Start with the missed morning
An employee has a new health concern before a shift. Getting it assessed could mean finding an appointment, arranging coverage and leaving work. A text conversation may make the first step easier. For a finance team, the next question is more specific: what changed because that access existed?
A helpful benefit and a positive financial return are related questions, but they need different evidence. Count who used the service, what care they needed afterward and which costs the employer actually paid. That gives the benefits committee something firmer than multiplying every conversation by the price of an emergency visit.
Define the covered group before the calculation
For a self-insured employee plan, begin with the employees and dependents included in the agreement. Use the actual quote, enrollment basis and period of coverage. A price per employee and a price per covered person are different inputs; the calculation should use the unit in the contract.
A resident program covers a different population under a separate local arrangement. An employee can live outside city limits, and a resident can work for someone else. A city’s employee benefit does not automatically give every household access to Tap.
This distinction also determines the evidence. Employee-plan analysis uses relevant claims and benefit costs. A resident-program review may also examine EMS operations, access and household effects. Keep those groups visible when presenting the results.
Understand what a Tap encounter can do
A member starts with a text. Tap confirms access, gathers the concern and relevant history, and connects the conversation with its clinical team of physicians, physician assistants and nurse practitioners. A photo, phone call or video conversation can help when appropriate.
The next step depends on the assessment. It may include advice, clinically appropriate prescribing, testing, follow-up or a recommendation for in-person care. A prescription is never guaranteed. Some problems need an examination or emergency services that a text encounter cannot supply.
That means a referral is not automatically a failed visit, and a finished text thread is not automatically an avoided claim. The analysis should recognize appropriate escalation as part of good care.
Measure changes you can defend
Agree on a baseline period and comparable post-launch periods with the plan administrator. Specify which visits belong in the analysis, how later care is connected to an initial encounter, and how enrollment changes will be handled. A claims decrease alone does not identify its cause.
Then compare plausible scenarios with actual results. Use the employer’s allowed costs rather than a hospital’s sticker price, and allow for people who would have waited, used primary care or sought no care at all. Those encounters should not all receive the same assumed savings.
For EMS, a separate hypothetical calculation could use $1,200 as an assumed avoidable cost per deployment. That is an editable example, not a verified national cost. Fixed staffing and equipment expenses may remain even when a unit handles fewer calls. Capacity released and cash saved should be described separately.
Use other communities to ask better questions
Tap reports a 20% decrease in non-emergent EMS runs in Ferris, Texas. That is a community-specific reported result, not an employee-plan return or a forecast for another city. Differences in population, participation, call definitions and the measurement period matter.
The Chesterfield case study offers another starting point, with reported community estimates and an explanation of their limits. Its EMS, emergency-department, wage and school measures benefit different groups. Adding them together does not turn them into money returned to the city treasury.
Use the interactive community calculator to make assumptions visible, then replace the defaults with local evidence. A resident-wide program still requires a community decision and confirmed terms. The strongest proposal explains what will be measured, who will review it and when the estimate will be revisited.
