Segments and metrics for outputs

Segments and metrics are the components that make up the outputs of our data analysis.

Segments: Overall and by major segment (product, geo, demo, etc)

Critical segments to figure out how to consistently separate: - How does the customer behave in the funnel/business process? - What pattern do they follow, according to the metrics we have identified as critical to the success of the business from our 108.20.50.10 Data Analysis - Acquitention Heuristic and discussions with the company?

• Easy segments are those with columns in the data.
  • Product, geography, etc - but these will rarely be interesting or important other than as support knowledge.
• Pareto segments
  • Show the 80/20, and the 96/4, and the 99/1. How do these segments differ from each other in terms of LTV, CAC, funnel, activity patterns, etc?
• RFM segments
• 108.30.10.20.20.10 Data Analysis - Segments - RFM
  • And RFM by segment - i.e. do an RFM segmentation within other segments which seem important so that we can identify the champions and losers there

Metrics: Revenue and engagement (customer count, LTV, interactions, etc)

Metrics can be different for different kinds of companies. E.g. A non-subscription consumer company will care about AOV because basket size is important, but this won’t change much for a subscription SaaS company so they will not care about it.

Example

To get the total number of outputs, take the number of segments and multiply by number of metrics. E.g. - Overall segment - Product segments (5 products) - Geo segments (10 geos)

(times)

• Gross retention
• Quick Ratio
• Net churn
• Revenue metric
• Customer count metric
• LTV metric
• AOV metric

= total of 16 segments * 7 metrics = 112 outputs, each with numeric grid and graph

Side note re. RFM segmentation

It would be a very fun and good experiment to include an RFM segmentation into the above outputs. It would add ~10 new segments so the number of outputs would be 26 segments * 7 metrics = 182 outputs.