Two Predictions for New Search Ranking Factors

Time to Content

I thought the most interesting news to come out of Apple’s WWDC 2015 event was that Apple would be releasing a public search index and that content would be indexed via NSUserActivity states. The presentation on Search APIs offered insight into what Apple will consider when ranking apps.

Particular emphasis was placed on the importance of delivering users straight to content, without any interruptions (“no interstitials”).

“This is in fact so important that we measure and estimate the time from the tap to the content shown, and use it as a factor in ranking.”

Side note: included in the talk was a demo contrasting a well behaving app with a misbehaving app. The bad example invoked laughter and mocking applause, but notice how it’s actually completely normal by web standards.

In 2012 Google mentioned they would start penalizing sites for overdoing it with ads. More recently they’ve dropped hints suggesting app interstitials on mobile screens may be frowned upon, and just a few days ago the Google+ team shared some data along with an announcement that they would be retiring their own app interstitials out of usability concerns.

Prediction #1 is that Google will soon put a finer point on this, framing it in the same Time-to-Content terms that Apple is using, with particular attention paid to interstitials and overlay ads. The argument will be simple: a link is a promise and prioritizing other content over the fulfillment of the promise is a violation of trust.

Publishers will be forced to consider difficult questions given the popularity of interstitial ads. Marketers will have to work harder to build their email lists. But the good news is that the usability advocates who have been complaining about this stuff for years will now have a stronger argument—and user experience on the web will get a little better.

Excessive Third-Party Requests

The big problem with ad trackers is not that everyone’s being watched, it’s that everyone’s being watched without any degree of transparency or consent. When you click a link you have zero control or visibility into what, how, and why data is being shared. The actual incurred cost of your free content transaction is entirely obscured.

This is a huge problem for users, but it’s also a huge problem for publishers—because publishers also don’t know what data is being shared. The ecosystem of interconnected marketing and ad technology platforms is incredibly complex, is driven mainly by third parties beyond the publisher’s reach, and the publishers are just as confused by it as anyone.

How can Google help save publishers from themselves?

Prediction #2: Eventually Google will correlate excessive third-party requests with probable disregard for user privacy.

Hostnames are the basis for security on the web. They’re how we manage trust. When you pass resources through your domain you take responsibility for it. Why not extend this principle to cover the entire range of transactions with the user?

Cautious site owners may need to think about a third-party request budget. Best practices may develop around using CNAMEs to authorize approved vendors. Products may begin to compete on privacy benefits, citing limited or no third-party transactions.

Whatever the responses will be it seems inevitable that the problem of too much data sharing will come to a head, and that cross-domain HTTP requests may be a useful measure of the behavior we want to curb.

For more stuff like this, follow me on Twitter.