WheelsTV, Azuki and the problem of getting video onto phones in 2008

In 2008 a video producer who wanted to reach phones had a problem that no longer exists in the same form. Producing the video was the easy part. Getting it onto a handset — encoded correctly, delivered over a slow network, billed, and reachable by someone who wanted it — required a partner whose entire business was that plumbing. The WheelsTV and Azuki Systems arrangement was one instance of a pattern that defined the period.

The short version

Mobile video before app stores was a partnership business, because the hard parts were not editorial. Two things then dissolved the arrangement: app stores gave publishers a direct route to the handset, and adaptive streaming turned per-device encoding into a solved, commodity problem. The platform companies did not disappear — they were bought by infrastructure vendors and became features of larger delivery systems, which is exactly what happened to Azuki.

The two companies

WheelsTV, operating as Automotive Networks Corp., is a video production company specialising in automotive content — test drives, vehicle profiles and pre-owned vehicle reviews. Its output is short-form, catalogue-shaped and built to be syndicated rather than watched on a destination channel of its own. It is still operating, with a library covering model years from 2008 onward, distributed through automotive retail and content partners.

Azuki Systems was a Massachusetts software company working on the delivery side: getting video to devices, across networks, with the commercial and rights machinery attached. Its later product positioning — "TV Anywhere" — describes the shape of the problem well: take one source of video and make it playable on whatever the viewer is holding, with content protection and personalisation intact.

That split is the whole point. One company had the content and no route to the handset. The other had the route and no content.

What made mobile video hard in 2008

It is worth being concrete about the obstacles, because every one of them has since been either solved or absorbed into infrastructure that publishers no longer think about.

• Device fragmentation. Handsets differed in screen size, supported codecs, container formats and player software. Serving a catalogue often meant producing and storing many variants of the same file.
• Network conditions. Mobile data was slow, expensive and metered. Video had to be encoded down hard, and a stream that guessed wrong about available bandwidth simply failed.
• No adaptive delivery. Adaptive bitrate streaming — the technique that lets a player switch quality mid-playback — was not yet the default. Delivery decisions were made up front, not continuously.
• Distribution gatekeeping. Reaching a phone user frequently meant getting onto a carrier's deck or portal, which meant a commercial negotiation rather than publishing.
• Rights and protection. Anything with licensing attached needed protection and reporting that a small producer had no reason to build.

A specialist producer facing that list has two options: build a mobile engineering department, or partner with someone who already has one. For a company whose actual asset is automotive video, the second is obviously correct.

Why a narrow subject was a reasonable bet

The instinct behind a single-subject mobile channel was sound and remains sound. Someone standing on a forecourt deciding between two cars has a specific, immediate question, and a short video answering it is genuinely more useful than a general entertainment feed. Narrow subject matter also sidesteps the two hardest problems in general-interest video: rights costs and discovery.

What the era got wrong was not the content thesis. It was the assumption that the delivery arrangement would remain a durable, defensible position.

What dissolved the arrangement

The single most consequential item is the first. Adaptive bitrate streaming changed the shape of the problem rather than just making it cheaper: instead of deciding in advance what a device could handle, a publisher packages several quality levels once and lets the player choose continuously. The mechanics are covered in how streaming video actually reaches your TV.

Once that became standard, "we can get your video onto phones" stopped being a business and became a line item.

Where both companies ended up

Azuki followed the path typical of the delivery-platform generation. In February 2014, Ericsson announced an agreement to acquire the company, folding its adaptive bitrate and content protection technology into a television and media portfolio that had recently taken on Microsoft's Mediaroom. A capability that had been a standalone product in 2008 became a component of a vendor's broader delivery stack.

WheelsTV went the other way, and is arguably the more interesting outcome. It stayed a content company. Because its asset was a catalogue rather than a distribution mechanism, the commoditisation of delivery worked in its favour: it can now syndicate the same library through whichever channels exist, without owning any of them.

What still rhymes today

The 2008 arrangement looks archaic, but its underlying shape has not gone away — it has moved up a layer. A service today does not need a partner to encode for each handset, but it does need a different application for each television platform, because the platform still decides what is possible. That is the subject of smart TV and streaming platforms.

Narrow, single-subject channels also came back, in the form of free ad-supported streaming television. A FAST channel devoted to one topic, distributed through platforms it does not own, is recognisably the same idea as a single-subject mobile channel — running on infrastructure that finally makes it cheap.

Why this URL still exists

This address is used as a reference for a specific fact about an early mobile video partnership. Removing it would break that reference. Keeping it and writing something new and accurate serves the reader who follows the link, without reproducing an article the current publication has no rights to. See the archive for how preserved addresses are handled.