Data Exchange Ecosystem
scroll
Problem
How do people navigate the world in which sensitive data are collected daily?
For this Instagram study, my team sought to understand where the social media platform fits in the broad ecosystem of organizations that collect data in exchange for personalized products and services. After all, Instagram isn’t the only place where people share data. Too often, tech companies analyze user behavior within the narrow purview of their own interface; yet, myriad on- and off-line touchpoints comprise the data exchange ecosystem.
What are people’s data concerns? Where do people find comfort? And how might tech companies leverage those sources of comfort to address people’s worries about data? These questions motivated our research, whose goal was to generate ideas for new data controls, privacy products, and education resources on Instagram.
Process
We began by developing a research plan that would both broaden our perspective on the data exchange ecosystem and hone our focus on the sources of concern and comfort within it. The Covid-19 pandemic also loomed. Our methods would have to be remote so we could accommodate social distancing.
We arrived at a two-stage qualitative approach.
-
Document a taxonomy of data touch points via remote focus groups using Mural and Zoom. Participants (n=19) discussed where they exchanged personal information
-
Explore people’s data practices via remote diary studies using dScout. Participants (n=30) logged their daily data exchanges over 14 days and shared the information sources where they learned how organizations use consumers’ data.
Research: Focus Groups
Leading remote focus groups in the Covid-19 era posed challenges They had never been done at Instagram. Whereas conversations would normally flow among 10 or more people in a room, we capped the size of groups at four to facilitate discussion in a disembodied setting where people’s social and gestural cues were often invisible.
Four exercises took place using Mural boards. For the first, participants documented all the on-and off-line touchpoints where they exchange data on pre-arranged sticky notes. After performing the exercise individually, everyone shared what they wrote. It was an opportunity to reflect on the vast landscape of data exchanges in our lives.
Second, participants copied their sticky notes and arranged them in the bullseye labeled “satisfaction with products.” The center signified those organizations with the most satisfying products; each concentric ring outward represented less satisfaction. The exercise invited participants to consider what they received in exchange for their data.
Third, participants again copied their sticky notes and arranged them in the bullseye labeled “comfort with data sharing.” The center signified those organizations whose data practice inspired the most comfort; each concentric ring outward represented less comfort. The exercise invited participants to consider what they found concerning about sharing their data.
Finally, participants brought their thoughts together and arranged their sticky notes on a shared board. The x-axis signified product satisfaction; the y-axis signified data comfort. What resulted was a wide-ranging conversation – generating ample attitudinal data – during which people reflected on the tradeoffs they make and the value they derive from sharing data.
Research: Diary Study
Participants recorded brief videos describing where they shared information: Amazon, banks, Google, insurance companies, and more. As for the articles people read, they ranged across CNN, TechCrunch, and USA Today. A common denominator was clear: nearly half (61/132) were discovered via Google search.
In the end, 246 submissions were collected, offering us a trove of data to analyze.
Diary studies are a useful tool to gather longitudinal data that trace people’s evolving attitudes and behaviors over time. We partnered with dScout to recruit 30 participants and to prepare their digital interface. Participants would log on every day and complete a series of tasks:
-
Log two data exchanges each day; record high and low experiences
-
Find and react to articles about organizations’ data practices
-
Find and react to organizations’ data policies
Data Analysis
We began by compiling the 402 data touchpoints mentioned in focus groups. Then, we assigned each a data comfort score by assigning the corresponding bullseye with 1-4 values descending from the center. Averaging each data touchpoints’ score yielded the following rankings of places where people were comfortable exchanging data:
-
Banks
-
Employers
-
Government agencies
-
Doctor’s offices
-
Social media platforms
The rankings showed not only where social media platforms fit in the data exchange ecosystem. It also shined a light on analogous data touch points, which Instagram could learn from for the sake of product development.
For our next step, we analyzed diary study data and transcripts from focus groups and identified the key sources of concern and comfort with data sharing. These were:
Sources of Comfort
-
Privacy laws
-
Organization’s customer service
-
Sensitivity of data categories
-
Organization’s security capacities
-
Avenues of recourse
-
ID authentication
-
Limits around data use cases
Sources of Concern
-
Receiving spam
-
Security breaches
-
Balance of data exchange
-
Data tracking
-
Using data for personalization
-
Showing information to others
-
Social inequity (surveillance)
Pairing the sources of data comfort and concern alongside each other allowed us to strategize how Instagram could relieve people’s worries about sharing data. What new products could be developed on the platform to bring about more data comfort?
We looked at the longitudinal data from the diary study for answers. When comparing participants’ attitudes at the beginning and end of 14 days, what came into view was that trust in data exchanges grew over time. Notably, the number of people who reported being “very concerned” about sharing personal data halved from 10 to five.
Remarkably, only 3/30 people became more concerned about their data after two weeks. By contrast, 8/30 people became less concerned after two weeks. The finding led us to hypothesize that greater familiarity with data practices strengthens people’s confidence in data exchanges.
It turns out that the sources of familiarity are key. Examining the news sources and data policies that diary study participants read over 14 days, we found a marked contrast in responses. The latter inspired more confidence in organizations’ data practices.
The reasons why data policies left people more comfortable than news articles did were not clear. Perhaps news tends to be more incendiary. The neutral tone of data policies could set readers at ease. Regardless of the explanation, the finding pointed us toward new transparency, education, and privacy products to support people’s comfort with data exchanges.
Results
The results offered foundational insights about Instagram’s place in the data exchange ecosystem as well as tactical recommendations for product development opportunities. These included new data controls for people to exercise agency over the categories of information shared with Instagram. There were also educational resources to familiarize people with how the social media platform uses their data to generate personalized content.
That’s all for now! Details about the final products won’t be shared for proprietary reasons.