§ 01 Abstract
Two ways in. One clear next step.
Objective.Preserve the speed of a familiar clinic phone menu while giving callers room to explain a request in their own words.
Approach.Clinic feedback led Skrypt to retain keypad input alongside open speech. Both entry points lead into the clinic’s configured answering, booking and handoff workflows.
Results.Among the observed first-input choices, 52% of callers spoke and 48% used the keypad. The near-even split supports retaining both routes as clinics introduce AI: a familiar shortcut for those who want it, and open conversation for those who prefer to explain.
§ 02 Background
The shortcut matters.
A regular patient knows the route: press 1 for appointments. A new patient has a scheduling question that does not fit the menu. Both are trying to reach the same clinic.
When we first discussed replacing phone menus with open conversation, clinic teams challenged the premise. Some callers wanted to explain the situation; others wanted the familiar button and a quick route through. The product needed to accommodate both.
§ 03 Methodology
Keep the choice at the greeting.
The design pairs a short keypad menu with an invitation to speak. Keypad input supplies the selected route; speech supplies the caller’s request in their own words. Either path should carry that context forward without asking the patient to start again.
Booking follows the clinic’s configured system and rules. Requests requiring staff or clinical judgment go to the appropriate team. The choice at the greeting should not alter those boundaries.
The reported measure is the caller’s first input after the hybrid greeting: speech or a keypress. It describes how callers began, rather than whether a booking or request was completed. Calls that end before any input are a separate measure of access and are not represented by the two-way split.
or explain the request
§ 04 Results
52% spoke. 48% pressed a key.
When offered both options, 52% of callers began by speaking and 48% used the keypad. In these observations, neither route was a marginal preference. Each served close to half of callers.
The finding supports a hybrid approach. Retain the menu for familiarity and ease of access while making conversational AI available alongside it. Adopting new technology does not need to mean removing a route patients already know.
§ 05 Discussion
The answer is both.
A near-even split gives the clinic little reason to impose one entry point on everyone. The keypad preserves a familiar, predictable way to make a routine request. Speech gives callers room to explain context when a fixed menu is too narrow.
As patients become comfortable with AI, clinics can offer its capabilities without making adoption a condition of access. Both routes can use the technology behind the scenes where the clinic’s systems support it. The patient chooses how to start; the clinic’s rules determine what happens next.
Keep the greeting short and carry the request forward. The next test is whether both routes deliver a clear outcome without repeated questions or unnecessary handoffs.
§ 06 Limitations
Entry preference is not the whole outcome.
The 52/48 split is a descriptive observation, not a controlled comparison or a universal ratio. The sample size and collection dates are not included in this edition. Clinic configurations and caller needs vary.
First-input choice does not establish which route resolves requests faster or produces more bookings. It also does not identify the caller’s age or comfort with technology. Those questions need separate measures, including calls abandoned before input.
§ 07 Conclusion
Keep the menu. Let patients speak.
There is no single right entry point for every caller. What we observed makes the case for both: preserve familiarity and ease of access as clinics transition to AI, and use the technology where it is enabled. Let adoption expand what patients can do without taking away the route that already works for them.