For marketers responsible for promoting mobile apps, ad tracking has become a mission critical component of most app marketing programs. By being able to attribute downloads, and post-download events like in-app purchases, to the marketing source, marketers are realizing a powerful way to gain insights that drive business and deliver outstanding ROI.
But there are challenges associated with choosing a mobile ad tracking solution. There is no industry standard for mobile ad tracking and reporting, so marketers are on their own to make sense of the different technologies and determine the best approach for their businesses. Several tracking solutions are on the market today, each employing very different approaches. It’s difficult to determine which technology to implement because many of the options have trade offs, and support for each technology can vary by ad network. Further complicating matters are the ongoing changes and privacy concerns, making it difficult for marketers to choose a tracking technology with confidence.
This paper helps you make sense of it all. It provides detailed overviews and comparisons of today’s mobile ad tracking technologies, and provides practical advice to help you decide what may be the right ad tracking technology for your business.
Tracking your app marketing campaigns is critical to your success. So let’s get started and help you select the right ad tracking technology.
Mobile Ad Tracking Technologies
Unlike the online world where most tracking is cookie based and works consistently across the industry, mobile uses a range of different tracking technologies that vary from ad network to ad network and by platform. Unlike its web advertising counterpart, in-app mobile advertising does not support third-party cookies – or even first-party cookies – which is a standard and easy-to-implement approach.
On Android, ad tracking is relatively simple, through use of the industry-standard Android Referrer. But on the iOS platform, ad tracking is more complex, with a range of ad tracking choices and varying capabilities that address different business goals.
The four primary types of ad tracking technologies available today are:
1. ID-Based Tracking (Apple Advertising Identifier, Facebook ID, MAC Address)
2. Android Referrer
3. Digital Fingerprinting
4. HTML5 Cookie Tracking
- Advertising Identifier, Facebook ID, MAC Address
- Unique IDs for click, conversion, post-download events are matched to attribute ad spend
- Requires integration between ad network and advertiser
- Provides accurate attribution
- Takes “digital fingerprint” of device configuration
- Matches configuration at time of conversion
- Privacy friendly – no identifiers are collected
- Uses statistical estimations
- Has an error rate
HTML5 Cookie/ Web Tracking
- Small file stored in local storage (“cookie-like”)
- Clicks and conversions involve a “flip” to Safari to gather ID data
- ID data is passed via a redirect to the Tracking DB
- Accurate attribution
- Impact on user experience
ID-based ad tracking approaches are used primarily on Apple’s iOS platform and are by far the most widely used technique for tracking on iOS devices. This approach involves capturing a unique device ID at the time the user taps an ad, and also at the time of a conversion. An ID-matching process then matches these IDs to determine the marketing source (ad network, ad creative, ad size, etc.) that should receive credit for the conversion. Each of these IDs is a unique anonymous identifier linked to the device on which the downloaded app resides. It is expected that advertisers will transition to the Advertising Identifier, Apple’s recommended ad tracking mechanism.
There are other ID-based alternatives for IOS including MAC Address, and MAC Address variants such as ODIN. With the release of Apple’s Advertising Identifier, support for ODIN has waned. There have been other attempts to create tracking identifiers, such as OpenUDID and SecureID – these never gained market acceptance and have been withdrawn from the market. In addition to general-purpose identifiers, Facebook has its own unique id-based tracking mechanism.
The UDID – Deprecated by Apple
As of May 1, 2013, Apple has discontinued approval of any app that utilizes the UDID. Historically, the UDID (Universal Device Identifier) was the most commonly employed device ID for iOS. The UDID is a unique number that is non-erasable and is tied to a specific hardware device. In September 2012, Apple released iOS 6 and included a replacement for the UDID, the Advertising Identifier. The Advertising Identifier works in a similar manner to the UDID, but allows the user to reset the identifier and has an opt-out mechanism from behavioral tracking.
App developers who are currently using the UDID in their apps should make sure to discontinue use of the UDID and switch to an alternate tracking method before their next submission to the App Store.
Following is an overview of ID-based ad tracking technologies:
Advertising Identifier: The Advertising Identifier, which was released with in September 2012 with iOS 6, is Apple’s recommended tracking technology. It is expected that the Advertising Identifier will become the dominant tracking method for iOS-based traffic.
MAC Address: This includes the ODIN standard, plus several hashed variants. This method is supported by a number of ad networks as their primary tracking technology. It is controversial because it is directly linked to the hardware device and cannot be reset.
Facebook: Facebook has a proprietary, ID-based tracking system that it utilizes in conjunction with its partners. Facebook has a set of guidelines around usage and carefully screens its partners to ensure compliance with its data security guidelines.
Advantage: Provides very accurate attribution.
Disadvantage: Requires transfer and synchronization of IDs between the advertiser and ad networks in order to attribute marketing spend properly.
Digital Fingerprinting is a tracking technique that attempts to attribute conversions to marketing sources without the use of direct ID-based identifiers. This technology provides broad access to both in-app and mobile web traffic and is viewed as privacy-friendly.
At the time of an ad click, a “fingerprint” is stamped with user device configuration data, such as device type, operating system, IP address and potentially dozens of other attributes. This technology then attempts to match this fingerprint to subsequent corresponding fingerprints generated during user actions such as downloads, launches, purchases and others.
