App user data, also known as app audience data, is vital to understanding any group of people you plan to reach. Knowing what apps they download tells you a great deal about their interests and preferences. With trusted, high-quality app user data, marketers are better able to anticipate and fulfill any audience’s needs, understand and appeal to their perspectives, and acquire groups of users that align with your business goals.
In this guide, we’ll cover the fundamentals of app user data to help you get your arms around the topic and make more informed decisions as you create or build a marketing audience.
At the highest level, mobile app user data tells you which users have which apps installed.
When you buy app user data, you receive a package of “mobile advertising identifiers,” or MAIDs, which are unique, pseudo-anonymous identifiers tied to a specific mobile phone. Each MAID includes a uniform resource identifier (URI), which is a code that identifies an app in the same way that URLs identify websites.
MAIDs also include a consumption type, which confirms the app in question is downloaded, as well as a timestamp and company ID that specify the data source. Some MAIDs also feature a referring URL, device IP address, a country code, carrier, and details about the device model.
Both iOS (Apple) and Android (Google) use slightly different MAIDs that work the same in practice but have vendor-specific names. The Apple MAID is called the Identifier For Advertisers, or IDFA. The Google MAID is called the Advertising Identifier, or Ad Id. Both are 32 hyphen-separated characters that look like this: 3f097372-f01e-4b64-984c-395ae5828ee6
What makes MAIDs extremely valuable is that they are persistent identifiers that can be linked to many other elements of an individual’s online experience, albeit anonymously. Let’s take a look at three ways you can put these connections to work for your marketing team.
Mobile app user data can be incredibly valuable for marketers, allowing you to create stronger campaigns, build more effective models, and make smarter decisions. Here are three common use cases that illustrate the true potential of app audience data.
The particular combination of apps on each of our phones reveals a profile of our individual interests that can be more useful than our demographics alone.
What any marketer wants is to find those people who are already predisposed to want the offer, so budgets can be used efficiently and deliver optimal results. App audience data provides a unique pathway to interested audiences, allowing you to build a custom-tailored audience.
If you sell yoga mats, for example, you can purchase app audience data for Daily Yoga, Yoga Studio, Down Dog, and other popular yoga apps. You can then use these MAIDs to run a new user acquisition campaign through Facebook, Google, or a demand-side platform (DSP) to target this specific subset of users with digital ads that will be more likely to resonate.
The mechanism is relatively straightforward. You acquire MAIDs associated with one or more apps and send them to Facebook, which has its own set of MAIDs for every user who has the Facebook app installed. By cross-referencing the two lists, Facebook can find matches and display your yoga mat ads within their Facebook feed.
You can repeat the process for all of the numerous properties Facebook, Google, and other DSPs own, so it’s easy to see the potential for audience building no matter what your marketing objectives are.
Competitive intelligence is a natural application of app audience data. If you are preparing to launch a new app, for example, you need to understand who is using apps that yours will need to outshine.
Imagine you’re developing a game in a new category. Your previous domain was puzzle games, but now you’re extending into sports games. Think about how helpful it would be to know which users are playing the most popular games in this new category, as well as which users are playing games from your direct competitors.
One Narrative customer did this and discovered that their new game’s potential users were not at all who they assumed they would be based on their previous experience. Competitive intelligence fueled by app audience data showed that their new audience belonged to a different segment of the population, and this simple insight changed the entire trajectory of their launch strategy and execution.
Competitive intelligence is just one kind of analysis app audience data enables. This gets back to the persistent and interconnected nature of MAIDs. Because MAIDs can be associated with activities that reveal demographics, you can cross-reference MAIDs for an app with MAIDs linked to age or gender or geography to create extremely detailed profiles of your audience.
App audience analysis creates all kinds of opportunities for marketing insight. Making connections between demographics and an app — also known as enrichment — can help you deploy marketing campaigns with greater precision.
Let’s say your offer is a membership at a high-end gym. Combining audience data for fitness apps and luxury retail apps can help you find an even more attractive audience than either one of those data sets on their own. Then add age, region, and gender to sharpen your focus.
Now that you know what mobile user data is and how to take advantage of it, all you need to do is acquire it.
Data is usually priced one of three ways. You can subscribe to a data feed (or data license), which lets your company ingest a specific data set that is updated on a regular basis. You can buy a large batch of data for a one-time fee. Or you can negotiate a custom arrangement with a data provider for a specialized need.
Price, unfortunately, is the most transparent aspect of data. Others are less so, including quality, even though the quality of the data set directly affects how well (or how poorly) it will help you solve your specific marketing challenge. Garbage in, garbage out, in other words.
That doesn’t mean marketers are powerless to determine data quality, only that it may take some effort. In general, your team will need to itemize all data sources, understand how the data was collected, ensure the collection methods meet your compliance standards, and ensure the data is truly anonymized. As with any provider relationship, asking for references is essential. Connect with them to ask about data quality, customer service, and issue resolution. Finally, get a sample data set that your team can test in your use case to see if it delivers as expected.
Keep in mind that comparing data quality across providers is inherently complex. There is not a standardized set of metrics that characterize data quality, and different providers do not measure every metric in the same way. At minimum, it will be important to ask every provider the same questions about what MAIDs include and how they were sourced.
Functionally, it is important to find out how much work you’ll need to do for enrichment. Most data brokers provide the full “firehose” of data, which means you can get a MAID-to-app list and a MAID-to-gender list, but performing the actual matching falls to you.
Another important factor to keep in mind is time. Sourcing mobile user data tends to be an extremely manual process, one that can take months to complete. In fact, the timetable for understanding what data you need, finding a provider, performing due diligence, completing the transaction, and managing the data might be six months or more. How much will your marketing challenge have changed in six months?
Consider also that this process does not scale. Every time you need new data… or have a new idea to test… or change your marketing strategy… or develop a new use case… you have to repeat the sourcing process from square one.
Narrative helps marketers overcome the systemic lack of agility and automation in data acquisition — especially for use cases designed to take advantage of fast-moving trends in app audiences.
The category of the digital consumption:
|Uniform resource identifier (URI)||
Identifies the mobile app
|ID||A unique identifier for a device or user. This can depend on the platform as well as the point of origin for the ID. One user is typically associated with multiple IDs: the mobile ad ID from their smartphone, hashes of their emails, browser cookies from various websites, etc. The full ID contains both an ID type and an ID value, separated by a colon. Common types of IDs include email hashes (sha1_email, md5_email) and mobile IDs (idfa, adid).||Always present|
The time the observation or data event occurred, expressed as milliseconds from January 1, 1970
A unique identifier for the partner that the data was sourced from
The referring URL passed in from an HTTP call. Roughly the resource accessed before the one represented in the current event.
IP address of a given device
A string representing the device the event was collected on
A2 (ISO) country code
Carrier or mobile virtual network operator (MVNO) that the phone belongs to
Device make and model