How Does a Data Analyst and a Business Anthropologist Dance Together? Part-1
Right at this moment, we are deluging ourselves in the world of Data. Data is everywhere. 1 Exabyte (1⁰¹⁸) of data is being created on the internet daily, equals to almost data in 250 million DVDs. Astonishingly enough, today we produce the same amount of data that it took from the dawn of civilization until 2003 to generate. Internet of Things has already become a reality and as more and more physical objects become connected to the internet, we will enter the Brontobyte (1⁰²⁷) Era. (WIRED.com)
We can readily observe website data — location, time, devices, internet providers, even age group and potential interests — with just a click in Google Analytics for free.
We can study more about social media data where we can find about social network interactions, pages visited, pages liked and shared (You can find a lot more on Saqiful Alam’s blog).
Digital marketers are now switching to programmatic marketing where data is analyzed from customers to advertise the ads to the right people at the right time.
We know now more about health data: patient’s description, potential symptoms, future risks and sometimes which can be linked into other forms of medicare services like Google Glass.
People data, where data is gathered in organizations to measure employee behavior data in workplace behavior. These types of data are analyzed to measure employee retention, performance evaluation, hiring and firing decisions. Even Google‘s Aristotle project has predicted with their people data about which employees might leave the company soon with a dominating accuracy.
There are also other measures of data such as texts, videos and images all of which can be analyzed and interpreted.
But what do we do by analyzing a certain dataset? We visualize the pattern emerging and try to ‘make sense’ of it. This visualization helps us in our everyday decisions. To be less wrong.
But what are the shortcomings of just quantifying numbers or analyzing them with only quantitative methods? Number ‘3’ does only mean 3. In a certain dataset, number 3 might mean a position in a marathon race. It might mean the 3rd year when a labrador was adopted. It might mean the number of surgeries a patient went to. How can we put the number in a context?
Moreover, how can we effectively put the number into such a context, that we can infer the highest intelligence possible from the analysis?
Let’s see how an anthropologist — a business anthropologist — might observe a context.
The Power of Proper Contextualization:
Big Data has limited value if not paired with its younger and more intelligent sibling, context. However, the layers of context that explain who stated the data, what type of data is it, when and where it was stated, what else was going on in the world when this data was stated, is just over the surface.
Clearly, data and knowledge are not the same thing.
Contextualizing Beer Consumption Patterns:
One of the most popular articles in putting contextualization into place is An Anthropologist walks into a Bar… from Christian Madsjberg and Mikel B. Rasmussen from ReD Associates. The problem they researched on was of a beer company who could analyze a trend of puzzling negative correlations with their core product with falling bar and pub sales, but a positive correlation with store sales. Something was missing. The pattern from the data was not enough to conclude an inference.
The Beer company commissioned a team of social anthropologists to visit a dozen bars in the UK and Finland to find out. The anthropologists immersed themselves in the life of the bars and started observing the actions, attitudes and the behavior of owners, staff, and regulars beyond the data without any hypothesis about what they might find.
In time, patterns emerged. Although the data showed how to use the promotional materials — coasters, stickers, T-shirts, and so on — the anthropologists found out how these materials were used, was actually unproductive, and at worst thrown into the cupboard. The team also figured out that the female servers felt trapped in their jobs where they had to be flirtatious with unknown people, which resulted in resenting their jobs and motivation. These female servers actually knew so little about the products themselves that they were unable to influence a buyer’s decision about a good experience with their product. They were just selling the product as just any other beer.
Finding the ‘why’ into the data with the help of anthropologists, lead the company to significantly change their strategy from one-size-fits-all promotional materials to a more contextualized customer experience for different types of bars and bar owners. They developed a design to help its salespeople understand each bar owner better and invented a tool to help owners organize sales campaigns. By understanding the context better, they were able to train waitstaff about its brands and won over female servers by providing taxi service for employees who worked late. The beer company’s pub and bar sales rebounded, and both sales and market share continue to grow.
Contextualizing SceneTap:
Another keynote example of contextualization comes from a recent San Francisco app called “SceneTap.” First mentioned in “Advancing Ethnography in Corporate Environments” by Chad R. Maxwell, he mentions about how Chicago bars were using the visual pattern recognition technology with this app. Smartphone users can readily connect to the bar scene data by in-bar cameras with facial detection analytics. The app posts information such as the average age of a crowd and the ratio of men to women, therefore giving out a pattern to help the barhoppers decide on where to go. A very qualitative and quantitative data, amorphized into a real time pattern for decision making abilities. However, it received a thrashing from San Francisco as the app made people feel spied on by the technology. Specifically, women.
SceneTap looks at a variety of characteristics to determine gender and age: the nose, the eyes, the jaw structure, mouth and overall face shape, forehead and skeletal structure. “It almost takes your face and creates a grid, matching general facial features to males or females, before determining how old you are,” [CEO Cole] Harper explains. “In a certain sense, it’s trying to find your look-alike in an anonymous database.”
Although the idea might sound awesome, people from San Francisco denounced the idea. An epicenter for technology and startups, San Fran technoligsts are the ones who are actually concerned more about privacy than others.
The one who keeps saying that you shouldn’t ever do anything on the internet you would ever be embarrassed about? Where is he from? Chances are higher that he is from menlo park!
San Francisco People boycotted the idea, and SceneTap’s PR were knee-deep in comments in various websites trying to defend itself from perceptions of women feeling like ‘hunted prey’ or ‘game animals.’ Issues of gender profiling skyrocketed. Later on, the board of directors restructured their company’s services to merge with BarVision (A liquor monitoring company), which continues till today.
