Many clinics using Jaz, TrackStat’s AI voice assistant, occasionally see calls that end very quickly - before Jaz has a chance to greet the caller or start a conversation. These are commonly spam or robocalls. This article explains the typical patterns, why they occur, how to interpret them, and practical steps to reduce the volume of these calls.
Common Call Outcomes You May See
In your call logs or feedback, these interactions often appear with notes such as:
Call ended before AI could speak with caller
Call ended before any live conversation between caller and AI
Call was interrupted - caller hung up before interacting
Caller ended the call before any conversation
These calls are usually very short (often under 10–15 seconds) and contain little or no speech from the caller.
Why These Calls Happen
Spam and robocall systems frequently:
Use automated dialers that probe for a live person.
Detect an automated or AI greeting and disconnect immediately.
Abandon the call if no agent is available on their end (common with predictive dialers).
Use local-looking numbers (including your area code) via neighbor spoofing or temporary numbers.
Because Jaz answers promptly, these systems often hang up as soon as they hear the greeting or confirm the line is active. This behavior is widespread across business phone lines and is not unique to AI receptionists.
How to Tell Spam from Real Potential Patients
Most short, pre-conversation hang-ups are spam. Clues that point toward spam include:
Multiple calls clustered in a short time window (e.g., several in a few hours).
Numbers that do not match any existing patients in your database.
Very short duration with no meaningful speech from the caller.
Numbers already flagged in public robocall databases (optional check if volume is high).
Real patients or potential new patients sometimes hang up when they hear an AI voice, but they rarely arrive in tight clusters with zero database matches and immediate disconnects. If you see longer calls or actual conversation attempts that then end, those are more likely to be people who prefer speaking with a human.
What This Means for Your Clinic
These calls do not typically represent lost new-patient opportunities.
Jaz is functioning as designed by answering the line. The disconnect comes from the calling side.
High volumes of these calls are a common industry issue for any business phone number.
Can We Reduce These Calls?
Yes. Most clinics can enable spam or robocall filtering directly with their business VoIP provider (the company that supplies the office phone number that forwards to TrackStat when no one answers).
Common features offered by major business VoIP providers include:
Spam / Robocall filters (often called Call Guard, Call Protect, Robocall Protect, or similar) that score incoming calls and can automatically block or divert high-risk numbers before they ring or are forwarded.
Adjustable filter levels so you can choose how aggressively to block (e.g., known malicious only vs. also telemarketers and robocalls).
Anonymous call rejection to block calls with no caller ID.
Selective call rejection and manual block lists.
Allow / whitelist lists to ensure important numbers always get through.
Some providers also offer challenge-style screening (asking uncertain callers to press a key before connecting).
Recommended steps:
Contact your business VoIP provider and ask about their spam/robocall filter for your main office number.
Enable the filter on the number that is routed to TrackStat’s voicemail line.
Start with a moderate setting (block high-risk / known malicious calls only).
Review the provider’s blocked-call log for a few days to confirm legitimate calls are not being stopped.
Use an allow list for frequent or important callers (referring providers, labs, etc.).
Optionally turn on anonymous call rejection.
Blocking at the carrier level is the most effective approach because the spam call never reaches the TrackStat voicemail number or Jaz. No system stops 100% of spam, but these tools significantly reduce the short abandoned calls you see in your logs.
Recommended Practices
Review short abandoned calls separately from true missed opportunities or completed conversations.
Avoid calling back numbers that show this pattern, as they are rarely productive leads.
Monitor trends: occasional clusters are normal; sustained high volume may benefit from stronger carrier-level filtering.
If you notice a pattern of local callers who engage briefly and then hang up, consider whether a short early option for a human team member would improve the experience for those patients.