Data-Driven Jobs-to-be-Done: How a Star Schema Can Help You Succeed
Unlock the potential of your Jobs-to-be-Done catalog with a star schema model that empowers you to make informed decisions based on real data
Some of you may be wondering how all of the data that we collect during our qualitative research gets used. I’d like to use an analogy to help understand. If you’ve done any work in data analysis, you’ve probably heard of a Star Schema. This looks like a spoke and wheel set up, where the hub is a FACT table, and the spokes are DIMENSION data.
This is more applicable to the actual data model that comes out of a Jobs-to-be-Done survey. However, it’s also a good way of thinking about what you will find in the Jobs-to-be-Done catalogs I share.
Think of the FACT TABLE as the Core Job Success Metrics. And think of Situations, Contexts, Use Cases, etc. as DIMENSIONS. We’re rating the needs in the Core Job but we’re using the other data to slice and dice them to see what changes given the various dimensions or combinations of dimensions.
There will also be other things you may want to add to a survey that I’m not cataloging for you. For example, if you’re studying the purchase journey for a physical product, you may want to ask questions about where a purchase was researched, where it was initiated, and how it was received (delivered, or picked up somewhere) to understand channel contexts and capabilities around right-channeling your customers (the balance between what they desire and what you can offer). That would also be a set of dimensional data you can use to understand segments, or the ranking of success metrics.
It’s important to think about the questions you need to answer before you begin your research and certainly before you construct your survey. Each question may require you to not only have specific dimension data, but have it in sufficient quantities to be statistically relevant. This is not like filling out a four forces diagram. Jobs-to-be-Done in its most effective form is a data-driven method and bad data - or assumptions - will result in bad outcomes.