Matrix Data Architecture Load Process

Let’s consider a scenario where a Marketing Automation software provider wants to analyze Email Campaign data. We can assume that we get fresh data everyday -- implies one record per Campaign per day. The date element in this example can be the Campaign execution date, which means that data for Campaigns executed on 11/4/2014 would come in on 11/5/2014. We can also assume that we have a User identifier that corresponds to the person who initiated the Campaign. This Id, for the purposes of this example, can live on the Contact object.

With these assumptions, the data feed has to have the following four columns:

1. campaign_id (acts as the unique identifier for each Campaign)
2. execution_date (tells us when the Campaign was run)
3. emails_sent (tells us how many emails were sent as part of the Campaign)
4. user_id (tells us who ran the Campaign)

Let’s also assume that the end goal here is to get to the following:

• Marketing Campaigns run per Account per week
• Emails sent per Account per week

Now that we know what the structure is, we can configure a subject area in the MDA to store this data. This can be done using the Tenant Management App available internally and Gainsight Services will configure during your implementation. Let’s call it Sample_Data_Load.

Technically, you could just populate this data into the collection without any linkage to SFDC Accounts, since you can build reports on this data and now even push them to Gainsight Home. This clearly is not enough, and is therefore, not the recommended way to use Gainsight.

There are broadly three ways to create the Account linkage:

1. Provide us the SFDC Account Id directly.
2. Provide us some other field on the Account record that can uniquely identify the Account, such as an External Id.
3. Provide us an Id that lives on an object that is related directly to SFDC Accounts. For example, this could be a User Id or Email Id on the Contact object, or something like a Group Id that lives on a custom object that connects back up to Account.

In scenario #1, Account linkage can be created directly in the previous step. For the other two scenarios, we will have to treat this subject area as a staging site, and then perform lookups to create the Account linkage.

For this example, we are assuming scenario #3 (user_id lives on the Contact object, which then connects to SFDC Account). Following are the steps you can take:

• Navigate to Administration > Connectors > Data Load Configuration and click +Add New. Then create a new project on top of the Subject Area called Sample_Data_Load.

Let’s start with four records for now. We can name this file: campaign_data.csv.

Let’s start with a simple INSERT operation. UPSERT is not recommended, as it negatively affects our performance. Exceptions can, however, be made if the volume and frequency are low. Please discuss and confirm with Gainsight before taking the UPSERT route.

We can name this file: sample.json.

The fully composed command would look like this:

This command can be used in two ways:

1. Customers own extracting the data and pushing it to MDA. In this case, all the code is created and maintained by the customer. Gainsight can provide guidance, but will not own the data load piece.
2. Using Informatica Cloud’s post-processing feature. In this case, the customer owns extracting the data and loading it to the SFTP in the agreed-upon format and frequency. Gainsight will own and manage the rest.

Before we start aggregating the data further, let’s do a quick recap of what we have so far.

Subject Area 1: Sample_Data_Load