Tuesday, July 27, 2010

This post originally appeared on ReadWriteWeb.

The interactive voice response (IVR) app was notably absent from the agenda at ReadWriteWeb's Mobile Summit in Mountain View, California. This is part of a larger trend: In the race toward capitalizing on the smartphone boom, application developers are focused on creating apps for smartphones, but very little attention is being given to creating apps for the voice network.

The phone call is no longer considered sexy. However, just like the human need for conversation, the phone call is never going to go away. People will always talk, and that conversation will continue to be a perennial part of any communication strategy. Building apps for the smartphone may be cool, but building apps for the voice network is essential.

Personalization to Stay Relevant

The nature of the phone call is rapidly evolving, especially as it relates to commerce. As businesses fight to stay relevant in the conversation - no pun intended - they see the value of using new data streams to access customers and utilizing existing streams effectively. They do this by using the telephone to send out a message that attracts customers.

There is a need to personalize the phone call. It's a direct response to the fact that people have little patience for the generic phone message, especially in this microblogging age. Unless it's apparent early on in the message that it was meant specifically for them, it's going to take a lot more to get the customer's attention. Not only does the phone call need to be personalized for each customer, but this personalization has to be able to scale with the business.

There are two aspects to personalization. The first is to unlock the information in disparate data sources, and the second is to interject this data into the conversation at the right time and in the right format. This has and always will be the dominion of the app developer with deep smarts in the business domain. These apps need to discern data quality, employ complex business rules and interject the data in real time into the conversation, all in an effort to make the conversation relevant.

This relevance can be in the form of traditional messaging, like personalized appointment reminders that give clients details of the appointment along with the ability to automatically reschedule, or it can be in newer methods, like in student education where key presses on the phone can be tied in real time to an online seminar.

No Longer a "Tool of the Big Boys"

Unlike the mobile device, these apps are hosted on the network, meaning it's up to the telecom provider to build the platform to support these apps. This is where the cloud-based IVR comes in, providing the full set of features and scale with very little up-front costs.

It's important to note that the technologies that make up the IVR developers palette have been around for decades; what's changing is the access to these technologies. For example, open standards like Voice XML, CCXML and technologies like text-to-speech (TTS) and speech-to-text have been available for a while, but until a couple of years ago, building an IVR that could customize a phone call was the dominion of large enterprise applications.

Building IVRs was a complex, multi-person endeavor, partly because of bloated protocols and high learning curves, and partly because of licensing and related costs. Not only was the barrier for entry high for developers, but the pricing of the technologies involved made the business case for a small development shop even harder to make.

To compound this problem even further, providing quality "copper-to-copper" termination at scale has always been an expensive proposition. It was the case that the only way to guarantee that 10 simultaneous calls would have the same call quality as 10,000 was to buy 10,000 phone lines worth of hardware, manage direct carrier relationships and keep dozens of servers idling, just in case the big flood came. Understandably, this made small- and medium-sized businesses even more wary of doing too much with their IVR other than what was available from cookie-cutter software. IVRs remained the tool of the big boys.

Cloud telephony is disrupting this old way of business. It has abstracted the complexity of telephony and the associated costs and provided an on-demand platform where anyone can create apps that are relevant to the conversation. Platforms provided by the likes of Adhearsion and CallFire are disrupting this status quo by creating an environment where the IVR resides in the cloud and can be invoked either by calling out to phones, or by consumers calling in to a hosted number.

"The Illusion of Simplicity"

The cloud solves the problem by creating the illusion of simplicity. The tasks of setting up trunk groups, interacting with carriers, and so on are taken care of by the service provider, who in turn provides open, on-demand APIs to connect, create and execute calls. In fact, many providers are ignoring VoiceXML and CCXML and are creating simplified standards that are more appealing to developers.

CallFireXML provides a reductionist XML spec that allows for great expressiveness by providing a simplified XML tag set. Simplification in cloud telephony means not having to worry about how the phone call is made; you simply tag your data, and the service provider takes care of the rest.

This trend has not gone unnoticed by the big boys. In fact, there's a mini-consolidation wave happening in the industry. The big players - the AT&Ts, Verizons, and the Level 3s of the world - are now on the prowl to dominate this space, either by innovation or by acquisition. Of course, their deep pockets make it easier to buy rather than build. BT's acquisition of Ribbit and Microsoft's acquisition of TellMe are perfect examples.

This is also a validation of the fact that the telephone call is not going away. As long as people want to talk, there will be a need for building apps that allow for person-to-person calls, and that's a great business to be in. What kind of hosted IVR apps would you like to build today?