Bringing the Phone Number to the Cloud
With some work, the phone number can be made more dynamic and more intelligent -- before it is finally eclipsed.
In my previous No Jitter article, "WebRTC and Contextual Identity on the Web," I examined the potential role that phone numbers can play as universal identifiers for next-generation online voice and video services (e.g., WebRTC-based solutions). Although many speculate that the phone number will become obsolete (along with the dial pad) within the next few years, it is still ubiquitous and strongly anchored in today's telecommunications systems. However, it needs to be adapted to the ever-changing online world.
Even with the emergence of email addresses, Skype IDs, and Facebook addresses, the E.164-formatted number designated in the international public telecommunication numbering plans is still the most interoperable and universal identifier we use to date. Most people have one, and it is often linked to a viable amount of information about the owner, including sensitive information like a home address, billing info, government or healthcare IDs. Moreover, unlike online identifiers, the phone number is portable. In many situations, this means users have the ability to keep their identities consistent when changing carriers.
Why Fix It If It Ain't Broke?
If the phone number is as universal as I claim, why even bother arguing about its relevance or its design?
The ways we communicate have radically changed with the introduction of smartphones, yet the phone number hasn't evolved like its hardware counterpart has. You still have one connected to your smartphone, right? It is still relatively "dumb." Society's approach to the phone number has grown into a love-hate relationship.
For example, phone number portability does come with a few limitations. If you move to a new country, you need a new number and your identity essentially becomes fragmented over many numbers, carriers, and countries. Getting a phone number can be more difficult than obtaining a common online identifier, like a Facebook login, since doing so usually requires a physical address, too. Most importantly, remembering someone's phone number is more difficult to do than recalling first names or business names. When we want to call someone, we do not type in a phone number. Rather, we look up a name in our contact list.
It seems that the phone number has become more of an identifying key rather than a person's end address -- just like secure keys used in API authentications. Moreover, the structure and use of the phone number has become a mismatch with the way we want to be identified and the way we want to contact each other.
Though there have been many initiatives for bringing the phone number to the Internet such as SIP trunking, which is Voxbone's area of expertise, and the E.164 Number Mapping (ENUM) standard, the phone number is not bound to last very long with respect to our technological advances and Moore's Law. Smartphones are rapidly turning into smart devices of all kinds, and we are developing more advanced human-computer interactions that will not work using complex and obsolete identifiers. Yet, there is no doubt that the phone number is still on everyone's mind today. As the phone number heads to a darker future, work on some things can help make it more dynamic and more intelligent before it is finally eclipsed:
- Over-the-top (OTT) providers and mobile virtual network operators (MVNOs) bringing the phone number to the cloud. The phone call feature of your smartphone is really just another app, and services like WhatsApp, Skype, and Viber have taken a big chunk of that app market. Though the phone number is at the center of such services (identification and two-factor authentication), a strong division between its circuit-switched and packet-switched (IP) uses persists: Someone calling me on my Verizon phone number will not end up as a WhatsApp call. What if I could port my number to WhatsApp? Google's MVNO project -- Project Fi -- and undoubtedly the rise of embedded SIMs allow your mobile number to live in the cloud and, just like your email, work across all of your devices. It's just another sign-on. By bringing our phone numbers to the cloud, we would not be bound to an area anymore. Our phone numbers would therefore not need to match a particular format or area code. What would a phone number look like? Could it finally be prime time for Voxbone's iNum initiative?
- OAuth and the phone number as a sign-on - As OTTs and MVNOs like Google's Project Fi introduce the concept of signing on to cloud services using a phone number, the concept could be exported to the way we sign on to everything we do online. Just like signing in to streaming services with Facebook or Google, we could be signing in using our phone number. OTTs and MVNOs could provide their own OAuth or OpenID identity provision mechanisms with the phone number at their cores. Using the phone number for identity purposes is already familiar to the Web world, as service providers often use it for two-factor authentication and proof of identity. Implementing the phone number to OAuth sign-on mechanism would enable us to sign on to Web communication services, like WebRTC-based services, using our phone numbers. Furthermore, this would lead to the opportunity to make these WebRTC-based services more interoperable with traditional telephony. For starters, they would have a similar identification system.
- Opening Cloud-based Caller Name (CNAM) and third-party services integration - The development of CNAM databases and the ability to deliver their information over APIs has given rise to a multitude of new third-party services that can provide information about anyone who has a phone number. The phone number does hold a great deal of information about the person using it, and CNAM APIs such as Truecaller or EveryoneAPI, along with Home Location Register lookup services such as Twilio Lookup and Nexmo Verify, have allowed third-party services to provide contextual and unified communication services. For example, services like FullContact offer ways to gain social information about people through the use of their APIs, which include CNAM lookups. In the meantime, Facebook has launched Hello. It lets you search for anyone or any business on Facebook and make a call directly from the app. Google also offers Smart Caller ID on Android, which provides reverse lookup of the phone number calling to provide the name of the business entity to the person being called. Coupled with the ability to use the phone number for sign-on, CNAM over API would enable consumers to gain power in determining how much of their information third parties will have access to, should they decide to share it with them (just like they can via the Facebook or LinkedIn sign-on mechanisms). And, as noted above, it will allow for contextual and UC experiences.
Issues With Bringing Phone Numbers Online
Though bringing the phone number to the age of the cloud seems appealing, we will need to tackle some important issues. LinkedIn has stopped providing its users' phone numbers through API requests for obvious reasons. Spamming is a big concern, especially with the emergence of full browser communications systems like WebRTC and the opportunity to automate calls.
These concerns also spread to privacy and security topics. CNAM databases offer a centralized way for third-party services to provide information over which the consumer has no control. Unlike when we authenticate to a third-party service using Facebook or LinkedIn OAuth mechanisms, we are not able to control how much information third-party services will gain when performing a CNAM API request.
Are we ready to receive the same amount of spam on our phones that we receive over emai? Clearly, security and anti-spamming mechanisms will have to be thought through very carefully. Facebook is already tackling the trade-off between being able to be contacted by anyone while keeping them safe from spamming.
I'll leave you with a few questions: Though it is ubiquitous, should the phone number still be relevant today? Even if we were able to successfully port its use to the cloud, would that actually fix the problem we have with the phone number's stiffness? Would there be a more ubiquitous way to identify people both online and offline? Maybe the solution is to tackle the interoperability among the different identification silos instead of choosing one universal identifier. Initiatives like Matrix.org are dedicated to tackling this scenario by providing a framework that all service providers can implement to provide identity and hence service interoperability. Your thoughts?