Who sees your business name when you call?

The short version

When your business calls a customer, three separate things decide what their screen shows: the number, the name, and any brand (logo, call reason) on top. Each one comes from a different system, run by different companies.

• The number travels with the call. US carriers are required to pass the calling party number to the next provider on the call path.6
• The name, in the US, usually does not travel with the call at all. The receiving carrier looks it up in a Caller ID Name (CNAM) database, and the name is only as accurate as that database.1 Many mobile carriers show it only to subscribers who opt in, sometimes for a monthly fee.2910
• The brand (a longer verified name, a logo, a reason for the call) is a newer, paid service. It is signed by the calling side, vetted by an intermediary, displayed only on certain networks and handsets, and billed per delivered call.141617

This page is about the name, not the spam warning; for a "Spam Likely" label see why your business number shows as Spam Likely.

How CNAM works: the receiving carrier asks a database

The FCC's own history is the best summary. Once the calling number travelled with calls, "To enhance the information provided to the called party, terminating providers began to query the Caller ID Name (CNAM) databases to identify and transmit the subscriber name associated with the number. The accuracy of the name presented to the called party depended upon the accuracy of the CNAM databases."1

That sentence explains most CNAM complaints. Your phone system does not send the name. It is a record attached to your number, and the customer's carrier decides whether to look it up (a "dip"), which database to ask and whether to trust a cached copy.

Where the record lives, and how it gets there

Your voice provider stores the name with a CNAM database operator. Two providers describe the process:

• Twilio says it keeps "its CNAM Line Information Databases (LIDB) up to date with the correct CNAM values", but warns that "not all carriers subscribe to every CNAM database, and Twilio can't control how carriers or database providers manage or update their records." After approval, "the display name may take 48-72 hours to propagate to all carriers in the United States."2
• Sangoma: "We then push this information to our CNAM vendor once a day (typically once a night)", and it takes "approximately 48 business hours for the information to begin displaying correctly on dips". Carriers that cache names may keep showing an old one.3

So there is no single national CNAM database, and a receiving carrier may subscribe to only some of them. The FCC's 2025 notice recorded "broad agreement among commenters that CNAM databases should not be used" to deliver verified caller names, "because, they asserted, CNAM databases are error prone and unreliable." It summarised the record as saying the databases "reportedly are not reliably accurate and are subject to manipulation."1 Those are commenters' views as the FCC reported them, not an FCC finding, but they came from trade bodies, analytics firms and CPaaS providers alike.

The 15-character limit

The name is short. Twilio: "Can be a maximum of 15 characters", it "Should begin with a letter and can only contain letters, numbers, periods, commas and spaces", and it cannot be "a generic value such as a City/State".2 Sangoma: "limited to 15 characters including spaces", with longer bulk entries truncated.3 That is why so many business names arrive as NORTHSIDE PLUMB or MAPLE DENTAL CL. Plan the abbreviation yourself, or a truncation will choose it for you.

The dip costs money, which is why it is not universal

Twilio's own price for an inbound caller-name lookup is $0.01 per request, "Only available in the US".5 Its documentation says the lookup "is billed per request, even if no data is returned", and that caller name "is only available for phone numbers owned by carriers in the United States".4 Multiply a cent by billions of calls and it is easy to see why carriers cache results or sell the lookup to subscribers as an add-on.

What US law does and does not require

The rules require carriers to pass the number. Under 47 CFR § 64.1601 they "are required to transmit for all PSTN Traffic the telephone number received from or assigned to or otherwise associated with the calling party to the next provider in the path". Nothing in that section obliges a carrier to deliver a calling name. Names appear in two places only. A carrier honouring a privacy request "may not reveal that caller's number or name", and telemarketers must transmit caller identification that includes, "when available by the telemarketer's carrier, the name of the telemarketer."6

The Truth in Caller ID provision makes it unlawful to "knowingly transmit misleading or inaccurate caller identification information with the intent to defraud, cause harm, or wrongfully obtain anything of value", and "caller identification information" covers "other information regarding the origination of" a call, not only the number.7 A name chosen to impersonate another business is a legal problem as well as a technical one.

Why mobile phones often show no name

CNAM was designed for landline caller ID boxes. On mobiles, the carrier's product decisions and the handset decide. Twilio puts it bluntly: "For mobile phones, most US carriers offer CNAM as an opt-in feature", and "If the recipient doesn't enable CNAM for their mobile phone, CNAM won't be displayed even if a CNAM is properly set for the number."2

The three national US carriers handle it in three different ways, according to their own support pages:

Scroll the table sideways to see every column.

So many mobile recipients see only your number, however carefully you registered CNAM. And the customer's saved contact beats everything: AT&T lists the address book first, and Rogers says that if a caller is saved, "the name you stored on your phone will display".1032 Getting into your customers' contacts (a vCard in a confirmation email, a "save this number" line on invoices) is the cheapest caller ID upgrade there is.