The challenge with Digital Fingerprinting is that it uses statistical estimation techniques that can cause false or incomplete matches, which leads to an error rate. For example, a conversion may fail to match if, in the time between the initial tap on an ad and the download, the user switches from the mobile network to Wi-Fi, thus changing the IP address of the device. In addition, many mobile networks have large numbers of users sharing the same IP address, and it is possible for users to have devices with the same attributes, something made more likely by the limited number of Apple device configurations.
Finally, Digital Fingerprinting often utilizes click URLs which redirect an ad click to a tracking server before going to the app store. Some networks, most notably Apple’s iAd network, do not support click URLs. In addition, some ad networks allow their publishers to opt out of click URL redirects, and there are those that simply choose not to support this capability. The impact of using a technology that relies only on click URLs can be significant as it may lock you out of as much as a third of the available mobile advertising inventory. For this reason, Digital Fingerprinting is often viewed as a second tier tracking method that should be used when other methods are unavailable, or when there is an overriding concern about privacy.
Advantages: Avoids privacy concerns associated with identifiers. Versatile technology that can be used for tracking mobile web traffic and other promotional channels.
Disadvantages: Not entirely accurate due to the use of statistical estimation techniques. Many publishers do not support click URLs, which lock advertisers out of as much as one-third of the available mobile advertising inventory.
HTML5 Cookie Tracking
This technology is utilized on iOS devices and involves storing a cookie-like file on the device. The process works in a manner similar to cookie-based tracking on the web, though it technically does not use actual cookies. When the user clicks on the ad, an identifier is stored on the device. This identifier can then be retrieved at the point of conversion. Because iOS does not allow apps to write cookies, this method redirects to the Safari browser, and writes to the device a lightweight cookie-like file with an identifier. Some marketers object to the redirect since it briefly flashes an interim screen before the app begins to download, which can hamper the user experience.
Advantages: Provides accurate tracking (there may be some tracking loss due to users abandoning during the redirect). Easy for marketers to understand because it operates similarly to tracking on the web.
Disadvantages: Requires a redirect to a Safari page on the ad click, and also on the first launch of the app. Utilizes click URLs, which certain networks do not support, and may lock you out of a portion of the available advertising inventory.
How Tracking Technologies Affect Your Available Ad Inventory
Ad Networks: Aggregate traffic from publishers and provide a single point of contact and ordering for advertisers. There is a great deal of variety in the tracking technologies supported by the networks.
Incentive Networks: Provide a reward to a user that downloads an app. Since a financial transaction occurs, it is critical that the user be identified reliably. For this reason, incentive networks support ID-based methods.
Premium Traffic: Publishers with their own ad serving capabilities and sell inventory direct to very large buyers or aggregators. They support a variety of tracking technologies.
Real-Time Bidding Exchanges (RTB): A rapidly growing advertising channel where high-speed, automated technology runs auctions on an impression basis. RTB exchanges support a variety of tracking technologies.
Facebook: Facebook provides an opportunity to reach precise sets of mobile users via its rich data capabilities. Facebook provides an ad unit designed specifically for promotion of apps, and has its own tracking technology that provides accurate mobile measurement. Facebook makes this tracking infrastructure available to selected partners.
How to Choose the Right Ad Tracking Technology for Your Business
At the end of the day, choosing an ad tracking solution is a business decision, rather than a technology decision. Following are some guidelines to consider when evaluating mobile ad tracking and attribution solutions.
If you track financial transactions… You should consider using ID-based methods and/or Android Referrer because this is the most accurate tracking technology, and it is critical that you identify the user accurately. You may want to avoid digital fingerprinting due to the accuracy concerns associated with its statistical estimation techniques.
If you need complete attribution of post-download events… You should consider using ID-based methods and/or Android Referrer because these provide the most accurate and comprehensive post-download event tracking.
If you want to optimize often throughout a campaign… You should consider using an attribution solution that is integrated with optimization technology. By using an integrated attribution / optimization solution, attribution insights are actionable immediately through the optimization engine, for better price/performance.
If you need access to large volumes of media inventory… You should consider using a tracking and attribution solution that employs multiple tracking technologies. Each network supports its preferred tracking method. Therefore if you employ only one tracking technology, you will only be able to work with the networks that support that particular technology – which means you will have access to less media inventory. By choosing an attribution solution that supports multiple tracking technologies, you will give your business access to as much media inventory as possible.
If you work with incentive networks… You should consider using ID-based methods and/or Android Referrer because you are providing a reward to a user and it is critical to identify the user accurately. ID-based methods and the Android Referrer provide accurate tracking and are the technologies typically used by incentive networks. You may want to avoid Digital Fingerprinting due to the accuracy concerns associated with its statistical estimation techniques.
If privacy concerns are paramount to your organization… You should consider digital fingerprinting because it does not use direct ID-based identifiers. Once it sees significant market adoption, Apple’s Advertising Identifier is another good option, since users have the option to reset it, and there is an opt-out for behavioral tracking.
If you want to run campaigns on Facebook… You should choose a solution that includes support for Facebook’s proprietary tracking solution.
Mobile application download tracking solutions by BigBalli