Devices vs. Users: A New Level of Ad Campaign Analytics in MyTracker
Author:
Olga Trushkova Product Manager at MyTracker
July 06, 2023
MyTracker has expanded its analytics toolkit. Now you can gauge the performance of your advertising platforms even if users interact with an app or website from different devices.
To learn more about how this works, visit our blog. In this article, we will focus on the differences between device-level and user-level analytics, listing their pros and cons.
The difference between device and user analytics
The number of devices that have visited your website or app is not always the same as the number of users. This is hardly surprising, as a single user may have several devices: a smartphone, a tablet, a laptop, etc.
Plus, a user can create several accounts on a single device, and analytics systems will treat them as unique users.
Device analytics
Analytics systems see a device as a unique set of anonymous identifiers associated with a physical device, through which a user interacts with your service.
In MyTracker, device-related metrics are marked by the index d – devices.
The set of identifiers that help detect a device depends on the specific platform – Android, iOS or WEB.
New Devices, d reveals the number of devices that interacted with your app for the first time on one of the platforms.
Active Devices, d shows the number of devices that interacted with your app on one of the platforms over a selected period of time.
Reattribution, d helps measure how many devices had previously interacted with the app, but were not active before being activated again on this device.
Attribution, d shows the total number of devices engaged or activated during the selected period.
The extent to which app owners can analyze user behavior on a single device depends on the platform, its privacy rules, and analytics restrictions.
For mobile apps, the identifier is more or less constant, enabling owners to see the user's journey from installing the app to target events: payment, login, adding to cart, completing a game tutorial, etc.
Browser cookies are used to identify unique devices in Web applications. Unfortunately, cookies are not as permanent as a physical device identifier. If the user deletes cookies or their lifetime expires, then for the analytics system, the user behavior on this device will stop being updated, and it will see a new device with a different set of identifiers.
Pros of device-level analytics
Device IDs can be shared by advertising platforms and analytics systems, which means that advertising platforms can better optimize ad impressions using signals from your tracking system about the efficiency of engaging a specific device.
Many device IDs are anonymous and therefore accessible and not requiring additional user consent (for example, IDFV for iOS or App set ID for Android).
Cons of device-level analytics
Users tend to change their devices. There are many reasons for this, including obsolescence, malfunction, loss, etc. Therefore, analytics for a user may be distributed across multiple devices. If your service has been on the market for a long time and you have already attracted a large audience, then it is highly likely that your existing audience is hiding behind new devices.
Analyzing devices on different platforms is impossible. For that reason, it is difficult to measure the audience of a service if its users interact with a variety of platforms, particularly if it is Android + iOS + Web. In this case, device-level analytics of the audience across platforms would be impossible, as various platforms use different device identifiers.
Some device IDs are updated too frequently, which complicates user analytics. This is the case with cookies for Web apps. As with changing devices, user behavior will be treated as generated by multiple unique devices. Tracing it to a single user is impossible.
Some users may employ several devices on one or several platforms. In this case, you will see multiple unique devices, while in reality they belong to a single user. This makes it difficult to analyze user experience or audience size.
Emulators and bots are seen as new devices. However, some analytics systems have built-in anti-fraud tools. For example, MyTracker’s Fraud Scanner provides automatic protection against all types of mobile and web fraud.
User analytics
A user is an individual using your service. To start doing that, they need to create an account and accept the terms of the user agreement.
For MyTracker, a user is an anonymous customUserId associated with accounts that are part of the same project.
MyTracker does not recommend providing personal data such email, phone number, and other sensitive information as user ID.
In MyTracker, metrics related to project users are marked by the index u – users.
Unlike devices, individuals generally have a limited number of accounts in a service. Customers understand that to use the service, they need to create an account and log in to:
save their progress level in a game;
link their bank cards for payments in online stores;
save and use favorite products or articles;
use the service across devices.
User ID helps analyze the project audience across different apps and platforms. This helps track user progress while they use your service, even if a device falters.
New users, u helps estimate the number of newly created accounts and the app in which the user was active for the first time.
Reattribution, u, shows the number of users that have not been active for a long time but became active during the selected period.
Attribution, u shows the total number of users acquired or activated during the selected period.
Active users, u helps analyze active audience among signed-in users.
Unlike device-level metrics, user-level metrics can be analyzed across apps that are part of the same project and estimate unique audience. To do this, just change the grouping of the report, leaving Project and excluding Apps or Platforms.
There are fewer active users within the project than the sum of active users in each individual app, as some customers used more than one platform and app to interact with the service.
Pros of user-level analytics
A single individual has a limited number of accounts, as opposed to devices or combinations of technical device IDs. Therefore, measuring signed-in users better reflects the service's audience.
User ID helps analyze the behavior of a user when they interact with several operating systems and platforms. It enables you to better understand the user journey.
The analysis of active users, unlike devices, shows the signed-in audience in your service.
Cons of user-level analytics
An individual may create and use multiple accounts in order to interact with the service.
User IDs cannot be shared by advertising platforms. An advertising partner will not be able to improve ad impressions if your service’s user ID is anonymous.
To analyze user behavior, they need to sign up and perform actions after login. You also need to configure the transfer of data on events to MyTracker by indicating the device and user IDs.
A new level of ad campaign analytics – free of in MyTracker
For accurate tracking, information on both users and devices needs to be collected. MyTracker is the first solution to offer free cross-platform and cross-device attribution and analytics.
If your project has user accounts, try our tool to ensure seamless analytics for user behavior on various devices. With this data, you will be able to optimize traffic purchases based on past actions of the user on different platforms.
