Note: This was originally posted on VentureBeat on 4/30/14, available here.
Imagine a phone that knows your tastes, your preferences, and desires, and patterns of behavior. This data about you — personal context — is combined in real-time with the phone’s environment (where it is, adjacent businesses, people, and events; weather, temperature, and pressure; ambient light, radio waves, and gradiometry) to create a personalized mobile experience.
This combination of the historical, personal context with the sensor-driven, real-time device context provides the data needed to anticipate your requirements, surface items of interest, and (perhaps most importantly) filter the signal from the noise in your information stream.
This learning-through-context is the basis of Anticipatory Computing and ultimately the way to a genuinely personalized mobile device.
More briefly: context allows the Who to inform the Where and the What.
However, while this scenario is certainly viable, we’re not there yet: personalization in mobile is confined largely to home screen optimization. The mobile phone is the most intimate, personal technological creation ever developed, yet your device still acts and responds like one that knows nothing about you.
I use the word ‘device’ generically here, but really apps (and the OS) make this happen. While most app publishers want to provide you with a personalized mobile experience, few are positioned to deliver. This is in-part due to the fact that they have little idea who the consumer is, and — much more importantly — have difficulty discovering. Fortunately this is changing, but we can point to four critical reasons for this unintentional ignorance:
User Disintermediation: Google and Apple have interdicted the Consumer. It’s their world and their app store; we just play in it.
App Architecture: apps operate independently; no data is shared programmatically between them, and an app learns only how you engage with it, and no other (data can of course be routed between apps at the platform level).
Privacy Protection: no mechanism beyond general prudence exists yet for users to control how their data is shared with apps.
Monetization Mentality: an advertizing infrastructure exists now to monetize your personal data, but not necessarily to improve your consumer experience; targeted content now is rarely for your benefit.
These four items conspire to hold back the development of personalized mobile experience, but there are ways around them. We are on the edge of an exciting sea change of Personalization in Mobile, and it’s largely due to data.
Many apps are doing an end-run around these blockers, usually with some kind of gateway that delivers a service while learning more about you (‘search’ in its most basic form also fits this definition). The most obvious way to do this is via a personal assistant such as Siri, or Microsoft’s recently released Cortana. A less obvious example of the platform circumvention is Emu, which bills itself as a mobile assistant, but at its core is an app that creates context around texts by tapping into your extant comms stream, enriching the content, and learning about you in the process.
Here, the focal point between the consumer and the device lies on the platform, routing in-part around Google and Apple’s control. Are these experiences personal? Not really, not yet; but they are positioned to be so by learning about you while enriching your experience — a virtuous circle of context.
Personalization, of course, is all about you.
But personalization is also about the publisher and developer: a hoard of well-rounded user profiles is gold for any app publisher. While large publishers will create Cortana, Siri, and Google Now to engage their userbase through OS integration and hooks into their broader ecosystems, few others have a seat at this table, so they are learning that they must look outside their own data to understand users better.
These smaller developers will create enrichment gateways similar to Emu. Others will look for data ‘hooks’ where they can tie their data into a much wider stream of content to learn more about the user. Location is of course the most ubiquitous hook. Location allows companies like Songza to customize content to the local weather, or to suggest alternative experiences and learn more about the user’s habits.
At Factual we see Location as a form of foreign key, a mechanism through which the widest possible array of additional data can be joined to obtain a more holistic understanding of users. Factual uses Location to build anonymous geographic, demographic, behavioral, and commercial profiles, which are employed by our partners to serve more relevant, personal content. Location is no longer about just the Where: it now allows all developers and publishers to learn about the Who and the What.