This is the phone version of deliverability, and it behaves the same way: your reputation attaches to the identifier you send from, it is earned through behaviour, and once it is damaged the fastest fix is a new identifier plus different behaviour.
The difference is that almost nobody in sales knows the phone system works this way, so number reputation gets ruined by accident.
STIR/SHAKEN, in plain English
Carriers now cryptographically sign calls to indicate how confident they are that the caller is who the caller ID claims. There are three levels of attestation, and the label your call carries follows you into the recipient’s handset.
| Attestation | Means | Consequence |
|---|---|---|
| Full (A) | The carrier knows you and knows you have the right to use this number | Best chance of ringing normally |
| Partial (B) | The carrier knows the customer but cannot vouch for the number | More likely to be scored down by analytics |
| Gateway (C) | The call came from somewhere the carrier cannot vouch for at all | Frequently labelled or blocked |
What this means in practice: numbers you own and register earn full attestation, numbers you borrow do not. That single fact ended the old local-presence industry, which worked by displaying local numbers from a shared pool that nobody could attest to.
Buying local is not spoofing local
The distinction matters legally and operationally.
Legitimate local presence: you buy or lease a number in the area code, it is registered to your business, a call back reaches you, and it appears on your outbound calls because it is yours. At $2.50 a month for a local number and $4.50 for toll-free, running one per market you sell into is a rounding error on any calling budget.
Spoofing: displaying a number you do not control. Increasingly detected, increasingly blocked, and in many jurisdictions unlawful when done to mislead. It is also self-defeating: the calls that get through reach people who cannot call you back.
The test to apply to any provider selling “local presence”: if the prospect rings this number back tomorrow, what happens? If the answer is “nothing”, you are buying the version that no longer works.
What actually earns a Spam Likely label
Carrier analytics score behaviour on a number, and the area code is almost irrelevant. The signals that matter:
- Dial volume with a low answer rate. Four hundred dials and a 4% answer rate looks exactly like a robocall, because behaviourally it is one.
- Very short average duration. A number whose calls average eight seconds is telling the network that recipients hang up on it.
- Bursts. Two hundred calls in twenty minutes from a number with no history is the clearest pattern there is.
- Complaints. Handset-level “report spam” taps carry disproportionate weight.
- No history at all. A brand new number with immediate high volume gets treated as suspicious, in the same way a brand new sending IP does.
Read that list again and notice that four of the five are volume and pacing. This is a ramp problem, not a technology problem, and the fix is the same as it is in email: start low, build history, keep volume per identifier plausible for a human.
A number strategy that survives
- One number per rep per market, owned and registered. Not a rotating pool of thirty.
- Ramp new numbers. Twenty to thirty dials a day for the first week, then up. A new number with no history is fragile exactly like a new sending domain.
- Cap dials per number per day. Pick a ceiling a human could plausibly produce and enforce it in the tool rather than in a policy document.
- Watch answer rate and duration per number, not per campaign. Degradation shows up on one number first, and that is your early warning.
- Retire rather than appeal. Remediation exists and is slow. If a number is flagged, correct the behaviour and move on.
- Make callbacks work. Voicemail on every number, with your name and company. It is both better practice and a reputation signal.
The thing that improves answer rates more than any of this
Local numbers are worth having, and they are a second-order effect. The first-order effect is whether the person recognises you.
An unrecognised local number gets answered slightly more than an unrecognised out-of-state one. A number belonging to someone whose email you replied to, whose LinkedIn invite you accepted, or who was introduced by a colleague gets answered at a completely different rate. That is why the call sits after the other channels rather than instead of them: the sequence is what makes the caller ID mean something.
Next: what you can legally dial, by market, or cost per meeting on the phone.
Frequently asked questions
Does local presence dialing still work?
Buying and using a local number you own still helps answer rates, because recipients answer numbers that look local and familiar. What no longer works is rotating through borrowed numbers you have no relationship with, which is what carrier analytics are specifically built to detect.
What is the difference between local presence and spoofing?
Local presence means you own or lease the number, it is registered to you, and a call back reaches you. Spoofing means displaying a number you do not control. The first is legitimate and attestable under STIR/SHAKEN; the second is increasingly blocked outright and can be unlawful.
Why do my calls show as Spam Likely?
Almost always volume and behaviour on that specific number: a high dial count with a low answer rate, very short average call duration, many unanswered calls in a burst, or complaints. Area code has little to do with it.
How many numbers should I rotate through?
Enough that no single number carries an unnatural dial volume, and few enough that each one builds a history. In practice one number per rep per market, with volume caps, beats a large rotating pool.
Can a flagged number be fixed?
Sometimes, through the analytics providers’ remediation processes, but it is slow and unreliable. Retiring the number and correcting the behaviour that flagged it is usually faster than appealing.