Beyond Content: Will Context Someday Reign?

Image by Patrick George/Alamy

What mobile permissions would you be willing to give news organizations if they were able to hand you personalized experiences in return?

That’s the question we posed at SRCCON in late July. (And may we just say Portland is lovely and thank you to the folks at Open News for having us!). At SRCCON, a conference for developers, designers, data analysts and others working in news, the lab facilitated a session about what mobile permissions readers might be willing to offer news organizations, and what experience we might be able to offer them if granted access to additional mobile data or functionality. We’re interested in this topic as we start to explore our third area of focus: contextual delivery, which we define as delivering news that adapts to a user’s needs and preferences throughout their day.

Let’s back up a second: What do we mean by mobile permissions?

Simply put: mobile devices gather and store information about us as we go about our day. As mobile users, we voluntarily grant apps and services permission to access this information in exchange for personalized experiences.

While all that information is gathered and stored, apps and services aren’t able to use it without your permission. A few examples of those uses: reading the user’s private information (such as contacts or emails), monitoring the device’s network connectivity, setting an alarm, and so on. Uber and Lyft work by knowing our precise location, Instagram recommends who to follow when we give it access to our contacts list, and Pokémon Go overlays the Pokémon onto the world around us by gaining access to our mobile cameras.

Such services accomplish these tasks by asking the user for needed permissions. That’s why being asked for access to location or even the camera is a common experience for smartphone users looking for things like weather information or trying to post to social media on apps or web. Other personal information that can be used by a service once permission is granted includes the ability to send texts on your behalf and access first and second degree connections.

Of course, as in the above examples — the permissions granted are often the cost of admission. The service runs on the expectation of the data being provided. The degrees to which apps require permissions vary, but, from the user’s perspective, agreeing to share this information either enables or significantly supplements the experience.

A service’s request for permission may also include a request for data from sensors, both from the mobile device as well as from devices that sync up with smartphones, like smartwatches, Fitbits, and other wearables. A few examples: the linear acceleration sensor and the gravity sensor, both of which are used to determine motion, and body-related data like heart rate or steps taken which can be collected by sensors in a phone or wearable.

A slide from the Mobile Innovation Lab’s session at SRCCON. View the full deck here. Google uses “normal” and “dangerous” to separate which permissions need to be explicitly asked for. “Dangerous” permissions the user needs to specifically agree to, while normal settings are allowed automatically.

In exploring contextual delivery, we’re starting with the premise that news organizations could be taking advantage of these new types of permissions to deliver news experiences geared to a user’s location, content preferences and even the time of day. However news organizations will need to be especially careful to respect readers’ privacy, be transparent about the reasons we’re asking for their information and clear about what we’ll do with it.

And it’s a fine line between offering valuable experiences and being invasive. Plenty of those who came to our session had their own experiences of being creeped out by digital services. For example: one attendee told us, the day after her wedding, the popular wedding site “The Knot” directed her to “The Nest,” their sister site about home decorating, babies and celebrating anniversaries…and she began to see a lot more baby-related advertisements around the internet. One reporter told us about having a source’s real identity revealed to her by LinkedIn through the “People you may know” section — the source had previously asked to go by a different identity, even with the reporter.

At SRCCON, we asked our group to explore that line between convenience and creepiness. How much information would they be willing to give up about themselves for a personalized news experience? As journalists and technologists, what products or features would they design with access to personal information?

The Workshop

We started by dividing the crowd into six small working groups who were given a set of cards with the names of 15 possible permissions or sensor data that they could access. Each team was also assigned a primary permission or sensor that their new product had to leverage. We told each group they could use any combination of permissions or sensor data they wanted to personalize a new product, and gave them 20 minutes to brainstorm.

The groups came up with some pretty interesting ideas.

Here’s a selection of some of what they came up with and the permissions involved:

With access to your contact list…

  • Relationship-based newsletter subscription offers, i.e. “Your friend gets this newsletter, do you want it?”
  • A news bot with access to SMS that would interact with a news organization and ask questions (Contacts + SMS)
  • A service that suggests you “call your mother” and intervene if she is reading or has shared content you think is terrible

With access to the accelerometer…

  • Podcast recommendations based on your physical state, including when you are awake, active, or maybe exercising (Body Sensor + Accelerometer)

With access to location…

  • A feature that would automatically activate your phone’s camera if you’re in the area of breaking news, in order to capture the surrounding events (Location + Camera)

With access to the calendar…

  • A service that suggests events to attend or things to do based on where your calendar says you are, or where you’re going
  • Provide up-to-date dossiers on people who you will be meeting with, timed to arrive just before you meet them (Calendar + Contacts)

Many of these ideas appear to veer into creepy territory, which we encouraged in the session as a thought exercise. And we’re not planning to pursue any just now (if ever). But they shed light on inherent tension between the possibilities for news and what is organizationally appropriate.

Mobile opens up opportunities for news organizations to be more targeted and personalized in their approach to coverage and delivery, but if users feel that their privacy has been transgressed by how we use this information, then we risk losing some or all of the trust we need in order to maintain credibility with readers.

What we aimed to do at SRCCON was start a conversation about new ways to take advantage of mobile functionality to create more contextually relevant interactions with our readers. What we are planning to do next is further explore what experiences news consumers would value and what sharing sensitive information based on their mobile habits would mean to them. From knowing a user’s daily routine and serving a well-timed podcast, to recognizing when someone is awake and triggering an up-to-the-moment version of a morning newsletter, the hope for our exploration of contextual delivery is that we can get a sense of how we can provide news that users want to know, when they want to know it.

Questions, suggestions or observations? Drop us a note: innovationlab@theguardian.com

The Guardian Mobile Innovation Lab operates with the generous support of the John S. and James L. Knight Foundation.