Sunday, September 29, 2019

Methods to Export Data Out of Adobe Audience Manager

My previous article was centered around the various methods to send data to Audience Manager. This post will cover some of the ways by which we can export data out of AAM. In this article, I will follow suit like before and link to individual posts I've written as well as reference the official Adobe documentation.

One important concept to understand is that we're typically exporting device IDs (cookies, Mobile IDs), 
segments IDs, aggregated metrics in the AAM UI (E.g. Unique Trait Realization) or hit level data collected in AAM (logs). I'll provide an overview of each of these at a high level. 


1. Audience Manager Reports

I wrote an article last year explaining some key AAM metrics which can be slightly complicated to understand at first. In this section, I'm not going to cover each report in detail as they're already covered here but just explain how you can export data from these reports can be used for at a high level. All AAM reports are located under the "Analytics" menu in the top navigation bar.

General and Trend Reports

General reports allow you to export data as CSV (traits, segments and destinations) as CSV for the last 1, 7, 14, 30, 60 and 90 days as well as for the lifetime of the customer.

Audience Reports

These reports require additional access (AAM Consultant/Support can provide) and the UI is built specifically in Tableau. There are various reports such as:
  • Audience Optimization- Media reports populated via pixels and Ad Server logs.
  • Trait Variation- Shows difference in trait volume based on standard deviation.
  • Unused Signals- Shows attributes collected by AAM that don't have traits created. Recently added signals is the recommended way to get unused signals now.
  • Overlap Reports- These reports allow you to look at the overlap between traits, segments or between traits and segments in one report. Common usage is to look for low overlap between 1st and 3rd party data sets to determine which 3rd party data sources to buy for prospecting.

All these reports can be exported using the default options which Tableau offers as shown below:



When to use this

When you want to export aggregated AAM metrics tied to traits, segments, destinations out of the AAM UI.

2. Audience Manager Destinations

Destinations (server to server) are the most common way of sharing AAM data with other 3rd party Demand Side Platform, ad networks or optimization platforms for downstream activation. The data which is shared is basically AAM cookies (3rd party UUID), mobile device IDs along with the segments (criteria) which need to be activated. The general expectation is for the destination partner to also have the AAM cookie present which they leverage for advertising purposes. The match rate is calculated based on how many AAM cookies can be found in the destination partner ecosystem so if the match rate is high then the the ability for the partner to reach users for advertising will also improve.

The other types of destinations such as cookie based on URL where you can share AAM segment qualifications on pages where AAM code is deployed.

People-based destinations is a new addition to AAM where we're able to send hashed email addresses (without the need for cookies) to social platforms such as Facebook for serving offers based on offline data. 

The Outbound File History report shows you how much data is shared with individual destination partners on a daily basis. 


When to use this

When you want to send AAM device IDs and their segment qualifications to 3rd party media partners for advertising activation (E.g. retargeting, frequency capping etc.).

3. Customer Data Feeds

Customer Data Feeds allow us to get raw Audience Manager data such as timestamp of events, page request parameters, segment/trait realizations, device, IP address to name a few on an hourly basis. I'm not going to cover this in detail as the official Adobe documentation is really good to dive deeper into it but I'll cover the core concept.

CDF is delivered to an Amazon S3 bucket (unique to each customer) on an hourly basis and needs to be requested via customer service as there's no self-service option to get these. The content of the file is constant and follows a similar pattern as shown in the example below.

Each CDF file (compressed) is accompanied by a .info file which provides a status of the numbers files, size etc. in JSON format. The other thing to note is that the timestamp included in the CDF file name may be different than the timestamp of the hit level data in the file in case there's some latency.

When to use this

When you want to store raw AAM hit level data (limited in the AAM UI) in your DataWarehouse and do data modeling or analysis on this data.

4. Custom Data Exports

This method is used to export IDs captured in Audience Manager and can range from UUID, MID, segment IDs among others and are set up by Audience Manager consultants or AAM support. This is typically set up when you need to send raw AAM Ids to a destination with which AAM doesn't have an integration such as some Email Service Providers for activation. Below is an example of an export and the various fields associated with it. Here's a link explaining what each of the file format options mean. There are numerous way to construct an export and it'll take a whole article to go through each so I'll save that for another day.



When to use this

If you want to test your CRM import (by exporting a custom file) or sending AAM IDs and their segment qualifications to a 3rd party partner which may not have a server to server integration with Audience Manager. 

5. Experience Cloud Audiences

We can also share Audience Manager segments and audiences to tools such as Adobe Analytics, Target and Adobe Campaign. Please note that all three integrations require the Experience Cloud ID service to be implemented.

Adobe Analytics

Audience Analytics allows you to share demographic data or offline data captured in AAM with Adobe Analytics. 

Adobe Target

It is recommended to leverage the server to server integration with Adobe Target to send AAM data for A/B testing and personalization. This is usually onsite data captured in AAM but can also be offline data combined with onsite.

Adobe Campaign

AAM data can be sent to Adobe Campaign for email marketing activation via the OOTB integration between the two systems.


When to use this

If you want to send segment qualifications to Target for A/B testing or personalization, send demographic data to Analytics or send segments to Adobe Campaign for further activation via emails.

There are other methods which can arguably be added to this list such as Rest APIs and DCS (S2S) API which can also be used to export AAM data. So, which of these is the most common method you use to export data out of Audience Manager or is there something is not included in this list?