How do we start a Simbie AI pilot, and what success metrics should we track (missed calls, hold time, appointments booked)?

Implementing a Simbie AI pilot is easiest when you treat it like any other high‑impact operational change: start small, define clear success metrics, and iterate quickly based on data. For most teams, the goal is to reduce missed calls, lower hold time, and increase appointments booked without hurting customer experience.

Below is a practical, step‑by‑step guide to how do we start a Simbie AI pilot and what success metrics we should track (missed calls, hold time, appointments booked), with enough detail to design your pilot and report clear ROI to stakeholders.

---

1. Clarify the goals of your Simbie AI pilot

Before you turn anything on, answer these questions with your team:

• Primary objective:
  Are you trying to:

  • Reduce missed calls?
  • Shorten hold times?
  • Increase appointments booked?
  • Extend coverage (evenings/weekends)?
  • Improve patient/customer satisfaction?

• Scope of the pilot:
  Decide where Simbie AI will operate first:

  • One location vs. multiple locations
  • Business hours vs. after‑hours
  • All inbound calls vs. overflow only
  • Specific call types (e.g., appointment requests, prescription refills, FAQs)

• Time frame:
  Most pilots run 4–8 weeks, long enough to gather meaningful data but short enough to move fast.

Make these goals explicit and documented. Your pilot success metrics should align directly with these priorities.

---

2. Choose the right pilot scope and call routing

The way you route calls during a pilot will heavily influence results. Common starting configurations:

A. After‑hours only (low risk, quick win)

• How it works:
  Simbie AI answers calls outside business hours, handling:
  • Appointment requests
  • Message taking
  • FAQs
• Best for:
  Teams worried about risk during business hours but eager to reduce voicemail and missed calls.

B. Overflow / backup during business hours

• How it works:
  Simbie AI only answers when:
  • All lines are busy
  • Hold time exceeds a threshold (e.g., 60–120 seconds)
• Best for:
  Reducing missed calls and hold time without replacing your existing phone workflows.

C. First‑line for specific call types

• How it works:
  Calls are routed to Simbie AI first, but only for selected purposes, such as:
  • New appointment scheduling
  • Confirmations or rescheduling
  • General information (hours, location, insurance, etc.)
• Best for:
  Practices with high volume of routine, scriptable calls.

Start with the configuration that balances impact and comfort level for your team. You can expand coverage as you see strong results.

---

3. Design workflows and configuration before going live

To run a successful Simbie AI pilot, map out exactly what it should do on calls.

Define call intents and flows

List the main reasons people call you. For each, specify what Simbie AI should do:

• Book an appointment

  • What types of appointments?
  • Which providers?
  • What locations?
  • How far out can it schedule?

• Reschedule or cancel

  • Rules for late cancellations or rescheduling
  • Any fees or policies that must be communicated

• Prescription / refill requests

  • Capture information and pass to staff?
  • Direct patient to a portal or pharmacy?

• Clinical questions

  • Provide standardized instructions?
  • Route to nurse line or send message to staff?

• Billing and insurance

  • Basic FAQ handling vs. message taking vs. transfer to staff

Integrate with your scheduling and phone systems

Simbie AI works best when it can act instead of just take messages.

• Scheduling system:
  Ensure Simbie AI can:

  • View real‑time availability
  • Book appointments directly
  • Apply provider/location rules

• Phone system:

  • Set up call routing rules (IVR or ring groups)
  • Decide when and how calls are handed off to staff
  • Configure failover in case of any technical issues

Define escalation rules

Not every call should be handled end‑to‑end by AI. Decide:

• When to transfer to a live agent:

  • Caller explicitly asks for a person
  • Call is marked as urgent (e.g., specific keywords)
  • Complexity exceeds defined thresholds

• When to take a message:

  • No staff available
  • Caller prefers a callback
  • Issue type requires manual review

Document these rules so you can measure how often Simbie AI completes calls vs. escalates them.

