How to Link Digital Behavior to Your Customers
by Eric Matisoff
posted on 10-20-2016
I’ve been working in the data industry for just under a decade and am excited to have joined Adobe as the Data Science and Analytics Evangelist. Over the last 10 years, I’ve been leading midsize to enterprise-sized implementations, analyses, and optimizations. I’m looking forward to adding this experience to help you make the most of your data and our products.
My role as an evangelist is to work with you, the data enthusiasts. I’m here to help you better understand Adobe’s vision with regard to data and analytics, to clear up any confusion that you may have about the Adobe Marketing Cloud (AMC), and to spark conversation about where our rapidly changing industry is headed.
In this post, we’ll be diving into Customer Attributes. You probably remember hearing about these at Summit last year; but if you’re like me, you’re busy and may have delayed their deployment. Well, the time is now. I’ve seen customers of all shapes, sizes, and industries that need this feature.
Here’s the simple definition: Customer Attributes is an AMC-level feature that enables imports of offline data based on a customer identifier, often a login ID. You can push in demographics data, loyalty-program data, and any other customer-based data that may affect your business or your customers’ behaviors on your digital experiences. And, you don’t have to worry if you push an attribute that doesn’t end up bearing any fruit, as they’re very easy to delete so you can start over again.
Customer Attributes Versus Classifications
Think Customer Attributes are just like Classifications? Well, they’re similar, but here’s how they’re different:
- Like Classifications, Customer Attributes enables you to upload labels to behavioral data retroactively (historical and future data). However, unlike Classifications, Customer Attributes are applied to the visitor rather than a specific action that visitor takes. This means that the attribute is linked to the entire lifespan of the cookie. In essence, Customer Attributes will apply to all visits — even those prior to the login ID being captured.
- There are no cardinality limits with Customer Attributes, so you’ll never see ‘Low Traffic’.
- Customer Attributes are available through the entire AMC, not just Analytics. So, you can target and report on these attributes in Adobe Target or automate a campaign in Adobe Campaign via Customer Attribute-driven segments.
- Customer Attributes can be set as Dimensions and Metrics. I have an example of this below for salary, but you could similarly use it to count loyalty-rewards points, customer lifetime value, etc.
- Customer Attributes do not require an eVar or prop. They are, instead, pushed in via the AMC ID service. eVars and props can be used as your source though; ClientCare can help you get this set up.
Now that you’re up to speed on the benefits, I’ve created a step-by-step guide to getting started with them. I’ve also included some screenshots of Analysis Workspace at the end to show you what the final product looks like. So, let’s dive in.
Step 1: Prerequisites — Get Your Marketing Cloud in Order.
For your organization to leverage Customer Attributes, you’ll need to ensure you have a few things checked off — and don’t worry; each item is relatively painless, and ClientCare will be happy to help you move along if you bump into any roadblocks:
- Enable your organization for AMC;
- Link your analytics company to AMC;
- Enable the AMC ID Service on your Analytics-enabled site(s), preferably via dynamic tag management (DTM); and
- Ensure you are set as an admin for Customer Attributes in the AMC Enterprise Dashboard.
Once these steps are out of the way, it’s time for the fun to start!
Step 2: Identification — Identify Your ID.
One of the must-have requirements for enabling Customer Attributes is identifying your customer. This almost always comes from a user login. It will be your ideal key for tying online behavior to offline customer data.
Most likely, if you’re using a tool like Adobe DTM, you already have a Data Element set up to capture this unique user ID. In fact, my test example of DTM has a Data Element called “unique id,” which you’ll see below.
Step 3: Push Your ID to AMC.
1. Click the gear to open the settings for the AMC ID Service:
- Open the Customer Settings accordion, add a friendly name for your ID attribute in the first textbox (labeled Integration Code), add the Data Element you’ve set up to capture the unique user ID in the second textbox (labeled Value), and select Authenticated as your Auth State.
3. Once you have everything set up, it should look something like this:
Note that I used “loginID” as my Integration Code value. Remember the name you chose, as you’ll need it later.
4. Validate that your ID is being populated on your site by leveraging the Adobe Debugger. You’ll see two rows similar to the screenshot below listed under your SiteCatalyst Image:
Step 4: Create a Comma-Separated Value (CSV) That Ties IDs to Your Attributes.
It’s time to link your unique ID data to your attributes. Consider starting with just one; they can be deleted and recreated an unlimited number of times and are historically retroactive to the visitor’s first hit.
Start with a blank CSV. In cell A1 (top-left cell), place the Integration Code name that you used in Step 3 (“loginID” in my example). In cell B1, place a friendly name for the Customer Attribute that you’d like to upload; I’ve selected Gender. I’ve also added Billing Address State and Salary in cells C1 and D1, respectively. Finally, import your customer-relationship management (CRM) data into this format.
Now, save your file as a CSV and keep it handy.
Step 5: Create Attributes in AMC.
Congrats! You’re done pushing data into AMC at runtime; now, it’s time to create your CRM attributes and integrate the data. This is done within the Profiles & Audiences section of AMC, accessed via the Product Selector dropdown:
Once here, switch over from Audience Library to Customer Attributes at the top:
Select the New button and provide a friendly name and description. Place the Integration Code name that you took note of in Step 3 (“loginID” in my example).
If everything works well, the File Upload module will become available to you. Note that you have the option to upload Customer Attributes via an easy-to-set-up file-transfer protocol (FTP) interface. We’ll be going through the browser import, but both will work great for you.
Once you’ve dropped in your CSV, Adobe will quickly analyze the file and provide you with a Schema to validate. For me, it auto-identified all three of my attributes and set each as a string; however, I’m actually going to adjust Salary to an integer:
You also have the option to adjust the Display Name and Description for each Attribute.
Now, go ahead and hit Save. Upon doing this, Adobe will automatically start crunching the data and providing you with a real-time update on the integration status.
Finally, click to configure a subscription: select your solution, report suite(s), and attribute(s):
If you have Target, this is also where you would enable the Customer Attributes for Target.
Go ahead and click Active, then Save — and, you’re done!
Extra Credit: View the Data.
Okay, I know I said “you’re done,” but … well, why not check the data? Customer Attribute data does take some time to populate, so you may want to double check it.
In Analytics, Customer Attributes are stored in a folder of their own within the Visitor Profile. Here’s what mine looks like now:
Open a report, and you can confirm that the data has been populated. You’ll notice a couple of features, including:
- Updated current AND historical data. This means that data dating back to the very first visit will be tied to attributes for any visitor whose ID has been captured at any time in that visitor’s cookie lifetime.
- Customer Attributes set as a Number can be added as metrics. In my example, this means that I’ll be able to add the numerical values of Salary to any report! I can easily pull in Salary as a metric to see which marketing channels are driving visitation from the highest-salaried customers.
Here’s a screenshot of my data in Analysis Workspace:
Now go have some fun with Customer Attributes!