Who is set up
The first fact is the footprint. data_sources_connected counts how many data sources the customer has connected. In the walkthrough, the 10:04 row shows 4 and the 09:12 row shows 1. Both are in the service.
Zero sources. Zero data sources means the customer is not in the service yet. A count of one is already a different state from a count of zero.
Are they still using it
checks_run is the volume. last_check_at is the recency. The 10:04 row has 38 checks, and the last one is at 16:20. The recording marks that row as still using the service.
The 09:12 row has 5 checks, and the last one is at 09:41. The recording marks that row as stopped after first usage. A first check can be the only check. Customer success watches whether use continued.
Are they getting the outcome
warnings_resolved counts findings the customer fixed. warnings_ignored counts findings left in place. The row with 4 data sources resolved 6 and ignored none. The row with 1 data source resolved none and ignored 3.
Read both counts. The recording keeps the two counts side by side. Six resolved and none ignored describes a different customer from none resolved and three ignored. It does not collapse the pair into a health score.
Succeeding in the service
The walkthrough calls a customer succeeding when all three facts are true: data sources connected, still using the service, and fixing what the checks find. The 10:04 row meets that bar. It has 4 data sources connected, a last check at 16:20, and 6 warnings resolved with none ignored.
The 09:12 row does not. It has 1 data source connected, checks that stopped at 09:41, and 3 warnings ignored. The useful output is a list of who is succeeding and who needs help, with the three facts that explain each row.
Episode 7 uses the same facts with the comparison reversed.