The Difference in Tracking for Mobile Web and Apps

Despite being for the same platform, mobile web and mobile applications work in completely different ways, especially when it comes to the difference in tracking for mobile web and apps. In this article, we’ll be explaining what differences there are so that you can get a greater understanding of where each platform excels and where each platform’s pitfalls are.

Tracking for Apps

Like online web tracking, tracking for mobile apps requires an identifier – many ad networks use various identifiers to help track individual users across applications, but due to privacy concerns, everything still has hiccups. Using cookies doesn’t work as well as it does on a desktop – they can be used somewhat, but tracking users through different applications is very difficult because of the fragmentation between different devices, apps and techniques used for tracking.

Identifiers Used for App Tracking

When it comes to the two most popular mobile operating systems, Android and iOS, Android_ID and UDID were often used as identifiers on an OS level, but other systems are available on Android too, including a mobile’s IMEI number, which is unique to each device. The MAC address can be accessed on both iOS and Android, and this is a much more widely used option. Ad networks can also create unique statistical-based IDs that are based upon certain information such as IP address, device type, operating system, OS version and so forth.

These identifiers are useful because they help ad networks to track individuals across multiple different applications. This means ads can show in different applications that are targeted to each user, but tracking a user when an ad is clicked is the difficult part when it comes to app tracking.

It becomes a problem when an ad network attempts to track activity via different sandboxes that are completely separate from each other. For example, if a user clicks on an ad in their app, then gets taken to the app store, it’s difficult for an ad network to detect what exactly happens when the user visits the app store. There are a few methods that are used here.

One method is called device fingerprinting. It’s fairly straight forward – if one device profile clicks an ad that leads to a download link, and then two minutes later a user with a very similar device profile then connects to the app that was being advertised, there’s a large probability it was the same device and the same user – for the most part this strategy is very accurate.

The Constant Struggle with App Tracking

There are two constant powers in the struggle for app tracking – privacy and ad networks looking to improve their tracking across different environments or applications. Because there are a lot of privacy concerns behind what ad networks are doing to track users across different environments on a mobile device, manufacturers are pushed to make their privacy regulations stricter.

A great example of this is how ad networks have used UDID as an identifier for iOS in the past. Previously, UDID was a great unique identifier that allowed ad networks to track devices through different applications, but because of the privacy concerns behind it, Apple has made it a lot less valuable to track a UDID. In fact, applications that utilize UDIDs are now not accepted into the App Store. Now, some ad networks rely on MAC as an identifier, which has proven to be just as reliable, but once again this could perhaps be devalued in the future by Apple if enough privacy concerns were to be raised.

In hindsight, it’s clear that tracking via mobile apps is possible, but it’s a bit clunky – it’s still being developed and the privacy concerns behind it are still only just being understood. It will take a while before a more stable solution is presented.

In the meantime, it’s incredibly important for marketers to understand just how an ad network or platform is tracking its users and ads so that they can make an informed decision on how accurate and informative a network may be for them.

Tracking for Mobile Web

Mobile web can use cookies, but in some instances, it can’t use them as well as the desktop can. There are still some concerns when it comes to tracking via the mobile web because of any lack of security software – cookies can often be left on a browser, and this brings up various privacy queries.

Because of the large fragmentation between devices, sometimes third party cookies are not immediately accessible either. This is especially the case for Apple’s Safari browser, which has third party cookies disabled by default.

Using Third Party Cookies

For the most part, third party cookies can be used on Android devices without any issue. Many browsers have third party cookies enabled by default – this includes Google Chrome, which is slowly becoming shipped with more and more Android devices and is becoming one of the most widely used Android web browsers.

When third party cookies are accessible, tracking users via the mobile web is easy. There are of course difficulties when it comes to tracking mobile web users into different apps, but these issues are often solved with various other tracking methods, like device fingerprinting or other methods described above.

Using Other Tracking Methods

Because some browsers do not have third party cookies enabled by default, namely Safari, ad networks have had to look for alternatives. Usually, a fingerprint is made, in a similar fashion to the device fingerprinting used for cross-app tracking. The accuracy of this fingerprint depends on a few factors such as how heavily the user browses the internet, whether javascript is enabled or what browser it is.

Problems with Privacy

The end goal for most ad networks and platforms is to be able to track users accurately throughout the entirety of a mobile operating system. Whether a user is browsing the internet, playing a game, or using another app, idealistically it would be beneficial for tracking to be accurate throughout all of these activities.

Unfortunately privacy does get pulled into all of this, and there’s only so much information you can pull, but more importantly store from a device before privacy becomes a real issue.