
Integrated Data Thinking™


Integrated Data Thinking™ is a framework developed by Sudden Compass® that allows teams to utilize both qualitative and quantitative data towards answering mission-critical optimization and discovery-level questions. It ensures that data analysis stays grounded in the question at hand, and creates a simple, common visual language that can align stakeholders from across teams, functions, and silos, regardless of their level of data literacy.
This simple approach offers a powerful intervention to several all-too-common patterns that prevent teams and organizations from achieving breakthrough results, such as over-reliance upon quantitative data, utilizing optimization-level techniques to answer discovery-level questions, litigating over which methods to use, and analysis paralysis. Integrated Data Thinking™ works by creating and operationalizing a shared understanding of the differences between discovery and optimization-level, and qualitative and quantitative.
Integrated Data Thinking™ is useful both as a method for planning pre-research, and as an assessment tool for evaluating data and research initiatives across an organization. Mapping multiple questions and data-related initiatives against the Integrated Data Thinking™ framework can help an organization better understand its gaps and needs around data-related initiatives, ultimately resulting in a more balanced and thoughtful approach to learning about customers.
THE COMPONENTS OF THE INTEGRATED DATA THINKING™ FRAMEWORK
- Unknown (discovery) questions are about variables that are emergent or unclear, where we cannot retrieve an answer from a known data set, i.e., “What is the future of mobility?”, “Where are the new markets for self-driving cars?” and, “What are the new products that our customers will want?”
- Known (optimization) questions are about variables that are already clear, well-understood, and could likely be answered via an existing data set, i.e., “Which feature is most popular amongst our customers?”, “How do we get our mid-tier customers to upgrade to the next tier?”
- Thick data (qualitative) questions cannot be answered by a number, and often involve a dimension of “why” or “how,” i.e., “Why are our sales declining?” or, “How do our customers shop for financial services?” The data points for qualitative data are presented in the form of stories through photos, videos, quotes, texts, and drawings.
- Big data (quantitative) questions can be adequately answered by a number, percentage, or ranking of a set of options, i.e., “How many of our customers are using this specific feature?” or, “Which of our pricing tiers is driving the most revenue growth?” The data points for qualitative data are presented in the form of statistics and numbers through charts, spreadsheets, and graphs.
Integrated Data Thinking™ is utilized to bridge the first (Ask) and second (Analyze) steps of Sudden Compass’s Unlock Sprints™ practice (Ask, Acquire, Analyze, Act). By starting with a question—not a data set or a methodology—Integrated Data Thinking™ ensures that the data acquired and analyzed in subsequent steps is actually well-suited to answer the question at the heart of the sprint.
Directions
- Create or print a 2x2 grid, such as the one pictured above, available in Google Slides format here. The x-axis should be labeled “UNKNOWN (discovery)” on the left, and “KNOWN (optimization)” on the right. The y-axis should be labeled “THICK DATA (qualitative)” on the top, and “BIG DATA (quantitative)” on the bottom.
- Identify a question that is timely and relevant to your team or organization. This could be a pre-defined business or research question, or a human question that emerges from using the Business Question to Human Question method. Write this question on a post-it note.
- Decide where your question falls on the spectrum from unknown to known, and qualitative to qualitative. Place it in the appropriate quadrant of the 2x2 grid.
- Use the resulting placement to determine the appropriate research approach. For example:
Qualitative/Discovery: Ethnography
Qualitative/Optimization: Targeted interviews, focus groups
Quantitative/Discovery: Exploratory data analysis
Quantitative/Optimization: A/B testing - Repeat the above process for other timely and relevant questions. Doing so can help identify areas where critical questions are not currently being asked. (For example, many organizations over-rely on questions that fall into the quantitative optimization quadrant.)