Only 54.8% of Calls Are Signed: Designing a Caller ID Attestation Gate for the Callbot Dialer Loop

Three Attestation Grades: What A, B, and C Actually Mean

STIR/SHAKEN is the framework that lets an originating carrier digitally sign a call, attesting that the caller is authorized to use the number it's dialing from. Signing comes in three grades. Full Attestation (A) means the carrier verified both the caller's identity and its right to use that specific number. Partial Attestation (B) means the carrier verified the caller but couldn't confirm number ownership. Gateway Attestation (C) means the carrier only knows where the call entered its network, with no verification of who originated it. A callbot or sales dialer pushing bulk calls through a SIP trunk can land in any of the three grades depending entirely on how the trunk and number registration are set up.

The grade isn't something the sender declares — it's assigned by the carrier. The same dialer can get different grades depending on which carrier handles the call and how the outbound number was registered, and a lower grade raises the odds that a receiving handset or carrier flags the call as spam.

Signed Isn't Equal: Spam-Flag Rates by Grade

TransNexus, which publishes monthly STIR/SHAKEN statistics, reported that in August 2026 robocalls made up 3.2% of all calls overall — but the breakdown by attestation grade showed a sharp gap. Calls signed with A-attestation carried a 1.6% robocall rate, B-attestation calls 3.8%, and C-attestation calls 7.8% — meaning a C-signed call was about 4.9x more likely to be flagged as a robocall than an A-signed one. In that same month, only 54.8% of calls were confirmed signed at termination, and the share carrying full A-attestation was still climbing — 27.6% in April, up to 30.8% by June. Most traffic is still sitting at B, C, or unsigned.

Wiring an Attestation Gate Into the Dialer: From Design to Operations

Start by declaring the target numbers up front: keep A-attestation share of outbound lines at 95% or higher, cap the spam-flag conversion rate (based on recipient reports) at 1% or under, and resolve a newly registered number's attestation grade within 48 hours. Without those numbers, teams tend to blame every grade drop on the carrier and never touch their own trunk configuration.

Failures repeat in a handful of recognizable shapes. First, setting the outbound caller ID to a number the trunk doesn't actually hold a registered right to use, so the carrier can't verify ownership and downgrades the call to B. Second, bursting too many calls from the same number in a short window, which trips a carrier's robocall-detection model regardless of the attestation grade it assigned.

Third, rotating outbound numbers frequently and pushing a brand-new number straight into high-volume dialing before its grade and reputation have stabilized. Fourth, never monitoring for carrier-side policy changes, so a line that held A-attestation for weeks quietly drops to B and the team only notices once the connect rate has already fallen.

Pre-declare the recovery branches. When a grade drops, immediately throttle volume on that outbound number, start the carrier's number-ownership re-verification process, and route traffic to a backup line with a stable grade until re-verification clears. Registering with a branded-calling or Rich Call Data service — which displays the caller's verified name or logo on the handset — can also partially offset the risk of being mistaken for spam even while a line sits at B.

The operational checklist should include test calls placed through each major carrier before deployment to directly confirm the attestation grade and how it renders on the receiving handset, a logging standard that makes the attestation grade a required CDR field, and a recurring pull of spam-report counts per outbound number from carrier portals. Caller identity and call content should never be logged in the clear — mask every identifier that isn't operationally necessary.

For the continuous-improvement loop, compare each outbound number's grade distribution and spam-flag conversion rate weekly to catch numbers drifting downward early, and keep grade-change history on the same timeline as volume-change history so you can trace which caused which. When switching number providers or trunk carriers, compare grade distributions for two weeks before and after the switch to catch any regression.

Takeaways at a Glance

Caller ID attestation isn't something you can fix after a call is placed — it's locked in by how the number is registered and the trunk is configured. Hold the line at 95% A-attestation and a 1% spam-flag conversion cap, and pre-declare volume throttling plus backup-line failover as the recovery branch for any grade drop, so the gate still holds even when a carrier changes its robocall-detection criteria underneath you.

References

STIR/SHAKEN statistics from August 2026 — TransNexus

STIR/SHAKEN trusted calling — Twilio