Ask why recall never happens and you’ll usually hear that staff didn’t prioritize it. That explanation survives because it sounds like something a manager can fix.
Look at what the front desk is actually holding. MGMA Stat asked 343 practice leaders about their biggest challenges on 19 August 2025. Margin and costs led at 41%.
Access and growth came next at 26%, then staffing at 20%, with medical assistants the hardest role to fill. Picture that desk. Short-staffed, absorbing the day’s calls. It will answer a ringing phone before it dials someone who is not asking for anything.
That is the whole mechanism. Recall by phone competes against triage every day and loses every day.
Watch one export prove it. A manager pulls 300 lapsed patients for Midtown in week one. Staff reaches 25, most of them voicemail. Week two brings a handful.
By week three, the sheet is under a stack of prior-auth paperwork. Next quarter, a fresh export prints the same 300 names, plus the new ones.
The recall call list never happens because nothing in the day protects it. Staff time on outbound recall calls is the first thing given up when volume spikes. Volume spikes daily. Automating overdue visit outreach fixes the structure instead of asking people to try harder against it.
Manual recall runs at roughly 5% execution across a full list, and it decays fast. Week one is hopeful. Week two is interrupted.
Week three has no owner, because whoever volunteered is covering the desk for someone who called out. The list does not fail loudly. It just stops moving.
Every manual attempt costs a scheduler several minutes. Pull the chart, dial, wait, leave a message, note the attempt. Roughly half reach voicemail.
A voicemail from a strange number in the middle of a workday gets deleted. The cost is certain, and the return is a coin flip. Busy desks decline that trade.
Across five sites, the problem multiplies rather than averages out. Each location gets its own export. Each works it to a different depth.
By quarter end, nobody can say how many lapsed patients were reached. Per-location recall reporting does not exist, because the work was never steady enough to measure.
Reactivation is the highest-return outreach the group can run, and it stays stuck at a fraction of execution. Measure the gap honestly.
It sits between running the campaign at all and never running it, which is a wider gap than any coaching conversation can close.
Setup happens once. Segments come from DrChrono visit history. Templates get written and approved at group level. Cadence sets how often a patient can be contacted.
Throttles set how many messages leave each site per week. After that the campaign runs on schedule, and the only repeat human task is booking the people who reply.
Throttling is the setting that decides whether a campaign helps. Match sends to open slots rather than to list size:
|
Location |
Open slots per week |
Recall sends per week |
Replies at 35% |
|---|---|---|---|
|
Midtown |
45 |
60 |
21 |
|
Westside |
30 |
40 |
14 |
|
Northgate |
18 |
25 |
9 |
Figures are illustrative. Watch the ratio. A site sends what it can absorb, so the campaign runs longer instead of flooding one Tuesday morning.
Interested replies arrive in the shared inbox with the patient’s context attached. A scheduler at any site can offer times and book. Nobody switches systems mid-conversation, and no separate campaign tool needs watching.
Segmented Recall Campaigns give each site its own sender identity, active segments, language mix, and throttle rate. One group-level template library governs all of them.
A dermatology recall and a primary care recall read differently on purpose. Neither gets written by whoever had time that week.
Per-location dashboards report sends, response rate, and booked appointments for each campaign. Recall stops being an activity people half-remember doing. It becomes a funnel with a conversion rate. That is what makes it defensible in a budget meeting.
One condition worth naming: campaign throttling and capacity only work if the open-slot numbers are current. Say a site changed provider hours last month and never updated its throttle.
It will either underuse the campaign or overshoot its schedule. The fix is a five-minute review, not a support ticket.
Reconversion on recall texts runs at 35% across our client base. Apply that to a 300-name segment. Staff effort moves from 300 outbound attempts to about 100 booking conversations with patients who already said yes. Every minute now goes to the productive end of the funnel.
Sends, replies, bookings, per location, per month. A site with a strong response rate and weak booking conversion has a scheduling bottleneck, not an outreach problem.
A site with a weak response rate usually has stale mobile numbers. Both diagnoses come from the same report, and neither was available when recall lived on paper.
Automation does not get tired at week three. Every site’s lapsed panel gets contacted in full, monthly. Segments refill themselves as new patients cross the lapse threshold.
Schedulers spend the day booking rather than dialing, which is the job most of them would rather have.
The call-down list fails structurally, not personally. Ask any team to make outbound calls in the gaps of an inbound day, and you get the same result. Blaming the team for it wastes the one quarter you had to fix the workflow.
Ask each location manager what happened to the last recall export. Those answers are the business case, and you can probably predict them.
Schedule a consultation and bring your open-slot counts by site. We’ll design segments and throttling against each site’s real booking capacity. Then we’ll set the reporting that tells you next month whether it worked.