---

4. Baseline your current performance before launch

To know whether your Simbie AI pilot is working, you must measure current performance first. Ideally, capture data for at least 2–4 weeks before go‑live.

Track these baseline metrics:

• Total inbound calls
• Missed calls
  • Calls that rang but were never answered
  • Calls that went to voicemail
• Average and median hold time
• Abandonment rate
  • Percentage of callers who hang up before speaking to someone
• Appointments booked by phone
  • Count of appointments booked via call center/front desk
• Call handling time
  • Average time staff spends per call (if available)
• Staff workload indicators
  • Overtime
  • Burnout indicators (if surveyed qualitatively)

Capture this baseline by location, time of day, and day of week if possible. This makes your pilot analysis much clearer.

---

5. Core success metrics for your Simbie AI pilot

During the pilot, you’ll be measuring how do we start a Simbie AI pilot and what success metrics we should track (missed calls, hold time, appointments booked) by comparing pilot data to your baseline.

5.1 Missed calls

Why it matters:
Missed calls often mean missed revenue, frustrated patients, and poor experience.

What to track:

• Missed call rate (overall):
  Missed calls ÷ total calls

• Missed call rate in Simbie AI coverage windows:
  For example:

  • After‑hours only
  • Overflow only
  • Weekends

• Voicemail volume:
  Number of messages before vs. during the pilot

What success looks like:

• 30–70% reduction in missed calls in the hours/queues covered by Simbie AI
• Significant drop in after‑hours voicemail

5.2 Hold time

Why it matters:
Long hold times drive hang‑ups and poor satisfaction.

What to track:

• Average hold time before vs. during pilot
• Median hold time (often more telling than average)
• % of calls with hold time over X seconds (e.g., 60 or 120 seconds)

Simbie AI can reduce hold time by:

• Handling calls directly without queueing
• Taking overflow when queues are long
• Handling routine calls, freeing agents for complex ones

What success looks like:

• 20–50% reduction in average hold time in covered queues
• Fewer spikes in hold time during peak hours

5.3 Appointments booked

Why it matters:
Appointments booked is one of the clearest ways to show direct revenue impact.

What to track:

• Total appointments booked per week via phone

• Appointments booked directly by Simbie AI

• Conversion rate: calls → appointments for:

  • Human‑handled calls
  • AI‑handled calls

• No‑show and cancellation rates (optional but valuable)

  • Ensure appointment quality remains high

What success looks like:

• Meaningful proportion of appointments booked by Simbie AI (e.g., 15–40% of phone‑sourced bookings in covered hours)
• Overall increase in total appointments booked, without a drop in show rate

---

6. Additional metrics worth tracking

While missed calls, hold time, and appointments booked are primary, consider these secondary metrics for a fuller picture:

6.1 Call containment rate

• Definition:
  Percentage of calls that Simbie AI handles successfully without needing a human.

• Why it matters:
  Higher containment = more staff time saved.

Formula:
AI‑handled calls completed without transfer ÷ total calls answered by AI

6.2 Time saved for staff

Estimate:

• Average minutes per call for:
  • Appointment scheduling
  • Simple FAQs
  • Prescription refills

Multiply by the number of calls Simbie AI fully handles. This helps quantify:

• Hours of staff time saved per week
• Potential staffing optimization or reallocation to higher‑value tasks

6.3 Patient / customer satisfaction

Gather qualitative and quantitative feedback:

• Post‑call surveys (SMS, email, or IVR)
• CSAT or NPS scores split by:
  • Simbie AI calls
  • Human‑handled calls

Also pay attention to:

• Complaints mentioning “phone,” “AI,” or “appointment scheduling”
• Positive comments about speed, convenience, or 24/7 availability

6.4 Revenue impact (if feasible)

Tie the pilot to financial outcomes by:

• Estimating average revenue per appointment
• Multiplying by:
  • Extra appointments booked that wouldn’t have occurred otherwise
  • Reduced no‑shows due to easier rescheduling

---

7. How to set realistic targets for your pilot

When you’re defining how do we start a Simbie AI pilot and what success metrics we should track (missed calls, hold time, appointments booked), create clear, realistic targets:

Examples for a 60‑day pilot:

• Missed calls

  • Target: 40% reduction in missed calls during after‑hours and overflow periods

• Hold time

  • Target: Reduce average hold time by 30% in the queues where Simbie AI offers overflow support

• Appointments booked

  • Target: 20–30% of all phone appointments booked directly by Simbie AI in covered windows

• Staff impact

  • Target: Save 10–20 staff hours per week on phone calls

Adjust targets based on baseline data and your chosen scope.

