Adobe Audience Manager (AAM) and Salesforce Audience Studio (Krux): Side by Side Comparison: Part III – Post-Implementation

Welcome to Part III! I could sense that you are hooked by now and asking for more. The post-implementation phase is exactly where the rubber hits the road. This saying is so true because by now you are almost out of professional services and 3^rd party consulting support and you are on the hook to show ROI of this tool. You spent months, hopefully less, in implementing code, polishing the taxonomy, talking to data on-boarders and providers and creating segments. It is GO Time!

I hope that you have some use cases at hand that you want to execute on but if you don’t, please stop and go to the whiteboard and start brainstorming or we will have a very expensive tool on the shelf burning through your budget. Let’s not be those guys.

The post-implementation phase is where you start using the DMP, you start discovering how things work and most importantly reveal its power as well as its limitations. In this post, I will concentrate on the top things that you will be concerned about in post-implementation:

• Reporting
• Privacy and Consent
• Integrations and Activation
• Modeling
• Identity and Customer Stitching

Reporting

There is a lot of data in the DMP, some of your own 1^st party, some you purchased from a 3^rd party provider, some of your websites and even maybe your apps. Bottom line, there is a lot of data and you will need to lean heavily on the DMP reports to understand how this data can evolve to be actionable audience segments. Both tools offer a good variety of built-in reports that we will cover in detail.

Adobe Audience Manager

AAM offers a blend of out of the box reports as well as some advanced reports that might require additional setup. Here are the top ones:

1. Overlap Reports

   These are out of the box reports to be enabled on the backend by an Admin or client care. This is where your data analyst will spend the bulk of their time analyzing the data. Overlap reports, as the name implies, shows you the overlap in the audiences between different traits and segments and most importantly the overlap between traits and segments. For example, this report will help you in understanding how your 1^st party audience traits overlap with 3^rd party data providers. This report runs once a week and has a requirement on the minimum size of a traits/segment to be included in the report. The report is rendered via tableau and allows you to export and download the data if needed.

2. Trend Reports

   We tend to use this report heavily throughout setup and early stages of post-implementation. The trend report allows you to see how your traits or segments have been trending over time. It is an excellent tool for troubleshooting or tracking website updates that might impact your data collection.

3. Daily Trait Variation

   If you are interested in seeing how your traits are shifting from day to day, this is the report you need to track. This report will show the traits that have a standard deviation greater than or equal to 1.7 in either direction over a period of time. It is a good indicator of what traits are getting more activity versus others.

4. Audience Optimization (Advertisers and Publishers)

   This is a bundle of reports that are catered specifically for advertisers and publishers. In the case of an advertiser, you will have access to reports that allow you to analyze segment performance and the optimal frequency of impressions for all your media (or non-media) campaigns. This data can be sliced and diced by a defined set of dimensions such as Data, Platform, Campaign and others. For publishers, it is mostly confined to publishers using DFP. Once the data is loaded from DFP, you can analyze data based on audience segments or inventory.

Salesforce Audience Studio

Salesforce offers similar reports to Adobe although I envision these reports will be getting a major revamp after Salesforce’s acquisition of Tableau. Here are the top ones:

1. Top Overlapping Segments

   This report offers a peek into the overlap between the segments in your Audience Studio instance. It is visually appealing and also allows you to export an excel version of the findings. The reports classifies overlaps based on the top segments or filtered down to specific segments or categories. The only down side is that it only supports segments and no attributes.

2. Frequency Reports

   Media and event pixels get a lot of reporting love in Audience Studio. Under frequency reports, you will see 3 reports: Frequency plot, delivery trends and frequency breakdown. The goal of these reports to give you a way to analyze your campaign performance when it comes to media impressions. It allows you to define the lower and upper limit for frequency and analyze how each campaign performed when it came to under, on and above targeted frequency. Very intuitive and useful set of reports.

3. Ideal Frequency

   This is a fairly new report that Salesforce launched about a year ago. It is similar to Adobe’s “Optimal Frequency Report” and the goal is the same. Figure out the ideal frequency to use in frequency capping tactics across all campaigns as well as selected campaigns.

4. Segment Level Reach

   Probably better known as segment indexing report where it allows you to select a seed audience and index it across a 2^nd and 3^rd party provider. This is a great audience discovery tool to help you in understanding the personas of your high value audience segments. The reports are not automatically created, you will need to select the seed segments and data providers to index against.

5. Campaign and Segment Performance

   These are two reports that are usually used for different use cases but grouped in one. For Segment Performance, you measure the Impressions, Reach and Clicks over a period of time. While in Campaign Performance, you can see clicks, impressions, reach and CTR. The report allows you to breakdown the campaign ids into audience segments, placements and even Ads.

