Real-time hypothesis exploration
Learn how to check and validate any hypothesis about the factors that are influencing your customer's behavior and business KPIs.
Outcomes
- Quickly find the driving factors of any KPI you’re trying to optimize
- Automatically self-serve insights using approaches that have been tried and tested by data scientists
Part 1: Building your dataset
Adding customer features to a dataset
Using aggregations to do quick exploration
- Using column shortcuts to create group by tabs [more info]
Adding activity features to a dataset
- How to add a features from an activity [more info]
📝 Practice: Start with a KPI (conversion rate or average) and segment (break it down by a customer attribute)
- Understand the impact on your KPI in aggregate
- Understand the impact on your KPI over time
Part 2: Generating your Analysis
Using the Analyze button to do real-time hypothesis exploration [more info]
- Specifying the inputs
- What is time between?
Understanding the analysis [more info]
- Understanding the baseline
- Assessing the impact of your features
Tracking the influence over time [more info]
- Scheduling analysis to be re-run
📝 Practice: Use the same KPI/customer column as before to generate a full analysis
- Does that customer feature matter?
- What action would you take from this analysis?
Part 3: Asking Good Questions
Tips for coming up with good hypotheses
- Starting with KPIs
- Starting from Customer journey
- Starting from another Narrative
Updated over 2 years ago