Costs and pricing

How the money moves

Customers pay rent, registration and per-message fees; the per-message fees hold most of the surprises.

These are rough ranges as of late 2026, and they move. Carrier fees, the DCA fee, the campaign minimum, number rent and international prices come from provider and aggregator pages, not from the carriers.

Coming from SaaS, these surprised me:

• Registration is marked up. One large provider charges about twice what TCR does for a campaign each month, and about doubles the vet fee too.
• Carrier fees move mid-contract. All three big US carriers changed theirs between January and May 2026, each by memo. If your pricing can't pass a new carrier fee through within a month, you're paying for carrier price rises out of your margin.
• WhatsApp's October 2026 change is bigger than it looks. In Meta's own worked example, one customer interaction goes from one charge on July 1 to five on October 1, so I'd expect customers who run a lot of support over WhatsApp to feel it first.

How companies of different sizes use it

Spend here follows message and call volume more than headcount, so a small app sending login codes abroad can outspend a mid-market manufacturer. I've kept the usual five sizes but anchored each to a typical monthly volume of US texts. Nobody publishes messaging spend by company size, so the volumes and spend figures below are my rough estimates, built from list prices, one buyer-data platform's contract figures and Twilio's filings. The tail is long: Twilio had more than 400,000 active customer accounts at the end of 2025, and dividing its revenue by that count (my own division, not a figure Twilio reports) gives roughly $12,600 a year per account, with the typical account paying far less.

A startup sends under about 50,000 US texts a month: sign-up codes, notifications, maybe support replies. One 10DLC or toll-free number covers it, on a self-serve account with a platform such as Twilio, Telnyx or Plivo, paid by card or a prepaid balance. Often there's no messaging contract at all, because marketing texts go out through an email and SMS marketing tool such as Klaviyo, and a company selling in Latin America or India adds WhatsApp through a partner like 360dialog or Gupshup. A company too new to have an EIN can only register as a sole proprietor, with one campaign and one number. A founding engineer owns it and the founder signs the registration. My rough estimate is $1,000-10,000 a year, though an app with users abroad or a voice AI product can spend like a much bigger company. What pushes it on is the carriers' cap for an unvetted brand, which a flash sale or a burst of login codes hits fast, users abroad, where each text costs several times more and pumping starts, or a pumped sign-up form.

A scale-up sends from about 50,000 to a few million US texts a month across several use cases, each its own campaign under a vetted brand: login codes, account notifications, marketing and support. It adds a toll-free number for national support, perhaps a short code for marketing at the top of the band, voice for sales or support, and WhatsApp where its customers live. It usually stays on one main platform, moves from pay-as-you-go to an annual commitment and starts buying add-ons such as verification, fraud protection and branded calling. Ownership splits: product engineering owns the integration, growth or marketing owns marketing texts (often in its own tool), support owns voice and WhatsApp, and consent records start to need a named owner. My rough estimate is $20,000-500,000 a year. The push to the next step comes from the vetted brand's daily cap, a launch that needs a short code and its weeks-long wait, another country, a TCPA demand letter or a carrier fine, and the carrier fee rises of 2026 showing up on the invoice.

A mid-market company sends a few million to about 20 million texts a month, more than one brand's daily cap at T-Mobile can carry. So it runs several programs and one or more short codes, toll-free for support, hundreds of voice numbers with 911 records for office and softphone users (see Emergency calling), WhatsApp in some countries and branded calling for outbound sales or collections. It signs an annual or multi-year committed contract with one main platform, and often adds another vendor for one channel or country, such as a WhatsApp partner or a Latin American aggregator. Competing quotes and longer terms win discounts, but only on the platform's own share, since platforms such as Twilio pass carrier fees through at cost. A named messaging owner sits in platform engineering, legal or compliance reviews consent and quiet hours, and procurement joins the renewal. My rough estimate is $100,000 to $1 million a year; the median Twilio contract in one buyer-data platform's figures sits near the low end. Going public (which brings Auth+ identity checks on the 10DLC brand), regulated customers such as hospitals and banks, and more countries push it on.

An enterprise sends tens to hundreds of millions of messages a month across channels and countries: short codes per program, login codes, marketing and service texts, a contact center on voice, branded calling, WhatsApp at scale and RCS agents. It buys committed-use contracts from one or two platforms, and some large senders contract aggregators or carriers directly. At this size each channel's rules matter more than the per-message price. WhatsApp's volume tiers add up across the whole business portfolio and reset every month, Meta now bills in local currency in India and Brazil, and class actions go after big senders' consent records. A messaging platform team owns the vendors, marketing operations runs campaigns, the contact-center team runs voice, security owns login codes, legal owns consent, and procurement and vendor risk own the contract. My rough estimate is $1-10 million a year. More countries push it further, each with its own sender registry and rules.

A Fortune 500 company runs all of that in dozens of countries, sending billions of messages a year. It splits traffic across several platforms and aggregators on purpose, for resilience and price leverage, and some keep direct carrier connections. WhatsApp is a main service channel in Latin America and India, and branded RCS and branded calling are standard. A central platform or architecture team owns the vendors, brand and marketing own the WhatsApp and RCS experience, the CISO and fraud teams own authentication, legal and privacy own consent, and procurement runs the multi-vendor strategy. Some gatekeeper rules treat the biggest public brands differently: one provider says T-Mobile doesn't put Russell 3000 companies on its default cap for unvetted brands, and public companies already go through Auth+. My rough estimate is $5-50 million or more a year, with the largest consumer platforms well above that. There's no next size up; the pressure comes from regulators country by country.

What changes as a company grows (volume and spend are my rough estimates):