Why Are Clinics Still Waiting for Patients to Call?

Years ago I crossed Australia with a circus. Actual circus. Trucks, dust, small towns, a different field every few nights.

I was convinced the aerial silk act would be the draw. It was the hardest thing anyone in that company did, months of conditioning, genuinely beautiful, the sort of act you'd put on the poster.

The crowds went to the fire chains.

Two guys spinning burning chains in a paddock, maybe six weeks of practice between them, and that's where people stood three deep. Not because it was better. Because you could see what was happening from anywhere in the field, at night, holding a beer, with a kid on your shoulders. Accessible beat impressive. Every town. Every time.

I think about that constantly now, watching clinics buy AI.

Why are clinics still waiting for patients to call?

Because the phone ringing is a problem you can see, and silence isn't.

Every practice knows its inbound call volume. Somebody complains about the hold times in every staff meeting. So when clinics start automating, they automate the incoming call, which is reasonable, and which most of the industry has now built for.

But the patients costing you the most money never call.

The referral sitting open for nine weeks. The post-op patient who's fine, probably, who knows. The 62 year old who's due for a colonoscopy and has decided not to think about it. The balance nobody chased. None of them generate a ringing phone, so none of them generate a complaint, so none of them get automated.

They just quietly don't happen.

What does proactive outreach actually recover?

Money and appointments that were already yours. That's the unglamorous part.

The numbers coming out of proactive AI outreach campaigns are getting hard to ignore. Practices are reporting 60%+ conversion on appointment scheduling campaigns and up to 90% on rescheduling. One ENT group hit 64% referral conversion on referrals that had been sitting dead. An orthopedic practice recovered outstanding balances from 47% of patients within a week of outreach. When a clinic closed for weather, more than half the affected appointments got rebooked without a single staff member picking up a phone.

None of that is new revenue. It's revenue that already existed and was leaking out through a hole nobody was measuring.

That's the fire chains. Not the most sophisticated thing AI can do in a clinic. Just the thing that works from anywhere in the field.

Why doesn't a reminder text do the same job?

Because a reminder tells a patient to do something, and doing something is where they fall off.

"Your appointment is Tuesday at 2, reply C to confirm" works fine for the patient who was always going to show. For everyone else it's a task added to a life that's already full. Reschedule means call the office. Call the office means during business hours. During business hours means when you're at work.

So it doesn't get done.

A real conversation closes the loop inside the same interaction. The patient says Tuesday doesn't work, and Thursday at 10 gets booked right there, and it's in the EHR before they hang up. No callback, no portal, no "we'll get back to you." That's the difference between a nudge and a resolution, and it's why the big platforms are all converging on voice as the primary outbound channel rather than a fallback.

We see the same thing. Around 85% weekly engagement on voice follow-up versus the 15 to 20% baseline you get from digital-only outreach, across five countries and three languages. The gap isn't the script. It's that a phone call doesn't require the patient to become an organized person first.

What should a clinic automate first?

Whatever is bleeding the most and requires the least clinical judgment. In that order.

For most practices that's a short list. Confirmations. No-show rebooking. Open referrals. Recall for overdue screenings. Outstanding balances. All high volume, all repetitive, all currently done badly because a human being with better things to do is squeezing them in between other tasks.

Start there. Not with triage. Not with anything that requires the AI to make a call about acuity on day one. You want the boring workflow that runs 400 times a month and fails 40% of the time, because that's where the compounding is. We map the usual starting points by specialty and workflow, and the honest answer is that the first thing you automate should be embarrassingly simple.

My daughter's 10. She told me once, mid-argument, that I work for her and not the other way around. She was right, and it applies to software too. The tool works for the front desk. If the front desk has to reorganize their day around it, you bought the aerial silk.

How do you know it's working?

Pick the number before you start. This is where most pilots die.

Not "patient satisfaction." Not "efficiency." A number somebody currently owns and can recite from memory: no-show rate, referral conversion, days to third next available, percentage of balances collected inside 30 days. One number, baseline captured before go-live, checked at week six.

If it hasn't moved by week six, the workflow was wrong, and that's fine, you change the workflow. What you can't do is run a pilot with no baseline and then argue about vibes in the renewal meeting.

The economics on this stuff are usually not subtle when the workflow is right. Clinics running our voice follow-up see roughly 31:1 return, and that's mostly recovered appointments and caught complications rather than headcount savings. We publish how the pricing actually works because the math should be checkable by the person signing, not just the person selling.

Key Takeaways

The AI conversation in healthcare has been stuck on the inbound phone call for two years, and inbound was never where most of the money was going. The patients who cost a practice the most are the ones who never call: the open referral, the no-show, the overdue screening, the unpaid balance. Those workflows are high volume, low complexity, and currently handled by whoever has a spare ten minutes, which means they're mostly not handled. Proactive conversational outreach closes them at rates that legacy reminder software has never come close to, and it does it without asking the patient to download, log in, or remember anything. Start with the boring workflow. Pick one number. Check it in six weeks. The accessible act draws the crowd.

FAQ

What's the difference between appointment reminders and proactive AI outreach?

A reminder notifies. Proactive outreach completes the task inside the conversation, so the patient reschedules, books the referral, or pays the balance without a callback, a portal login, or a second touchpoint.

Will patients actually talk to an AI on the phone?

Consistently, yes, when the call is short, relevant, and about something they already know they're behind on. Engagement drops fast when calls feel like marketing and stays high when they feel like the clinic keeping track of them.

How long before a clinic sees results from outbound automation?

Usually inside six weeks for the simple workflows, because rebooking and referral conversion move immediately. Revenue effects from recall and screening campaigns take a quarter or so to show up properly in the numbers.

If you want to look at which of your workflows is leaking hardest, book a slot and we'll go through it.