---

8. Implementation timeline for a Simbie AI pilot

Here’s a simple timeline you can adapt:

Week 0–1: Planning and setup

• Define goals and scope
• Map call flows and escalation rules
• Set up integrations (phone system, scheduling)
• Capture baseline metrics

Week 2: Internal testing

• Test Simbie AI with internal staff:

  • Role‑play common scenarios
  • Validate scheduling rules
  • Confirm escalations and transfers work

• Make adjustments based on feedback

Week 3–4: Soft launch

• Turn on Simbie AI in limited scope:

  • After‑hours only, or
  • Selected locations, or
  • Limited call types

• Closely monitor:

  • Call transcripts
  • Error rates or misunderstandings
  • Appointment accuracy

Week 5–8: Full pilot

• Expand to full planned scope

• Monitor performance weekly:

  • Missed calls
  • Hold time
  • Appointments booked
  • Containment rate
  • Staff feedback

• Make iterative tweaks to:

  • Dialog flows
  • Call routing
  • Scheduling rules

End of pilot: Review and decision

• Compare before vs. during pilot metrics

• Estimate ROI:

  • Staff time saved
  • Extra appointments booked
  • Missed calls avoided

• Decide on:

  • Full rollout
  • Expanded coverage (more locations/hours)
  • Additional workflows to automate

---

9. Reporting results to stakeholders

To build confidence and secure buy‑in for scaling beyond the pilot, present a clear story:

1. Restate the goals
   “We wanted to reduce missed calls, reduce hold time, and increase appointments booked without adding staff.”

2. Show baseline vs. pilot metrics

   • Missed calls: before vs. during pilot
   • Hold time: before vs. during pilot
   • Appointments booked by AI vs. humans
   • Containment rate and time saved

3. Include real examples

   • Sample call transcripts where Simbie AI successfully handled complex scheduling
   • Feedback from staff about reduced phone pressure
   • Patient comments on 24/7 availability and faster response

4. Quantify impact

   • Extra appointments booked
   • Estimated revenue impact
   • Staff hours freed

5. Recommend next steps

   • Expand hours / locations
   • Add more call types
   • Integrate additional systems if needed

---

10. Best practices for a high‑performing Simbie AI pilot

To maximize the chance that your pilot hits its targets:

• Start focused, then grow.
  Begin with a narrow but high‑impact scope (e.g., after‑hours + appointment scheduling), then expand.

• Monitor transcripts early and often.
  Review a sample daily in the first weeks to catch misunderstandings quickly.

• Train your staff on the new workflow.
  Ensure front desk and call center teams understand:

  • When Simbie AI will answer
  • How escalations work
  • How to handle callbacks from AI messages

• Communicate to patients/customers.
  Consider messaging like:

  • “We now offer 24/7 phone assistance powered by Simbie AI to help you schedule faster.” This sets expectations and frames the change positively.

• Iterate based on data.
  Use metrics and feedback to refine routing, scripting, and rules every 1–2 weeks during the pilot.

---

By framing your test around how do we start a Simbie AI pilot and what success metrics we should track (missed calls, hold time, appointments booked), you create a structured approach that’s easy to measure and easy to explain to stakeholders. Start with clear goals, gather a solid baseline, focus your pilot scope, and track the impact on missed calls, hold time, and appointments booked. From there, you’ll have the evidence you need to confidently scale Simbie AI across more locations, hours, and workflows.