Column Headers

App events data must be custom exported in CSV format like the following example header row. It is important to use the exact column headers shown in the example below. The cohort day is specified in each metric column header as a “Dn” suffix where n is the cohort day. The example includes D0 and D1, but your file should include all cohort days selected in your pricing plan.

Date,Country,Installs,Sessions D0,Sessions D1,Retained Users D1,Revenue Amount D0,Revenue Amount D1,Revenue Events D0,Revenue Events D1,Unique Payers D0,Unique Payers D1

For files that contain data for multiple apps, the following example header row must be used instead where each line specifies the app's store ID in the Store ID column.

Date,Store ID,Country,Installs,Sessions D0,Sessions D1,Retained Users D1,Revenue Amount D0,Revenue Amount D1,Revenue Events D0,Revenue Events D1,Unique Payers D0,Unique Payers D1

Metric Columns

The table below specifies the column headers that must be used based on the metrics you selected in your pricing plan. Remember that each column header that represents a cohorted metric must include a "Dn" suffix as described above for each cohort day selected in your pricing plan. Any column header that represents a metric that can never be cohorted must never have a "Dn" suffix.

The table also specifies the type of aggregation required for each cohorted metric. Cumulative aggregation refers to the sum of all cohort days through the given cohort day. Incremental aggregation is similar, but the sum must start from one day after the preceding cohort day. For example, if two revenue amount columns are included, Revenue Amount D1 and Revenue Amount D7, Revenue Amount D1 should represent the revenue that was reported on day 0 through day 1 and Revenue Amount D7 should represent the revenue that was reported on day 2 through day 7. On day aggregation refers to the given cohort day only. For example, Sessions D7 should represent the number of sessions counted only on day 7.

Metric	Column Header	Cohorted	Aggregation
Installs	Installs	Never	N/A
Reinstalls	Reinstalls	Always	On Day
Sessions	Sessions	Always	On Day
Retained Users	Retained Users	Always	On Day
Revenue Amount	Revenue Amount	Always	Incremental
Ad Revenue Amount	Ad Revenue Amount	Always	Incremental
Purchase Revenue Amount	IAP Revenue Amount	Always	Incremental
Purchases	Revenue Events	Always	On Day
Unique Payers	Unique Payers	Always	On Day

If Ad Revenue Amount, Purchase Revenue Amount, or both are contained in the file, Revenue Amount must be omitted. Revenue Amount is total revenue so when Ad Revenue Amount and/or Purchase Revenue Amount are included, it is the sum of both, which we compute automatically.

If you are using a data partner integration to share marketing data and are only sharing app events data via custom export, the column headers you must use will be different than those documented in this article. Please contact your Customer Success Manager to receive the required column headers.

Cumulative Metrics

If you'd prefer to import one or more metrics with cumulative aggregation instead of the required type of aggregation stated in the table above, you can indicate such metrics by including the text __cumul-{column}__ anywhere in the file name, once for each cumulative metric, where {column} is the Column Header of the metric in the table above.

For example, if you'd prefer to import both the Revenue Amount and Purchases metrics with cumulative aggregation, you need to include the cumulative metrics in the file like:

events_123456789__cumul-Revenue Amount____cumul-Revenue Events__.csv

Last Touch Columns

If you’d like to be able to easily compare last touch attribution and incrementality data side-by-side in the dashboard, you should include last touch dimensions like in the example below.

Date,Channel,Campaign,Country,Source App,Installs,Sessions D0,Sessions D1,Retained Users D1,Revenue Amount D0,Revenue Amount D1,Revenue Events D0,Revenue Events D1,Unique Payers D0,Unique Payers D1

If you wish to include multiple apps in the file, you must the following example header row instead where each line specifies the app's store ID in the Store ID column .

Date,Store ID,Channel,Campaign,Country,Source App,Installs,Sessions D0,Sessions D1,Retained Users D1,Revenue Amount D0,Revenue Amount D1,Revenue Events D0,Revenue Events D1,Unique Payers D0,Unique Payers D1

The last touch dimensional columns (Channel, Campaign, and Source App) are restricted based on the granularity you selected in your pricing plan. Source App usually refers to the publisher app or sub-publisher on SDK or video ad networks and keyword ID on most search platforms, but it can contain any sub-campaign level you wish. If you choose to include last touch dimensions, simply omit any columns from your CSVs that are not included in your pricing plan based on the following table.

Granularity	Supported Columns
Channel	Country*, Channel*
Campaign	Country*, Channel*, Campaign
Site	Country*, Channel*, Campaign, Source App

An asterisk (*) indicates a required column that must be present in the file even if there's no value.

When including last touch dimensions, it is critical that the country level app event totals are accurate since those are the only data points that are used in the model. That means app events attributed to organic demand must be included in the export. Otherwise, they’d be missing in the country level totals. Those app events must appear in the export under a Channel and Campaign both called Organic (case-sensitive).

Updated on: 25/01/2024