6. Reach Opportunities

   Another version of segment level reach and persona indexing. This one looks across all the audiences loaded in Audience Studio and how they index against the 3rd party data providers enabled for your instance.

My POV

Reporting in the DMP is underrated. It is surprising to most but the overwhelming volumes of data pushes some to completely ignore the report functionality and instead concentrate on pulling the data out and layer a Data Viz App on top of it. When it comes to the offering of these two platforms, their reporting offering is OK, not great but OK. It is not great for two reasons:

1. Slow: given that this is powered by Tableau and pulling so much data, it might take minutes for some reports to render.
2. Limited: There is a minimum volume that a segment or a trait needs to have to be included in a report. Moreover, some of these reports run once a week.

It is hard for me to say who has better reporting but if we look at specific use cases then:

1. Media Specific Reports: Audience Studio wins, Campaign Performance Report is beautiful
2. Flexible Reports: Audience Manager wins, segment to trait overlap as well as trait to trait overlap
3. Audience Discovery: Tie since Audience Studio has Segment Reach while Audience Manager has Adobe Analytics connection
4. Frequency Analysis: Tie since both offer optimal frequency or ideal frequency reports

Privacy and Consent

General Data Protection Regulation (GDPR), ePrivacy Regulation (ePR) and California Consumer Privacy Act (CCPA) are some of many regulations that will change the way we do things today. These are strong indicators that data privacy and consent are no longer an option, they are a necessity. They are a crucial part of any initiative that is collecting, hosting or leveraging consumer data (on the device and personal level). DMPs are in the heart of it all. Both platforms took measures to cope with these privacy laws and we will cover them here.

Adobe Audience Manager

AAM always had strong Access Control (known as Role Based Access Control, RBAC) and Data Exports Controls to handle who has access to the data and what data can be used for targeting, retargeting and personalization.

On the other hand, AAM is considered as a processor and not a controller which means that it follows instructions from a controller to process data. This is all good until a request comes in to remove someone from the platform. In this case, AAM leverages GDPR API and Adobe I/O to flush someone’s information. This is constructed in a form of a JSON request that outlines the data associated with this device (cookie or mobile id) to be flushed out.

Salesforce Audience Studio

Salesforce takes a different approach to privacy where it allows you to manage consent on every piece of data you are collecting. This can be done in different ways:

• API
• JS Consent Tag
• Mobile SDK
• In the tool
• File based

Regardless which way consent is set, there are consent flags that are configured that impact how this data can be leveraged down the line. Specifically, there are 6 flags:

• dc: if set to true, allows data collection
• al: if set to true, allows analytics
• tg: if set to true, allows targeting
• cd: if set to true, allows cross-device linking
• sh: if set to true, allows data sharing
• re: if set to true, allows re-identification

These flags can be hooked into a front-end consent management tool that sets the them accordingly to visitor’s privacy preferences.

My POV

Although neither of the platforms offer a consent management interface, Salesforce Audience Studio is ahead of the curve on this one and is better positioned to give advertisers and publishers control to ensure data flows into the platform with the proper permissions.

Integrations and Activation

Both platforms have robust integrations with the programmatic landscape of DSPs, Ad Servers and Exchanges but they vary a bit when it comes to their own Marketing Cloud Integrations.

Adobe Audience Manager

I mentioned this before and will repeat it here, AAM’s integration with Adobe Experience Cloud is a major advantage on many DMPs and even CDPs.

1. Adobe Target

   In terms of activation, AAM has a server to server integration with a powerful testing and optimization tool, Adobe Target, that allows you to do multi-variant testing and personalization based on audiences from AAM on the website, app and even single page applications (SPA).

2. Adobe Analytics

   AAM integrates with Adobe Analytics where audience segments can be pushed to Analytics for further analysis as well as audience discovery. This is not exactly a native integration and only applies for audiences that interact with your owned properties (site and app). In other words, only when visitor A visits the site/app, then his/her mapped segments to Analytics will be transferred.

   AAM also integrates with Adobe Campaign, email marketing automation tool, but I will not comment much about this here. I wrote few posts on this in the past. In short, match rates and lag time are two issues that limits this integration.

Salesforce Audience Studio

When it comes to integrations with Testing Platforms, the Optimizely integration stands out. Audience Studio and Optimizely built an integration from the early Krux days that enables audience segments in Optimizely Classic and X. This integration, similar to AAM’s with Adobe Target, allows site and mobile app personalization and testing based on Krux’s audience segments. This does come with a price ticket on the Optimizely side to enable this integration.

