Editor’s note: Christian Henschel is the CTO and co-founder of adjust, a mobile app attribution and analytics company.
Cookies are the base tracking mechanism for most companies running online marketing initiatives, yet cookies are dying out rapidly because mobile usage is increasing and most mobile browsers don’t support third-party cookies. In fact, cookies simply don’t work across all mobile devices, and multi-device usage is no longer the way of the future; it’s happening now.
After announcing plans to move away from cookie tracking last year, Microsoft, Google and Facebook have developed their cookie alternatives, further eroding the existence of cookies as having a viable future for tracking. The big players like Facebook and Twitter are already using a hybrid model where the fundamental technology is user-based. In due course, these companies will be determining how to build ad models around this form of attribution. The transition to mobile dominance has happened so rapidly that brands and marketers are still playing catch-up in this post-cookie online marketing ecosystem.
The market is really moving towards the unique device Identifier for Advertising (IDFA) or device fingerprinting solution as the answer to tracking in a non-cookie-compliant world. Additionally, the thinking across mobile and platforms needs to shift towards a user-based attribution model. “People-based marketing” (i.e. user-based attribution) is when each user is attributed to their source and what they do in the app. This blend of technologies allows attribution with greater accuracy.
With the variety of technology required to track apps across all stores from the Apple App Store and Google Play to the Asian-market third-party stores, it has become necessary for publishers to blend tracking options and include device fingerprinting in the mix. Particularly as re-targeting platforms have entered mobile – AdRoll just joined the game – there are serious gains to be made from adopting cross-platform practices on this same basis, and in which case, cookies become obsolete.
How It Works
Device fingerprinting identifies the mobile device after an ad has been clicked on. A set of non-personal data, such as the operating system version, screen site, language, time and more are gathered. If the user installs and opens the app with an integrated SDK that has tracking abilities, the same process happens.
The attribution window for ID-based attribution is seven days across the board. Meaning, an install can be matched to a click on a source up to seven days. After those seven days, the install will be attributed to organic traffic. Fingerprinting allows publishers to match a click to an install lifespan of five hours (with 95 percent confidence). After that, data points may have changed to an extent that a certain attribution cannot be made anymore.
A blend of technologies for attribution is important to provide the most accurate results, especially when re-attribution and more advanced tracking is thrown into the mix. For example, although 90 percent of the time with Google Play the referrer is correctly transferred at the time of install, more matching that includes fingerprinting in the mix doesn’t hurt tracking accuracy – it improves it. Without fingerprinting, publishers lose the ability to track any third-party store, such as many stores in Asia that don’t support referrer passing as there are no device IDs available at the time of click (to undesirable results).
It was just a year ago that collectively the industry started addressing the notion of web-based cookies dying out in a few years. With the proliferation of mobile devices and cross-platform marketing, as well as the fingerprinting solutions now available in the market, the rate at which this transition is happening in the industry has rapidly increased to a point that we can truly say the cookie is dead, long live user-based attribution.