Scripts, cheat cards, and quick answers for the front desk and office manager. The Seven Steps for every inbound call, and what to say when a patient reacts to the AI voice-assist. Print the cheat sheet, keep it next to your phone, and use the huddle drills to warm the team up.
Before you answer their question, get their name and callback number. Every call. It's the single biggest lift on conversion — and the easiest step to skip when the phone is stacked up.
These are the full source pages. Bookmark them — this playbook is the summary; those pages are the whole thing.
Greeting → connect → take control → uncover chief concern → schedule → gather info → wrap up. The framework Dana wrote for every front-office team member.
10 patient scenarios: explaining why the AI answered, defusing when someone's frustrated, and recovering when the AI got a detail wrong.
Tap any step to open the full detail on the source page. Click Print in the corner to get a desk-side version.
Filter by what's happening on the call. Tap "Do / Don't" for the coaching notes underneath.
Left column reads as cold, defensive, or blame-shifting. Right column keeps the practice warm and in control of the story.
| Say this instead of… | Preferred wording |
|---|---|
| "our robot" · "the bot" · "the AI" | "our virtual phone assistant" · "our after-hours assistant" |
| "the system got that wrong" | "let me get that fixed for you right now" |
| "the AI can't do that" | "that's something I handle here at the desk — let me take care of it" |
| "it won't happen again" | "you've got me now, and I'll make sure this call is a great one" |
| "we're short-staffed" | "every line was busy and I didn't want you on hold" |
| "you must have misheard" | "let me pull up exactly what we have on file and we'll fix it together" |
| "that's just how the technology works" | "the reason we use it is so no one reaches voicemail — but you're with a real person now" |
Front desk handles most reactions. Anything past mild frustration goes up the ladder — same-day, not "next week."
Mild frustration, mild surprise, correcting a wrong detail. Use the defuse / recover scripts above and finish the call yourself.
Patient asks to opt out of the assist, wants privacy details, wants a specific promise about the tech, or is choosing not to schedule because of it.
Patient says they were misled, mentions a complaint or review, or wants to speak above the OM.
Any mention of legal action, HIPAA, recording law, or public exposure (media, social). Same-day route — no defense of the tech in the meantime.
Pick one, read the setup, and role-play with a team member. Rotate through them week over week — everyone hears every scenario.
Role-play Steps 4–5. One person plays a patient with a toothache. The other has to uncover the concern, confirm insurance intent, then offer two specific appointment times.
Play patient B-1 ("I hate talking to a robot"). Team member has to defuse without arguing the AI is "actually pretty good."
Play patient C-3 (told on the phone you take Delta, now told you don't). Team member has to apologize, pivot to accurate options, and offer the membership plan.
Team member picks up a mock call. Patient asks "does your office take Aetna?" — team member has to redirect to Step 3 (get the name and callback number) before answering.
Read the "Say this instead of…" table out loud together. Each team member picks one phrase from the left column they've said before, and commits to the right-column swap this week.
Play patient who mentions the word "complaint" or "review." Team member has to take the callback number, not defend, and route to OM the same day.
Tap "Show answer" after you've thought about it. If you missed one, open the source script and read that section.
What phone situation do you run into that isn't covered here yet? Dana + Jessica review incoming topics monthly and add the ones that show up most.