In the past year, Salesforce enabled the integration between Audience Studio and Salesforce Marketing Cloud to enable audience sharing from the DMP to Email Studio. The integration is enabled by the service support team and straightforward to use. Make sure that your consent flags are set properly to allow data to be shared to Email Studio as this is sharing anonymous profiles to PII system. Unfortunately, match rates are a limitation here as well.

My POV

In general, these integrations look better in a sales pitch than in practice. They have their share of limitations such as time lag, match rates and unsegmenting capability. Nevertheless, they serve a purpose. From my perspective, Adobe offers more integrations (Analytics, Testing) than Salesforce when it comes to their own Technology Stack. Another key differentiator would be activating on social channels such as Facebook and Twitter. Today both platforms utilize pixels that fire from Web/App and populates audiences on the social platforms which is a horrible way for audience sharing but unfortunately the only possible way (without paying extra for an data onboarder), THANK YOU WALLED GARDENS. The differentiator here is that Adobe is rolling out “people-based destinations” later this summer which will allow a better integration with social platforms, similar to the one Liveramp and CDPs have, and move away from pixel-based data sharing. We will wait to see how this pans out.

Modeling

Look-a-like modelling is another selling point in the DMP. Choose your best performing audience (best bookers, Most Valuable or big spenders) and look for prospects that share similar traits like them and most likely to behave like them if converted. Find these look-a-likes and target them. I am sure you heard this before.

Both platforms offer modelling capabilities, but both are limited. Let’s summarize their offerings:

• Neither platform allows you to bring your own model or give you the ability to alter the model. You are stuck with one option.
• Adobe allows you to choose whether you want to model against 1^st party data, 3^rd party data or both. Salesforce limits you to one option, either 1^st or 3^rd. You can’t combine both and you definitely can’t alternate between them. There are different licenses and costs associated with each option.
• AAM allows you to model based on a trait as the seed audience while Audience Studio limits the seed audience to segments only.
• Both platforms process models on a weekly cadence.
• AAM displays a list of influential traits that are common between the seed audience and helped in the audience modeling. Audience Studio doesn’t.

Identity and Customer Stitching

We discussed privacy earlier in this post and how each platform handles data governance to ensure it goes in line with the regulations. Now that privacy is accounted for, we will dive into how these platforms allow you to stitch customer profiles and hook into identity and device graphs.

Adobe Audience Manager

Adobe’s approach to identity is by explicitly separating device (cookie and mobile id) from a known/authenticated person. Once these are separated, you have the option to bring these two worlds together or keep them separate. This is done in the form of a feature called “Profile Merge”, also known as Profile Link. Profile Merge is only applicable on the segment level which allows up to 3 different rules of mixing device level data with authenticated profile data. These rules allow you to personalize and target in non-authenticated space based on authenticated attributes or leveraging behavior from one device to target on another. Really powerful.

In addition to the customer stitching, profile merge allows you to hook into device graphs to expand reach. Today there are limited options of accessible device graphs: Adobe Device Co-op, Liveramp Identity Link and Tapad.

Salesforce Audience Studio

Salesforce refers to this as Cross-Device Identity Matching (CDIM) which allows probabilistic and deterministic approach for cross device stitching. This is also enabled on the segment level offering 5 levels of device similarity and from there you can create new segments. For the deterministic portion of CDIM, it relies on the first party user match id that links to a person and usually exposed when logged in on web on in mobile application.

My POV

The cross-device capability, definitely the deterministic portion of it, is underwhelming in Audience Studio. It serves a better purpose of a device graph rather than an identity management solution. Adobe’s profile merge is ahead in this space where you have control over what and when data needs to be merged between device and an authenticated profile. This is crucial if your use cases are individual and household driven, where specifications change from one segment to another to make sure you are showing the relevant offer to the right person. In addition to the rules, Adobe’s access to multiple device graphs to expand reach is not possible in Audience Studio today.

That being said, Profile Merge is not a perfect product. It has limitations especially when it comes to merging onboarded data with attributes from multiple devices. It is the type of limitation that you won’t notice until you start testing and wonder why a fraction of your segments are not performing as well as you expected them to. Moreover, reporting based on profile merge can be spotty where the real time reporting doesn’t always reflect the right numbers due to how the data is connected on the backend. There seems to be a fix for this on the horizon and towards Q4 2019.

This wraps up the post-implementation phase where we covered the main things that you will be involved in within the DMP as you start executing on your use cases. Please share your POV and thoughts on this post.