Understanding data aggregation methods is crucial for accurately comparing KPIs across systems. We’ll discuss how these approaches impact the analysis of key metrics like Cohort Analysis, Lifetime Value (LTV), and Actual event-based metrics, helping you effectively interpret the data in our system.
| Cohort Analysis (by install time) | LTV (Lifetime Value) (by install time) | Actual Date Analysis or Real-time analysis (conversion time) |
What is it? | Cohort analysis involves grouping users based on shared characteristics or experiences within a defined time period. Cohort Analysis involves grouping users based on their app install date, typically when they first engage with the app after an advertisement. It's a study of user behavior within this specific group over time. | LTV is a metric that measures the total value of a user to a business over the entirety of their interaction with a product. In this context, LTV analysis is often conducted with respect to the install date of the app. | This approach focuses on analyzing data based on the date of events occurring within the app (for example, in-app purchases). Actual analysis helps you understand the performance of your business as a whole on a particular date. |
What does the report date range mean? | The report's date range indicates the install date of the user cohort. It provides insights into how these users, as a group, interact with the app. Your report may show you events that occurred outside of the date range of the report. | The report's date range pinpoints the initial user engagement period. It reflects the cumulative value of users starting from the install date and extending to the current analysis date. | The date range of your report represents the dates on which the events occurred. Since you're measuring by dates, your events cannot occur outside of the date range of your report. |
Examples | 1d, 3d, 7d, 14d, 30d, 60d, 90d, 120d, 180d, 360d
Select the date Jan 1, 2024, the metric Revenue D3, and the period 3d to see all the revenue garnered from users who installed the app on Jan 1, in the 3 days after installation. | Select the date Jan 1- Jan 7, 2024, the metric Count, Revenue, and “Cost per Event name” and the period LTV to see all the number of events and revenue garnered so far from users who installed the app on Jan 1 - Jan 7. | This aggregation is based on the actual date on which the event occurred rather than based on the install time. Select the date Jan 1, 2024, the metrics Count and Revenue, and the period Actual to see how much revenue was gained from all the users of the app on the actual day of Jan 1. |
Pros and cons | Pros: Allows for the analysis of user behavior and engagement over time; identifies long-term trends; useful for understanding user retention and churn. Cons: Can be complex in interpretation; requires a sufficient volume (30 days for Cohort period 30 days) of data for reliable conclusions. | Pros: Provides information on the long-term value of a user; helps in evaluating the effectiveness of marketing campaigns. Cons: Less effective for new or recently modified products. | Pros: Allows for quick response to changes in user behavior; useful for optimizing the product and marketing strategies in the short term. Cons: Less focus on the long-term value of the user; can be less predictable due to external events. |
Splitmetrics Availability | ❌ To compare cohort periods accurately, adjust to view cohort data based on calendar dates in your system. | ✅ USE for comparing data. | ⚙️ Not available yet. |
☝️Note: On the Splitmetrics Acquire dashboards all Count metrics are based on the install date, aligning with LTV, or are presented in a Cohorted manner.
How do I align Apple Ads data with Tracker data?
When you have aggregated data from Apple Ads based on the download date (the date when a user clicked 'Get' in the App Store), and you want to correlate it with install and in-app event data from trackers, it's beneficial to display tracker metrics and calculate other KPIs based on the install date (LTV) rather than the event date (Actual date). This is especially true for metrics such as Cost per Install (CPI) or Cost per Registration and others.
Below you can see the advantages of basing on the install date:
The sequence of Events:
When a user clicks on "Get" in the App Store (recorded in Apple Ads), it leads to the installation of the app. The install is the first key user action, followed by various in-app events (like registration).
By aligning these data with the install date, you get a more accurate picture of how an ad click translates into an install and subsequent actions.
Attribution and Accuracy:
Attribution of installs to advertising campaigns is based on the installation time (and click time if available), allowing for precise determination of which ad led to the installation.
Using the install date (or click time if available) helps avoid confusion if the same user performs different actions (events) at different times.
Advertising Efficiency Calculation (CPI, CPA):
Calculating Cost per Install (CPI) or Cost per Action (CPA, e.g., registration) becomes more straightforward and accurate when based on the install date (or click time if available).
This enables marketers to accurately assess ad campaigns' ROI (Return on Investment) and determine their effectiveness.
In general, using a single time criterion (install date) simplifies the comparison of data from different sources. It provides a clearer and more consistent view of the performance of advertising campaigns.
How do I choose the right approach?
The choice of approach depends on your goals and data specifics. Here are some recommendations:
For Long-Term Analysis: Cohort analysis and LTV by install date are best suited for understanding users' long-term behavior and value.
For Short-Term Changes: Analysis by event date is useful for quickly assessing the effectiveness of campaigns or product changes.
Combined Approach: The best analysis is often achieved by combining different methods to get a complete picture of user behavior and value.
❗Note: Working with metrics in actual time data grouping is currently only available for BI integration. Please contact support for more information.