No Analytics for You; How Children's Education Apps Could Suffer Under New Regulations

Actors often talk about how different it is to act in a movie vs. a live theater production. In film production, an actor can't measure his performance until months or years later when it finally makes it to the multiplex. In a Broadway show, there is instantaneous feedback. The actor knows immediately whether the audience laughs or cries at the right points. The theater production has information that can be used to make the show better the next night, and the next, and the next. The audience response is providing a form of analytics for the show's actors and director.

Mobile apps are somewhere between a movie and live theater. App makers can't watch over the shoulder of their users, but they can infer what they like from how users interact with the app (do they laugh, do they cry?). They do this through analytics services that aggregate usage data to provide reports explaining how customers use the app. Developers take that information to update apps so that the 1.1 version is better than 1.0, and so on. These services are very beneficial to small app developers (comprising 88% of the marketplace) because they do not have the thousands of dollars to spend on market research.

Take, for example, a popular mathematics app. The analytics service finds that students are having particular trouble with long division. The developer can then modify the program to better explain long division to the students. Without analytics, the students' difficulties with long division may never have been noticed and the developer would have been left operating in the dark when updating his or her app.

This process of continuous discovery and improvement is how successful apps are built, and analytics have been at the heart of that process. Unfortunately, a whole segment of app developers recently lost their ability to learn how their apps are used and, in turn, how to improve the educational experience for users.

Flurry Analytics, the leading provider of app analytics, recently decided to no longer allow its services to be used by apps for children. The company hasn't publicly indicated the reason behind this decision, but it comes right on the heels of the Federal Trade Commission's (FTC) efforts to update the Children's Online Privacy Protection Act (COPPA).

The FTC has proposed updates to existing child safety regulations, and has issued a "Notice of Proposed Rulemaking". Mostly, these changes represent sensible steps, including an explicit statement that mobile applications are covered, just like websites. However, the 122 page document goes beyond the basics and attempts to break new regulatory ground, and broad revisions like this one can have unforeseen consequences.

The proposed regulation generates considerable concern for educational app developers, the kind highlighted in a recent 60 Minutes special on "Apps for Autism". Under the new proposal, the definition of personally identifiable information (PII) would be modified to also include the device identifier, IP address, and screen handles -- elements that do not directly identify users and must be combined with other data in order to derive an identity. While this may sound innocuous, it could create problems for the collection of analytics and other tools developers use to improve their products.

The proposed rules seek to define as PII "an identifier that links the activities of a child across different websites or online services," and it allows for the use of this information only for the "support for the internal operations of the website or online service." Further explained, these operations are defined to include only "those activities necessary to maintain the technical functioning of the website."

Thus, while the information shared with Flurry is necessary to measure effectiveness and make improvements in the app, it is unclear whether it will be considered "support for the internal operations of the website or online service." Moreover, Flurry may want to use that information to improve their own product, and to provide information to others about what people are interested in and focusing on.

Flurry's withdrawal is very disappointing, but hard to fault. The public relations and litigation risks that would arise from one of its clients violating COPPA are hazards a startup like Flurry is eager to avoid. The company's decision underscores the real concern that many startups and small businesses face with the uncertainty and potential regulatory hurdles of the new definitions.

While it is unlikely that the FTC sought this outcome, it is an example of the unintended consequences that may force others to abandon creating children's educational programs. Given the amazing interactive educational apps that exist for tablets and smartphones, it will be a tremendous disappointment if our children lose this valuable learning resource.

COPPA was written to protect children from online predators and unsavory activity. It is hard to see how education apps fit under this rubric, but educators find themselves in the same space nonetheless. The FTC has already proven its enforcement measures have teeth, fining small business violators up to $50,000 and bigger companies several million dollars.

We hope that Flurry and other analytics services can find a way to continue providing valuable data to children's app makers, but this may be difficult without changes or clarifications to the FTC's rules. Analytic information is essential for developers to fine tune their products and make them compelling educational resources. As the FTC finalizes its proposed regulation, it should make the necessary amendments to ensure that its unintended consequences don't stunt the growth in one of the fastest growing sectors of our economy.