Skip to content

Dead End: MMS, the Picture Message with a Toll Booth

Abstract

The Multimedia Messaging Service was the mobile industry’s plan to repeat the most profitable accident in its history. SMS (a 160-character afterthought bolted onto GSM) had become a money-printing machine, billed per message at prices wildly out of proportion to the bytes transmitted. MMS, standardized around 2001 and launched commercially in 2002, was supposed to do the same for photos: every camera-phone snapshot a billable event. It never happened. Punitive per-message prices, tiny size limits, and chronic incompatibility kept usage marginal, and when smartphones with flat-rate data arrived, WhatsApp and iMessage did the same job for free. MMS then outlived its own relevance by a decade and a half: Vodafone Germany switched it off in early 2023, and Deutsche Telekom, O2, and 1&1 followed at the end of June 2026, one of the rare dead ends with an official date of death.

The SMS Goldmine

To understand MMS, start with the business it was meant to extend. SMS had been added to the GSM standard as a minor signaling feature (160 characters squeezed into spare network capacity) and had turned, to the industry’s surprise, into a cultural phenomenon and a revenue miracle. By the early 2000s, carriers worldwide were charging on the order of ten cents and more per message for a service whose marginal cost was effectively zero. Per byte transmitted, SMS was among the most expensive data services ever sold, and customers paid happily, billions of times a day.

At the same moment, cameras were arriving in phones: Japan’s carriers had launched photo messaging (“Sha-Mail”) in 2000 and proved people would pay to send pictures. The European and American industry’s conclusion was obvious and, in hindsight, fatal: build a standard that makes every photo a message, and bill it like SMS, only higher. As one later summary put it, MMS was built on SMS’s model precisely so that “service providers [could] collect a fee every time anyone snaps a photo.”

The Standard: A Message That Was Secretly a Download

Technically, MMS was a hybrid, and the hybrid shows what the designers cared about. Specified by 3GPP together with the WAP Forum (and later maintained by the Open Mobile Alliance), an MMS did not actually travel like a text message. The multimedia content was uploaded to a carrier server and fetched by the recipient’s phone over a data connection using the WAP protocol stack; only the notification traveled the SMS path. An MMS was, in plain terms, a small file download dressed up as a message, dressed up that way so it could be billed that way, per item rather than per byte.

The dressing came with corsets. First-generation MMS was capped at 50 kilobytes; later revisions recommended 300 kB (MMS 1.2) and 600 kB (MMS 1.3), limits that shrank photos to thumbnails of themselves just as phone cameras improved past two megapixels. “Content adaptation”, transcoding messages for whatever the receiving handset could display, worked unevenly across brands, so a slideshow composed on a Nokia might arrive scrambled on a Motorola, or not at all. Recipients with non-MMS phones got a text message with a web link and a password instead of a picture. Every phone needed correct carrier-specific configuration settings before its first MMS would move.

Launch: CeBIT, Free Trials, and High Hopes

The hardware arrived first: Ericsson showed the T68 at CeBIT 2001, the color-display handset that, in its Sony Ericsson T68i revision, became the standard-bearer for the new picture messaging. Commercial services followed in spring 2002: Vodafone Germany launched MMS in April 2002 (free at first, to seed the habit) with Deutsche Telekom following in mid-2002 and other European carriers close behind; Verizon brought MMS to the US in July 2003.

The free-trial phase was the high point. Once billing began, the proposition was stark: a single squashed, 300-kilobyte photo cost several times the price of an SMS at a time when the same customer was starting to carry a phone that could email the same photo, or soon, send it over the open internet. Usage settled at a small fraction of SMS volumes in most of Europe and stayed there. The one major exception proved the diagnosis rather than the product: in the United States, where carriers folded MMS into unlimited messaging bundles (removing the per-picture toll) traffic actually grew into the tens of billions of messages as late as 2013. Where MMS was priced as its designers intended, it failed; where it succeeded, the pricing model that motivated it had already been abandoned.

The Kill: Flat Rates and the Free Photo

What ended MMS was not a better multimedia message but the disappearance of the category. The iPhone (2007), Android, and data flat rates turned the photo back into what it had always technically been, a file moving over the internet. WhatsApp (2009) sent pictures, videos, and group messages over the data connection at no per-message cost; iMessage (2011) silently diverted Apple-to-Apple traffic away from carrier messaging altogether. Against apps with no toll booth, no 300 kB corset, and no transcoding lottery, MMS had nothing to offer. Usage in Europe collapsed during the 2010s to statistical noise; the messaging story continues in Instant Messaging: From ICQ to Signal.

The infrastructure lingered on, unloved, for years, carriers kept paying licensing and operating costs for a service almost nobody used. Then the plugs were pulled, market by market: Swiss carriers shut MMS down in 2023 and 2024; Vodafone Germany ended the service in early 2023; Deutsche Telekom, O2, and 1&1 switched off MMS at the end of June 2026, citing license fees, operating costs, and negligible usage. (Austria’s A1, as of mid-2026, had no shutdown plans, dead ends close at different speeds.) The industry’s designated successor, RCS, is pointedly not a descendant of MMS: it is an IP-based messaging standard in the mold of the apps that won, and even Apple adopted it in 2024.

The Lesson

MMS is the purest case study in the vault of a technology designed around a billing model rather than a user need. The engineering worked; photos were delivered. But every design decision (the WAP delivery detour, the size caps, the per-message tariff) served the goal of making multimedia meterable, and every one of those decisions made the service worse than the plain internet that was arriving anyway. SMS had been a goldmine because it was accidentally cheap to run and customers had no alternative. MMS tried to engineer the same scarcity deliberately, in the exact years the alternative was being built. The carriers got their toll booth; the traffic took the free road.

📚 Sources