Apple’s latest Intelligent Tracking Prevention (ITP) update aims to increase user privacy by preventing unwanted tracking – but could have a major impact on digital marketing, particularly when it comes to technology such as analytics and A/B testing tools. Jean-Noël Rivasseau, CTO and Founder, Kameleoon, explains ITP, how it works, the impact it has on digital marketing tech and how it can potentially be overcome.
Successful digital marketing is driven by data, which should be underpinning every decision you make and every change to the user experience. But what if you can’t trust your data? This is a very real fear since Apple’s latest Intelligent Tracking Prevention (ITP) update, which threatens to have a major impact on digital marketing technology in general, and A/B testing in particular.
What is Apple ITP?
Apple has made a clear commitment to safeguarding the privacy of its users, and Apple ITP supports this by preventing unwanted tracking when using the Safari browser. As part of this the latest version, ITP 2.3, limits all JavaScript based cookies to a seven day lifespan.
The main issue for digital marketers is that almost all the front-end based technology platforms that they use (including the likes of Google Analytics), rely on these cookies to identify a specific visitor.
Essentially this means someone making a first visit to your site on a Monday using Safari, and then returning on the following Tuesday (i.e. more than seven days later), would be seen as a completely new visitor by these tools. Thus, the new visitor metric for Safari traffic will appear much higher than it actually is. Many other e-commerce metrics, such as time to buy are also completely distorted. To sum up, it’s now hard to trust any data from a cookie-based analytics solutions!
The impact on digital marketers
Given that Safari is the default browser on all Apple iPhones, it removes confidence in your user data and therefore the decisions you make based on it.
There are even more severe problems for digital marketers who use A/B testing or personalisation platforms to make any key decisions for their business. When running an experiment, a variation (such as a different home page) is usually chosen for a visitor. To ensure that this is remembered for when the same visitor returns to the website, most front-end optimisation platforms store that information (the association between experiment and variation) in a cookie. With ITP, a visitor returning after seven days runs the risk of seeing a completely different variation, meaning A/B tests simply do not provide reliable results.
Overcoming the drawbacks of ITP
So clearly digital marketing tech vendors need to adapt to overcome the challenge that ITP brings and some have already taken specific actions to solve it. While there are several options available, such as setting cookies as server-set HTTP cookies, we will focus on the one that we believe enables digital marketers to be confident of the data they are looking at, while requiring a minimum investment: embracing Local Storage.
Local Storage, a standard web technology supported by virtually all browsers, is basically a data store in the browser, so it can handle data in the same way that cookies do. This makes it a good candidate to replace cookies, as ITP has fewer restrictions on it than for cookies. Prior to 2.3, ITP ignored Local Storage entirely. With version 2.3, ITP only deletes Local Storage if the user is navigated to your website with a link decoration (which usually makes up only a small percentage of the overall traffic and does not disturb A/B experiments unduly).
There’s one major drawback however – it is normally limited to a single exact subdomain. So if you host an e-commerce site at https://www.mygreatshop.com but also use https://secure.mygreatshop.com within the user journey, your analytics solution will report two different visitors when a complete purchase occurs, one seen on each domain.
When it comes to A/B testing this leads to significant issues. For example, if you run an experiment modifying the navigation menu, with two variations, the same user could very well be exposed to the first variant on the main site, but to the second one down the funnel! This leads to a disastrous user experience, and of course makes the results of the experiment completely invalid.
The only way round this is to implement cross-domain Local Storage, which is more complex and time-consuming to program for vendors, but means visitors are recognised correctly, no matter how many subdomains they cover in the user journey.
With growing concerns from users over privacy, other browsers are likely to follow Apple’s lead and introduce similar tracking technology. That means that digital marketers need to ensure that they are working with the right vendors to deliver accurate, reliable results in an increasingly cookie-free world.
By Jean-Noël Rivasseau