We are witnessing a watershed moment on the Internet — with Apple’s content-blocker announcement, the way we see and understand our users on the web is going to profoundly shift.
It doesn’t seem like a very big deal. Adblockers have existed on desktop browsers for years, and products like Google Analytics have still become the industry standard for measuring and monitoring websites. But this is all about to change — individual products and categories will be massively disrupted by a simple technical shift in mobile Safari, and organizations with a web presence will have to adapt or risk being put at a competitive disadvantage.
Why is this the moment? Simply put, the legacy system for web tracking is already broken. It fails to accurately capture web engagement due to existing adblockers, and adding mobile content blockers will further skew the data. It fails on single-page web applications, because it is built around pageloads. It fails more severely in apps — where 90 percent of the action is — for the same reason, and because the multi-stack data is often poorly integrated. The biggest way legacy systems have failed, however, is by siloing data — putting key data about your customers in an inaccessible place, where integrating data and doing automation is impossible without costly and brittle API integrations.
Mobile-content blockers are the final nail in the coffin. They are likely to be much more compelling for users than desktop adblockers — bandwidth savings and faster pageloads really matter on mobile. The privacy gain is just a perk. With the Internet shifting inexorably to smartphones, losing visibility into iOS users — the users who spend the most money and time online — will skew the data and render it unreliable.
It won’t happen on Day One. But adoption of similar technologies on Android will rise with iOS adoption, and users will be even more likely to adopt desktop blockers as they become familiar with these technologies. Browsers may even build them in.
It’s Not Just Publishers
This technology will deeply affect digital advertising and attribution, which will in turn affect publishers and other companies with ad-supported revenue models. But that’s just the beginning.
Inspect any e-commerce or B2B website and witness all the cookies and requests (often made by Google Tag Manager). These networks (and tag managers), will be dramatically impacted — along with all the creative agencies and marketers who work with them. More profoundly, this will affect the companies that sell on these platforms — how will e-commerce companies reliably retarget users?
Reducing bandwidth and cleaning up the tremendous amount of cruft on mobile are worthy goals.
The most technologically savvy users will be much harder to reach; in essence, there will be growing adverse selection on ad networks. Decaying attribution models will reduce confidence in ROI, making ad programs — and any marketing campaigns — harder to quantify with confidence and harder to justify concretely.
At the most basic level, understanding user behavior on the web will become more challenging. Analysts who try to marry data in Google Analytics with data in internal systems already know that a meaningful schism nearly always exists between the two, and this is only going to get worse as more data is withheld from third-party systems. Not having good data will render many product and marketing teams blind — unable to see what’s really happening with their users, and unable to react intelligently.
What’s Next? The Burden Will Shift
Just as the app ecosystem created work, chaos and change, the shift in web analytics will not be easy.
But this presents some fundamental problems. First, only technically savvy developers will be able to confidently adjust server-side code. For most webmasters, they will rely on their hosts or on WordPress plug-ins to do this work for them. Ultimately, this will create more friction and delays around deploying these solutions, reducing their appeal. For technically savvy developers, the question will arise, “If I’m already deploying client- and server-side code, why use a third-party system at all?” Open source analytics solutions will gain traction, and more organizations will opt to do more of their analytics and intelligence in-house.
The other issue is scalability. Most web properties today are optimized to serve pages, not to handle and process an influx of analytics data. The new server-side solutions must be simple, reliable and cohesive. If we add dozens of requests and pixels without thinking about performance, as is the case today, we’ll end up flooding our own servers, adding costs and potential latency. The winning solutions will marry front-end data to server-side handlers in a single, well-designed pipeline, then export the data to data warehouses (and third-party services) from the server, probably in batches.
During this transition, many organizations will fail to act and will rely on imperfect data that is getting steadily worse. Numbers won’t add up and biases will lead to poor intelligence and decision making. It will be a chaotic time for developers, analysts and marketers, and there will be a bifurcation between sophisticated teams that adopt new tools, and those who choose to stick with the old.
We’re Due For A Revolution
Albeit chaotic, this shift will open a major market opportunity to disrupt existing players by going beyond the content-blocking issue to solve deeper issues around web tracking — marrying front-end data with multi-stack internal data, accounting for a web experience that is no longer centered around pageloads, but instead around events and sessions on the web and in apps, and making first-party, owned data a reality for more organizations.
The investments companies are making into marketing and product will demand solutions that really work, and new tools will evolve to fill the void — and, because of the increasing complexity of these tools, a new class of developer-analysts with deep measurement expertise will crop up and will be in high demand.
What should the ideal solution look like? Given that both front- and back-end stacks are diverse, the system will need to be highly flexible. One exciting guidepost for where things could be going is third-party bug tracking — services like Bugsnag have libraries in virtually every language that can operate at the server level and REST-fully interact to stream data about bugs and exceptions — analytics could work the same way.
Hopefully, however, organizations use this opportunity to claim their data for themselves, and open-source tools gain traction that help stream data into first-party data warehouses like Redshift, where it can be joined with internal data and used more flexibly. The community should come together to define a common schema that makes sense for Web 2.0 applications, and third-party BI services like Looker should sit atop this owned data instead of siloing it in inaccessible places as Google Analytics does today.
If this happens, we’ll be trading in an easy-but-broken system of web analytics for a much richer, more complex system capable of delivering much deeper insights. This will require a significant investment and more specialized skills — but the payoff will be better intelligence — and, ultimately, better products, user experiences and business outcomes.
Ultimately, It’s Great For Users
Contrary to claims about content blockers being Apple’s secret scheme to drive more activity into native news readers, the bottom line is that this change is great for users. Reducing bandwidth and cleaning up the tremendous amount of cruft on mobile are worthy goals. Some of this bandwidth will undoubtedly be shifted toward the origin, as described above, but overall, the experience should be better for mobile Safari users (and, ultimately, everyone else as we clean things up across the web).
Organizations with a web presence will have to adapt or risk being put at a competitive disadvantage.
There is also the privacy consideration — removing third-party cookies should reduce the embarrassing ads about items we abandoned in online shopping carts that follow us all over the web. But overall, the impact on online privacy will likely be muted as companies find other ways to share data — most notably, by simply sharing email lists. Facebook has already brought this approach into the mainstream with Custom Audience targeting. Universal IDs and IP reconciliation will also become more widespread and important to marketers and content platforms.
It’s Also Great For The Web
Apple is forcing us to rethink the Internet — by shifting us to mobile, shifting us into apps and by challenging the status quo of how we measure, track and target on the web, starting with mobile Safari. Just as the app ecosystem created work, chaos and change, the shift in web analytics will not be easy. But it will also create incredible opportunities and new winners. It will open the door to better, more flexible mechanisms for analyzing and understanding our users and products across the ecosystem. It will force us to be better — more integrated, more intentional and more knowledgeable.
It’s about time.