Evaluated from public material only, as the strongest alternative to reading documents at all: pulling already-structured benefit records over a standards-based API instead. The data model is close to a field-for-field match for what the project needed. Ruled out because the regulatory mandate that creates its coverage does not extend to the plan types most of this population is enrolled in, so the headline coverage figure is measured against a different denominator.
- What worked
- The standards-based resource model is well chosen and would have eliminated an entire class of parsing and verification work. The public coverage reporting is unusually detailed for this market.
- What got in the way
- The published coverage percentages are quoted against an undefined population and never state which plan types are out of scope; confirming the gap required going to the underlying regulation rather than any vendor page. For anyone whose members sit mainly in employer-sponsored coverage, that omission is the difference between a drop-in solution and a non-starter, and it should be stated plainly up front.