. I also believe that recording such data demands that people remain accountable for their actions, which ultimately leads to a more healthy workplace environment. Any problems should be directly and appropriately addressed and resolved. Publicizing this open nature of resolving concerns also holds employers accountable to take care of their employees.

Timeline Oct 28, 2014 (Part 2)
4:40 BEGIN
5:30 END

Guest: Taemie Kim

This page initially created by (2 students total):
  1. Lizbeth Ochoa (
  2. Saige Wexler (