Handsets now add their own layer. Apple's iPhone guide says the phone "can help identify incoming calls using information from Apple Business Connect, supported carriers, and supported call identification apps". The current version also offers Call Screening: calls from unknown numbers are answered automatically, and "After the caller shares their name and reason for their call, your iPhone rings".28 On a screened iPhone, the first impression your business makes may be the sentence you say to a machine, not the name on the screen.

Rich Call Data: the name moves to the calling side and gets signed

STIR/SHAKEN, the call-authentication framework US and Canadian carriers were required to deploy131, signs the calling number and an attestation level (A, B or C). It says nothing about the name. Rich Call Data (RCD) extends that signature to identity information. The FCC describes the two key differences from CNAM: "instead of the terminating provider querying a third-party database to obtain the name associated with the originating telephone number, RCD relies upon the caller's service provider to provide and transmit the caller name", and RCD "relies upon authentication by the originating provider and verification by the terminating provider within the STIR/SHAKEN framework."1

The two IETF standards (July 2025)

RFC 9795, "Personal Assertion Token (PASSporT) Extension for Rich Call Data", Standards Track, July 2025.11 It defines an rcd claim inside the signed token carried in the SIP Identity header:

• nam: the display name (required; an empty string if there is none);
• apn: an optional alternate presentation number;
• icn: an optional HTTPS URL to an image (a logo);
• jcd / jcl: an inline jCard or a link to one (mutually exclusive);
• crn: a call reason, kept separate from rcd because call reason "is not the same as information associated with the caller and may change on a more frequent, per-call basis";
• rcdi: integrity digests for anything referenced by URL, so a logo cannot be swapped after signing.

For SIP, the signed nam must match the From header's display name.11

RFC 9796, "SIP Call-Info Parameters for Rich Call Data", also July 2025, carries similar data in the SIP Call-Info header without the token. It adds a jcard purpose, a call-reason parameter "intended to convey a short textual message suitable for display to an end user during call alerting", a verified parameter whose only valid value is true, and an integrity parameter.12 The difference matters: RFC 9795 data is cryptographically signed, whereas RFC 9796's verified=true is a sender's assertion.

The ATIS layer, and its open problems

Carriers implement the IETF pieces through ATIS's SHAKEN documents. The FCC cites ATIS-1000094 version 2, "Calling Name and Rich Call Data Handling Procedures", published in April 2025.1 ATIS's own January 2026 filing to the FCC describes it as introducing "mechanisms for authentication, verification, and transport of calling name as well as other enhanced caller identity information (e.g., images, logos) and call reason". The same filing is candid about maturity. It says "there is currently no industry testbed with respect to RCD for interoperability testing between implementations", that "there are multiple RCD specifications (IETF, IP-NNI Task Force), and not all these specifications are aligned", and that "Interoperability is dependent on vendor implementation."13

ATIS also corrected the FCC on a point small businesses should understand. The FCC had said the RCD standard requires the originating provider to vet the caller identity information. ATIS replied that "there are situations in which the originating service provider cannot and does not vet the caller identity information". The name may instead come "from a trusted non-SHAKEN entity such as an Authoritative Database".13 We expect that, in practice, the trusted entity will be the vetting company you pay. The ATIS standard is not published openly, so this page describes it through the FCC's and ATIS's public filings, not from its text.

Privacy is kept. The FCC quotes the ATIS standard: the terminating provider "shall not convey any rich call data to the called party device if the calling party has requested privacy".1

Branded calling programmes in 2026

"Branded calling" is the commercial wrapper around these standards. Some programmes are standards-based, some proprietary, and the FCC itself admits it is "not completely clear to what extent any of these services are proprietary or use RCD in whole or part".1 Each row below reflects only the provider's own pages.

Scroll the table sideways to see every column.

A few details in those pages matter more than the marketing around them:

• Name lengths differ by network. Twilio's Basic tier allows 15 characters on Verizon and 32 on T-Mobile, letters and numbers only; the name "Must not contain Personally Identifiable Information (PII)" or web links.16 If you brand across networks, write for the shortest.
• Reputation can disqualify you. Twilio warns: "If you get a high number of SIP 608 or 603 responses in your outbound calls, you might not be able to use Branded Calling" (608 is the response an intermediary, typically an analytics engine, returns when it rejects a call35).16 Branding does not rescue calls customers already reject.
• Apple's "BCID" is not CTIA's BCID. Apple announced Business Caller ID on 16 October 2024, promising a business's "name, logo, and department appear on the inbound call screen".26 Its developer documentation uses the same abbreviation for a different system, requires "an explicit, unambiguous association between a phone number and a delegated brand", and says that otherwise "the iPhone will rely upon fallback measures."27 It gives no price. Brand customisation in Business Connect is "for free"26; the partner hosting the RCD is a separate commercial relationship.
• Google's Verified Calls has gone. Telnyx's support page (a secondary source) says "Google Verified Calls is no longer offered by Telnyx. This product was sunset by Google."29 We found no Google page giving a date. On Android, branding now comes through the carrier (T-Mobile's "Verified Business Call" setting20) or the handset maker: Twilio notes that "Samsung devices include a 'Smart Call' feature" that shows a business name "regardless of the recipient's mobile network".19

