AEP vs. Marketo Engage: Where to Build Audiences?

[Michaela_Iery3]

More Marketo Engage customers are bringing Adobe Experience Platform (AEP) into their marketing tech stack or are new-to-Marketo customers who have purchased AEP and made the decision to simultaneously move to Marketo as their new marketing automation platform. So more and more, I receive the question – where should I build my audiences for marketing?

As we know, Marketo makes it easy to use smart lists to identify people that have certain attributes and/or behaviors and then use that smart list to take action on those people. If you are not familiar with AEP, the platform also creates audiences by querying for attributes and/or behaviors based on the data available in its data lake and person profiles. This data could be sourced from multiple systems, usually including Marketo.

A great advantage with AEP is that you can create an audience in one place and then share that audience to multiple destinations at once. For example, you can create an audience of all people who were invited to an event but have not yet registered for it and share that audience to Marketo for a reminder email, to Target for a web personalization experience and to Google Customer Match for ad retargeting. You can even update this audience in AEP as the people in those audiences complete the calls to action and register for the event. This has obvious efficiencies compared to building and updating the same audience (people who are invited but not registered) in multiple applications as well as seeing in one platform the results.

As a marketer, I’m excited to have one place to create audiences and push them to multiple marketing platforms at once. It feels like a great way to view and evaluate my audiences in an omni-channel way. But as a Marketo Engage user for more than a decade, it feels a little strange to me to build audiences somewhere else when I probably have the same data about the audiences in Marketo.

There is no one-size-fits-all answer here. I lean toward creating as many audiences as makes sense in AEP due to the efficiencies described above. One of my teammates called it “AEP as a ‘center of gravity’ for audience development” and I like that. It demonstrates that AEP should be where you start but it doesn’t mean you’ll always land there. I advise my clients to consider the following:

How AEP activates audiences to Marketo Engage

The current AEP connector to Marketo Engage can only create new person records or put new and existing records on static lists that represent an audience. This means, AEP can give you a list of records based on whatever criteria it has used to make an audience, but it can’t put any data on the existing records that already live in Marketo. So if AEP wants to tell Marketo that the audience it is sharing has Favorite Color = Green, it can put the audience on a static list that represents Green but cannot actually update the field Favorite Color on any records that exist in Marketo already. In many use cases, this may not matter much, but if Marketo wanted to personalize an email on Favorite Color, this creates additional complexity in Marketo.

Fortunately, enhancements to the connector, announced at Summit this year, will allow AEP to update Marketo person records. However, it’s important to understand that for optimal importance of AEP and its destination workflows, Adobe has a soft performance limit (guardrail) that recommends no more than 50 data attributes be mapped to a single destination, like Marketo.  This is a soft limit, not a system-enforced limit, and you may not experience any performance issues by going over that limit, but it is important to note.

Similarly, as you’ll see in the same document, Adobe also recommends – as a performance guardrail – that no more than 200 total audiences be shared with a single destination. We’ve certainly had clients go well beyond these limits, but it’s important to understand that for Adobe, these are the guardrails and you would be required to accept any performance impact (most likely in how long it takes data to be delivered to that destination).

So some of my clients choose to focus AEP audience building on audiences that will be shared with more than one destination (Marketo and Target for example) and use Marketo to build audiences when only Marketo will be using that audience. (Don’t worry, AEP can still see all the marketing activities Marketo is doing when you connect Marketo as a source to AEP.)

What is the source of the data used to build the audience?

There will be many reasons that data that exists in AEP will also exist in Marketo. For example, if you are using Marketo forms on your website (or other forms but leveraging Marketo’s Forms 2.0 API), Marketo is going to know about that form fillout event as quickly as, if not more quickly than, AEP. Isn’t it more efficient to simply build the audience in Marketo with data it sourced itself rather than having AEP do it?

Let’s take another example. You’ll want to connect Marketo directly with your CRM even if that CRM is also connected to AEP. (Why? Remember that even with the enhanced AEP connector, AEP will only be able update the Marketo person record. But as B2B marketers, we all know that we leverage accounts and opportunities and other CRM objects in Marketo too (for personalization, for further flow step choices, etc.). So we want to keep that connectivity.

And it offers us an advantage – let’s take SFDC for example. The SFDC source connector to AEP is a “batch” connector. You can have multiple batches in a day, but it’s still batching. As we know, Marketo and SFDC have an always-on, bidirectional sync that starts again 5 minutes after it last ended. So Marketo’s going to get updates in SFDC often much more quickly than AEP and would always get it faster than AEP could share it with. In addition, any time audiences in AEP are built using objects like Account or Opportunity or

Program/Campaign Membership, those audiences in AEP are only updated once per day. So if you wanted to send an email in Marketo based on an opportunity reaching a particular stage and timeliness was important, you’d want to use Marketo to build that audience.

Who is building the audiences?

This one is important and I think it often is overlooked.

Marketo was designed by marketers for marketers without needing coding or data engineers to use it. Building smart lists is pretty easy in Marketo because Marketo has a very flat, person-centric architecture. All of the data objects and actions you want to use build audiences are right there on the right hand nav, grouped intuitively, with drag and drop functionality.

AEP is not a database, relational or otherwise. It collects data from various systems and ingests them as data sets that you can then use to build an audience. It also uses an intuitive drag and drop interface to pull attributes and events from data sets, but it isn’t – to me as a marketer – always as intuitive about where to find the attribute or activity I want to use. The data labels may not be as obvious to someone who isn’t a data engineer. It is an extremely ROBUST platform, as it should be, but is going to have a learning curve for marketers who are used to Marketo or really any marketing automation platform.  

With that in mind, if you make a decision to build many of your audiences in AEP, you will need to consider this:

• Is my organization willing to commit the resources to the additional enablement that will be required to get your Marketo users comfortable using AEP to build and activate audiences they need in Marketo (and other destinations)

or

• Do I have my AEP users build the audiences and is my organization willing to commit the resources to create the processes (including service level agreements – you don’t want bottlenecks) for requesting audiences be built and activated to Marketo

If your organization can’t commit to one of those two decisions, then I’d recommend that you’d use AEP for “big bucket” audiences only (our Top 500 accounts, our buyer personas, our opted-in prospects, our regions) and then have Marketo users use those “big buckets” and fine tune the audience in Marketo with additional filters, etc.

In conclusion, you absolutely can build nearly 100% of all audiences in AEP (there might still be edge cases where there is data AEP can’t actually ingest) but in my experience and considering the above, you’ll probably find that it’s not an All or Nothing proposition.