By Steve Wunker
In the past 30 years, AT&T has undergone several transformational shifts. Government regulators broke up the monolithic “Ma Bell” in 1984, and dozens of successor companies sought to become top dog in the telecom food chain.
It turned out that the old, integrated model had many merits, and dealmakers thrived during the tortuous process of putting Humpty-Dumpty back together again. Today, AT&T is again a provider of complete telecom services, albeit lacking the hardware business it once dominated.
Recently the company announced a further step toward integration, opening three “Innovation Centers” around the world. The company says these centers will enable it “to work directly with device makers, application developers and network equipment providers to expedite development of an ecosystem of mobile and wired broadband services and capabilities for consumers and business users.”
Why should AT&T, one of the world’s largest telecom companies, need to create these centers? After all, vendors are eager to work with such a dominant firm. An answer lies in a pathbreaking 2001 article by Harvard Business School Professor Clayton Christensen, titled “Skate to Where the Money Will Be.” Christensen argued that industries almost always get their start through supplying integrated offerings to customers, as that is the only way that companies can get performance to be good enough to meet basic market demands. As industries mature and those performance needs become increasingly sated, customers start to value other things such as customization, speed, and price. This sets the stage for modular solutions to emerge. The PC industry provides an excellent illustration, with its shift from the integrated IBM model to the highly modular ecosystem of today that has different firms provide operating systems, microprocessors, services and so on.
It turns out that government pushed telecom into the modular world too early. Major shifts in the technical architecture of telecom networks were difficult to engineer when key players had diverging interests. It was also tough to change customer behavior when the network “owned” the customer financially, even as device makers created the machines which could dictate most of the customer experience. AT&T and Apple tried a different approach with the launch of the iPhone, tightly integrating the device and the network, but that task has certainly proven challenging.
I have had my own frustrating experiences in this world. As CEO of Brainstorm, a middleware developer for mobile networks, I worked with one of the largest European operators a decade ago to launch their picture messaging services. These services worked differently on the dozens of handsets the operator supported, and network equipment makers were inconsistent in how they handled the messaging protocol. The operator tasked my firm with harmonizing this highly diverse, constantly changing cast of actors. The result was no less than 50 changes in our software platform as the industry rapidly evolved. This is no way to orchestrate the introduction of new customer behaviors.
AT&T’s Innovation Centers seem a promising step toward uniting the right players under one roof. The question is whether these people will have the autonomy needed to create truly integrated solutions. At Samsung, innovators have thrived in its Value Innovation Program Center, bringing together key players in device development, but fundamental to their success has been Samsung’s ownership over a broad collection of assets from chip design to display manufacture. Will AT&T’s innovators be empowered, for instance, to tweak an operating system or communications protocol? Even more critically, will they have the freedom to suggest innovations in the business model? The current business models in this industry are rapidly heading toward commoditization, and one of the brightest hopes for new offerings has to be a game-changing approach to monetizing the ubiquity of telecom throughout customers’ lives. Otherwise, AT&T’s innovations may become marginal differentiators, not the performance breakthroughs that justify this integrated approach.