Every one of the 169 workflows carries two scores out of 100: opportunity, built from 8 inputs, and implementability, built from 6. Priority is their mean, and the threshold dividing the four quadrants is 60.
The two axes and their inputs
Opportunity asks how much is at stake in a workflow: how often it runs, what it costs to staff, what it loses when it goes wrong, and how far the damage travels. Implementability asks whether the work is tractable: how standardised it is, how good the data behind it is, whether the rules can be written down, and whether anyone would pay to have it done.
Each input is scored 1 to 5 for every workflow, then the axis is rescaled to 100. The weights are equal within an axis. That is a choice, not a finding, which is why the sensitivity test below matters more than the ranking does.
| Axis | Inputs |
|---|---|
| Opportunity | Workflow frequency, Function labor cost, Practice revenue loss, Inside-out error cascade, External error impact, Human skill required, Payer variation, Platform variation |
| Implementability | Implementation complexity, Data maturity, Process standardization, AI competition, Knowledge codifiability, Customer willingness to pay |
The sensitivity test
Every weight was jittered by a quarter and the ranking recomputed 5,000 times. A workflow's top-three stability is the fraction of those runs in which it stayed in the top three, and top-five stability the same for the top five. Both are published for every workflow on the ranking table, including the ones where they are low.
A stability of 0 does not mean the workflow scored badly. It means its position is an artefact of the exact weights and should not be quoted as a rank.
What this method cannot tell you
Edition 2026.1-analyst is scored by analysts. No operator panel exists, so no agency has independently rated a single workflow in it. Where the scores are wrong, they are wrong in the direction analysts are wrong: toward what is legible in the regulations, away from what is only visible from inside a branch office.
The two axes measure how worthwhile a workflow is to automate and how hard it would be. They do not measure whether any particular product does it, and no product was tested to produce them.
Disclosure
Moative builds AI systems for home health and hospice agencies, and published this atlas. Two of the inputs in the original scoring measured how well a workflow fits Moative's own tools. Both were removed before any score on this site was computed, and every number here was recomputed without them.
The scoring code, the inputs and the full ranking are published so that a reader can check that claim rather than take it.
Licence and citation
The atlas is published under CC BY 4.0. Reuse it, including commercially, with attribution.
Suggested citation: Moative Research. Home Health AI Atlas, edition 2026.1-analyst. 2026-09-22. https://moative.health/home-health-hospice/