How to Run a 90-Day EVV Pilot Before Rolling Out Agency-Wide

Switching EVV tools for an entire caregiver roster on day one is how agencies end up stuck with a tool that doesn’t fit: either because a technical issue that only shows up at scale wasn’t caught early, or because caregiver resistance builds faster than anyone can address it once it’s agency-wide. A short, structured pilot de-risks the decision before it’s a full commitment, and it costs almost nothing beyond the time to run it deliberately instead of rushing it.

Why a Pilot, Not a Big-Bang Rollout

The instinct to just flip the switch agency-wide is understandable. It’s faster, and it avoids running two systems in parallel. But it also means any problem you discover, you discover everywhere at once, with your entire caregiver roster and your entire client base exposed to it simultaneously. A geocoding issue with one client’s address, a caregiver whose phone doesn’t support a required permission, a gap in how the sync handles a specific service code: any of these is a minor fix caught in a 10-caregiver pilot and a full-blown operational problem caught in a 200-caregiver rollout. The pilot isn’t just caution for its own sake; it’s how you find the failure modes while they’re still cheap.

Days 1–15: Small-Group Rollout

Start with five to ten caregivers, ideally a mix of tenures and comfort levels: a couple of your most tech-comfortable staff and a couple who are more hesitant about new software. Resist the temptation to pilot only with your strongest performers; if the app only gets tested by people who’d adapt to anything, you won’t learn what actually breaks for the caregiver who struggles with technology, and that caregiver still has to use it eventually. This surfaces real usability problems while the group is small enough to fix issues individually rather than through a mass communication: a client address that doesn’t geocode cleanly, a caregiver whose phone has location services disabled, or a caregiver working across two different waiver programs whose task list needs to reflect that.

It also helps to pick caregivers who serve a genuine cross-section of your client base rather than clustering the pilot around one easy neighborhood or one waiver type. If your agency serves PCA, CFSS, CADI, BI, and EW clients, try to have at least one pilot caregiver touching each program, since documentation requirements and task authorization can differ meaningfully between them.

What to track: clock-in success rate, average time-to-clock-in, and how many exceptions (missed clock-outs, missing GPS) show up per caregiver. Write these numbers down somewhere you’ll actually look at again; a shared spreadsheet is fine. The goal at this stage isn’t a polished dashboard, it’s a baseline you can compare against as the pilot expands.

Days 15–45: Confirm the HHAeXchange Sync

This is the phase where you verify the technical promise actually holds: visits logged in the new app should be appearing correctly in your HHAeXchange account, with the right service codes and timestamps, without your billing team doing anything differently. This is arguably the most important checkpoint in the entire pilot, because it’s the one that’s hardest to catch informally: a caregiver-facing usability problem gets noticed the moment it happens, but a sync issue can sit quietly in the background for weeks before it surfaces as a denied claim.

What to check: pull a sample of synced visits (not just one or two; aim for a few dozen spread across different caregivers and days) and compare them against what caregivers actually logged in the app. Confirm your billers aren’t seeing anything unusual in claim submission, and specifically ask them whether the visits are landing the way they’d expect from a manual entry, with no extra cleanup step quietly added to their workflow. If your billing team starts doing anything differently to make the new app’s data usable, that’s a sign the integration isn’t as clean as advertised, and it’s much better to catch that with ten caregivers’ worth of visits than two hundred.

It’s also worth confirming what happens with the caregivers you didn’t onboard for the pilot but who might cover a shift for a pilot client. Make sure a substitute caregiver clocking in on an unfamiliar app doesn’t create a gap in the record.

Days 45–75: Expand and Stress-Test

Widen the pilot to a larger group, roughly 25–40% of your caregiver roster, including caregivers serving different waiver types (PCA, CFSS, CADI, BI, EW) if your agency spans more than one. This is where problems that only show up at volume tend to appear: a spike in GPS exceptions concentrated in one part of your service area, a scheduling edge case (overnight shifts, split shifts, multiple clients per day) that the small pilot group didn’t happen to include, or a caregiver segment (newer hires, for instance) who need a different onboarding approach than the group you started with.

What to check: exception rate at scale, caregiver feedback on the clock-in experience, and whether your scheduling and billing staff are noticing any change in their own workload. It’s easy to focus entirely on the caregiver-facing experience during a pilot and forget to check in with the office staff who touch the data on the other end. Their feedback at this stage is just as important as the caregivers’.

This is also the point to test what happens when something goes wrong on purpose: have a caregiver simulate a missed clock-out, or intentionally clock in from outside the expected radius, and confirm the exception actually gets flagged and routed the way you’d expect. It’s much better to discover a gap in the exception-handling process during a controlled test than during a real audit.

Days 75–90: Decide

By day 90, you should have enough data to answer three questions plainly: Is the exception rate low enough to trust at full scale? Are caregivers actually using it without resistance, not just tolerating it, but not generating a steady stream of complaints or workarounds? Is the HHAeXchange sync clean enough that your billing team isn’t doing manual cleanup?

If yes to all three, roll out to the rest of your roster, using what you learned about onboarding pace and communication from the pilot group to plan the wider rollout. If not, you’ve spent 90 days finding that out with a fraction of your agency, not all of it. Whatever the sticking point turned out to be, you now have a specific, well-documented problem to solve rather than a vague sense that “the new app isn’t working,” spread across your entire caregiver base.

Common Pilot Mistakes

A few patterns show up often enough to call out directly. Piloting only with caregivers who are already comfortable with technology gives a falsely clean result: the group that struggles most is exactly the group you need real data on before a full rollout. Skipping the HHAeXchange verification step because “the vendor said it works” is the single most common source of a rollout that looks fine for weeks and then produces a wave of denied claims all at once. And treating the 90 days as a fixed script rather than a floor, moving straight from a clean 15-day check to a full rollout because things seem to be going well, throws away the exact stress-testing value the middle phase of the pilot exists to provide.

Why This Matters More for a Free Tool

Because there’s no cost to piloting a free EVV app, there’s no reason to skip this process to “save money” on a trial period. The usual argument for rushing a rollout, that a slow pilot delays the return on a paid tool, simply doesn’t apply here. The only cost of a slow rollout is time. The cost of a fast, bad rollout is caregiver trust, which is much harder to rebuild than a 90-day timeline is to run. A caregiver who has one bad first experience with a new app tends to carry that skepticism into every future software change your agency makes, long after the original bug gets fixed.