The telecommunications services business, worth about $1 trillion in annual, global sales, is starting to feel the effects of cloud computing, one of this era’s most important advances in high tech.
For the most part, the gear for telecom work comes from incumbents like Cisco or Avaya.
The more software-based alternatives for what is called “unified communications” are things like Microsoft’s Lync.
The telecom gear business has not been touched by the crushing economics of software delivered via cloud computing, the kind of low-price, high-versatility product that has caused so many problems for companies making business software, and for many traditional high-cost computer hardware makers.
But that’s starting to change.
Twilio is a cloud-based business that produces the means to create any number of communications services, primarily through servers it rents at Amazon Web Services. According to Jeff Lawson, the company’s co-founder and chief executive officer, the company has provisioned “millions and millions” of phone numbers in over 40 countries. Revenue last year was about $50 million, more than double the year before, he said.
Twilio generally sits in the background while larger companies use its technology to quickly create things like customer call centers, collaborative communications networks or text and image-messaging functions.
Another company, called Tropo, also allows for faster deployment of communications applications, while others create cloud-based call centers. Plivo may be Twilio’s most direct competitor, with cloud-based voice services that can be added to apps. Netflix, among others, uses Plivo for its internal phone system.
When Uber texts clients to tell them their cab is five minutes away, or eHarmony links two strangers interested in dating, Twilio is in the background.
“Our average selling price is one penny a minute, versus multimillion-dollar boxes from the legacy equipment providers,” Mr. Lawson said.
As with much of the cloud business, the initial attraction is making things cheaper and faster to do. Mr. Lawson said Twilio beats the incumbents by 90 percent or more on both counts, but it’s hard to judge that claim independently (though even 50 percent would be impressive). A Cisco spokesman said the two products are not directly comparable, since Cisco’s product can do many more things.
What usually happens next with these cloud products, however, is where it gets more interesting.
Software developers find new things they can do, once factors like speed and cost become less limiting. Think of cognitive computing and supercomputing as a service, or Snapchat, the disappearing image-sharing business hosted on Google’s computers.
Put that capability on cloud-based applications, and you are likely to get a large number of new communications services.
Billing applications can have a feature that puts you directly in touch with a person if you want to dispute a charge, and the person on the other end of the call instantly sees the same charge. People at a sports event have had their season tickets scanned, and by linking that information to their mobile numbers, promotions for, perhaps, two beers for the price of one might be texted to their phones during the game. (How creepy this might seem is for society to work out.)
“Every company wants to do things with less friction and more customer awareness,” Mr. Lawson said. “Communications companies have historically worked on a development cycle of decades – the cloud takes that to weeks.”
Down the road, the most likely competition for Twilio is Google, Amazon or another cloud provider that wants to offer its own communications applications. Google has email, texting and video, but so far these are pretty diffuse offerings, and Google’s cloud service has offered ways for developers to tie into Twilio.
Amazon Web Services, which seems to be in an ever-increasing number of services and features, so far seems content to have Twilio inside its system, as one more thing developers on Amazon can use. They could partner with it, build their own version or copy it. What is unlikely to happen is for the phenomenon to stop.