Contextualizing ‘Market Research’:
The power of contextualization is so hidden sometimes, that, business, moreover, which is multinational and operates from a global perspective fails to comprehend the minor technicalities that can either result in a huge gain or a huge loss.
A key example is shared by Julia Gluesing, a professor from Wayne State University. She was called for to help address pressing problems in the product development, sales and service aspects of a computer manufacturing business.The comany had offices primarily around France and in the United States, and designated English as the common language. There were about thirty teams, from which only six were performing according to the company’s expectation. Although each team had undergone weeklong orientation program and learn about their mission and how to work together as a team, they were consistently being reported as being behind schedule in their tasks and difficulties in getting along. They all had process facilitators to help get their work done, but it was just not making the level of progress as expected.
The data pattern was only able to show the numbers. However, inferring a conclusion from the data, was problematic. A business manager from the top might view it as a process or structural problem that might arise from how the french or the US team organized their work ethic. However, Julia Gluesing, once employed as an anthropologist, conducted an ethnographic research project where she was assigned to determine what was contributing to the success or failure of the global teams.
Julia participated in the team sessions; both face-to-face and video calls. She got to know the team members informally over lunch, dinner and other activities. She invested time to listen to their stories about work in general, about their teams, about their lives, where all counts as ‘data’ and contributes to the research. She interviewed the managers outside the teams and tried to understand the organizational contexts that influenced how the team members saw their work behaviors.
Julia’s ethnography resulted in a typology of characteristics of successful and unsuccessful global teams and uncovered many understandings within the teams and as well as the teams and managers that were previously poorly understood. One simple but major problem that cost a lot of money for the company turned out to be the many different definitions for commonly used English terms, as innocuous as “market research.” Team members used the term assuming that they all had the same understanding of what market research meant. This seemingly deceptive, but inoculated thought made a huge discrepency in the planning and execution processes in different branches. In France, market research meant the opinions of tech experts and top managers inside or outside the company. However, in the United States, market research was different; it was conducted by talking almost exclusively with customers and potential customers, even those who were customers of competitors. Because of different meanings, one team was unable to compare market research data gather in France with that gathered in the United States.
As a business anthropologist, with her insight, she was able to adjust the training and facilitation to include more emphasis on language and on understanding the clarigication across contexts and cultures. Managers also decided to rotate the quarterly face-to-face global team meetings across locations to give all members the opportunity to learn about each other’s work contexts and standard work practices to have a shared understanding about their goals and creating a composite international team culture over time.
Contextualizing an Online Course on Internet:
A online course is certainly one of the best takeaways from Internet of Things. However, as I was looking through one of the popular courses from Wharton, I wanted to extract the data of people registered for the course, as the field itself is comparatively new. The majority of the participants are from the North America and Europe. Given the depth of data, a fellow data analyst offered me these possible inferences:
- Countries from the northern part of the world are mostly accustomed with taking online courses. They are more adaptive on online course cultures than other countries. (which I don’t believe as I have seen so many people from so many countries on other courses)
- They have more money to be able to afford these online courses. (maybe)
- They know about these courses because they need them. People from the south are not implementing these technologies, therefore the logic is a two plus two. (maybe)
Can it be one of them? Or can it be all of them? How would we know?
As my data analyst friend and I started working together — which we brand as ethnographic analytics, thanks to Chad Maxwell! — the people from some of the southern countries, for example in India, Bangladesh and Vietnam gave us some interesting answers.
Some told us that the first analysis they started with was the examples of NFL draft. A lot of people taking these courses haven’t ever have seen what NFL is, and don’t know the rules of the game. So the contextualization is missing.
As one of the first key example from the Week 1 of the course was about NFL drafting, a lot of people had failed to comprehend it, the analogy experience turned out to be puzzling and they dropped out within Week 1. Would they be more interested if the examples were based on football or cricket? Majority said yes. But will the Professor from Wharton create an example based on the customer needs? We don’t know. Moreover, this is not just a contextualization problem in terms of online content creation. It is also a contextualization problem in terms of research. How many NFL drafts research are there compared to Cricket in the academia?
Therefore, the problem is not so simple. You can see the data, observe the data, but to be actually be able to feel the inference from it, is tough.
The Takeaway:
Take the example of a company that has invested heavily in business intelligence (BI) software that organizes internal data. In an increasingly connected world, this company has not leveraged its data to its potential. Why not? Because the company’s internal data is isolated from the rest of the data universe including news, social media, blogs, and other relevant sources. What if you could join the dots between all these data sources and surface hidden connections?
For organizations and businesses to survive today, they have to contextualize their data. And to contextualize, a data analyst and an anthropologist needs to dance together. Just as a doctor diagnosing a patient with diabetes based on body temperature data alone is incorrect, so is making business decisions derived from data out of context. A doctor needs to know about the patient’s age, lifestyle, diet, weight, family history, and more in order to make a probable and guarded diagnosis and prognosis. Contextualization is crucial in transforming senseless data into real information — information that can be used as actionable insights that enable intelligent corporate decision-making.
At the end of the day, our overworked minds want to be spoon-fed insights. We want key signals and tightly packaged summaries of relevant, intriguing information that will arm us with knowledge and augment our intelligence.
The Don of ubiquitous computing, Mark Weiser, stated in 1991 that “the most profound technologies are those that disappear. They weave themselves into the fabric of our everyday life until they are indistinguishable from it,” much like the telephone or electricity. They have seeped into our surroundings, playing an integral role in our everyday lives.
Then, how do we extract real intelligence from data? When a data analyst and a business anthropologist can dance together.