What it costs, at published list prices

CNAM registration itself is often free. Sangoma states "There are no additional charges for Outbound CNAM".3 Branded calling is not. Published list prices on the date of this page, before call charges and tax:

Scroll the table sideways to see every column.

As a worked example only: a business whose staff place 300 branded outbound calls a month would pay about $36 at Twilio's per-call rate, $42 at Hiya's per-call rate, or $52 on First Orion's 500-call plan (plus the $25 one-off), on top of normal calling costs. Note what you are buying: delivery on supported networks and devices. Hiya says outright that "Not every outbound will be branded."25

We think it is clearly worth it where revenue depends on customers answering your calls: clinics confirming appointments, contractors returning quote requests, field-service dispatch. It rarely is for businesses that mostly receive calls and only call back people who just rang them, because those customers already have the number in their recent calls.

The FCC proposal: a verified name whenever the checkmark shows

In October 2025 the FCC proposed making some of this mandatory. The Ninth Further Notice of Proposed Rulemaking (FCC 25-76, adopted 28 October, released 29 October 2025) proposes "to require terminating voice service providers to transmit verified caller name for presentation on consumers' handsets whenever they transmit call authentication information indicating that the originating number is unlikely to be spoofed."1

The detail is narrower than it sounds:

• The duty would bite only when a carrier chooses to show an A-level indicator: "we do not propose to require terminating providers to transmit to consumer's handsets whether a call has received an A-level attestation".1
• The proposed minimum is "a verified name, whether personal or business", because "some callers, such as individual callers, will not have a brand logo".1
• The FCC's reasoning: presenting an A-level indicator with only the number "provides little benefit to consumers because they might not understand the meaning of the indicator, mistakenly taking it to indicate that the call is not a scam".1
• It also proposes that originating providers verify the caller identity information they transmit, and it asks whether to require RCD on IP networks.1

Status: still a proposal, as far as we could find. A separate Know-Your-Customer notice adopted on 30 April 2026 asks how those rules can "complement any Call Branding or Caller Name requirements we may adopt".34 We found no order adopting the caller-name proposal.

Note that the evidence the FCC cites for branded calling's benefits comes largely from companies that sell it, including TransUnion's statement that customers are "up to 105% more likely to answer a branded call" and survey figures from TNS, First Orion and Hiya.1 Those numbers may be right, but none of them is independent measurement, and we would not plan a budget around them.

Canada: the name comes from the calling network

Canadian caller name delivery differs from the US at the root. The CRTC's 2018 decision on caller ID spoofing records parties' submissions that "In the United States, CNAM is provided by a look-up to separate databases, while in Canada this information is provided by the originating network". Those parties warned that authenticating names would need "a unique Canadian CNAM specification".30 That is a 2018 summary of submissions, not a standard, but current vendor documentation agrees: Talkdesk, a contact-centre platform, says Canada "generally does not rely on a centralized CNAM database lookup" and that the originating carrier places the 15-character name in the SIP signalling. It also warns that a call from a US number crossing "legacy interconnection points" may lose it.33

What follows:

• US CNAM tools do not apply. Twilio cannot assign CNAM profiles to Canadian numbers2, and its caller-name lookup covers only numbers owned by US carriers4. For a Canadian number, your provider sets the name in its own network, and you ask that provider.
• Seeing names on a Canadian mobile is a separate feature. Rogers distinguishes Call Display (the number) from Name Display ("both a caller's name and number"): "you need to have Call Display to add Name Display", and you contact Rogers to add it. Names may not appear for calls from a "business switchboard, an 800 number", for calls "made outside of Canada", and for forwarded calls, where "you might see the contact information for the line that forwarded the call."32
• Mobile-line names are personal by default. A Rogers wireless line's caller ID is "the first and last name attached to your wireless phone number". Postpaid customers can change it, with no special characters or numbers, and a change "could take up to 24 hours to appear on your outgoing calls."32 A sole trader calling from a personal mobile shows their own name, not the business.
• STIR/SHAKEN is mandatory, names are not signed. The CRTC made STIR/SHAKEN for IP-based calls a condition of service "effective 30 November 2021", although carriers had objected to the timing because issues including "handset display standards" were unresolved.31
• Branded calling reaches Canada only partly. Twilio's Non-US Basic Branded Calling lists Canada (with Germany and the UK) as a public beta with a 32-character name.19 First Orion says it supports Rogers and Bell Canada.22 Apple's Business Caller ID is "limited to the USA".27

We found no CRTC decision regulating branded calling or Rich Call Data as such (CRTC documents here were read through Internet Archive copies). In Canada, the useful question for your provider is concrete: "What name do you send in the signalling for this number, and can I change it?"

What a small business can do today: outbound calls

When you call customers, in rough order of value for money:

1. Know which identity each call presents. A staff member's own mobile shows their number and, where a name shows at all, their name. A call through your business service shows the business number. Choose the one customers should learn and can call back. (One 2026 option bridges the two: First Orion's T-Mobile programme brands calls placed from business mobiles and routes callbacks "to a centralized business number rather than individual employee lines".24)
2. Register CNAM on every outbound business number (US). Write the 15 characters yourself, wait the 48 to 72 hours providers quote23, then test: a landline, and a phone on each of T-Mobile, Verizon and AT&T. Expect the name on the landline and on T-Mobile, and possibly only the number elsewhere8910.
3. Get into your customers' contacts. A saved contact overrides every database1032. Put a "save us as a contact" link (a vCard) in booking confirmations and on invoices.
4. Make sure your provider signs your calls. Branded calling products build on STIR/SHAKEN (Twilio's Enhanced tier requires it17), and the FCC proposal ties verified names to A-level attestation1. Ask your provider in writing what attestation your outbound calls receive.
5. Prepare for call screening. On iPhones with Call Screening on, an unknown caller is asked for a name and reason before the phone rings28. Give staff a short line to say ("This is Northside Plumbing about your Tuesday booking") and leave a voicemail when nobody answers.
6. Pay for branding only where answer rates are revenue. Test a month on one number with a per-call plan against an unbranded number. Your own answer rates are evidence; vendors' figures are marketing.
7. Never borrow someone else's name. Misleading caller identification with intent to defraud or cause harm is unlawful in the US7, and branded-calling programmes vet names against your business records1516.

A spam label, as opposed to a missing name, comes from carrier analytics; the Spam Likely article covers it.

Inbound calls: what you see when customers call you

When a customer calls your business and the call is forwarded to a mobile, that forwarded leg is a new call, and the mobile's carrier decides what to show for whichever number the forwarding service presents:

• If the service passes through the customer's number, your staff see that number. They see a name only if the customer is saved in their contacts or their own mobile plan shows names (free on T-Mobile, paid on Verizon and AT&T, a separate feature on Rogers)891032.
• If it presents the business number, staff know the call is a business call but not who is calling. Rogers notes that on forwarded calls "you might see the contact information for the line that forwarded the call."32
• A service can look up the caller's name itself (Twilio charges $0.01 per lookup, US numbers only45) and announce it in a whisper before connecting. That adds cost per call and inherits all of CNAM's accuracy problems.

Which of these a service does, and what it means for voicemail and callbacks, is tested call by call in business call forwarding to a cell phone: caller ID and voicemail. One cheap fix whatever the setup: have every staff member save the business number as a contact ("Shop line"), so a forwarded call that presents it is recognisable at a glance.

For transparency: Auto Attendant, the service this site describes, is inbound-only (calls routed to ordinary mobiles, no outbound calls), so it sells none of the outbound branding above; the inbound questions here apply to it as to any forwarding service.

Our view: caller ID is becoming a paid trust layer

Follow the money. Under CNAM, the caller's provider registered the name (often at no charge3) and the receiving carrier paid to look it up5: near-zero cost to the caller, and near-zero guarantee. In 2026 the trustworthy version of the name is priced per delivered call on the calling side, between about 12 and 14 cents at the list prices above1825, with vetting intermediaries in the chain. On two of the three national US carriers, the recipient also pays a monthly fee to see names at all910. A business pays to be recognised, and its customer pays to recognise it.

We think that has three consequences for small businesses:

1. The basic CNAM name will keep decaying. Industry commenters already call the databases unreliable1, and handsets increasingly prefer carrier and platform identification28. Register it, because it is nearly free, but do not count on it.
2. Branded calling rewards volume. Per-call billing, monthly tiers and integration for logos222325 suit contact centres. A plumber making 40 callbacks a week can buy it, but a clinic confirming 2,000 appointments gets far more per dollar.
3. The FCC proposal is the variable to watch. If a verified name becomes the required companion to every A-level checkmark1, a basic verified name could become part of ordinary phone service, and the cost would move to originating providers, which pass it on in their prices. If the proposal stalls, the market stays a toll road. Neither outcome was settled on the date of this page34.

So: register CNAM, earn a place in customers' contacts, get your calls signed, and buy branding only when a test on your own calls shows it pays.