Intro to Data Cloud: How to Navigate Data Ingestion and Identity Resolution Like a Genealogist

Everything old is new again. And also, everything new is rooted in history.

One of Salesforce’s newer offerings is Data Cloud. This is a system for unifying information from a single organization’s multiple different Salesforce CRMs and other data sources, and consolidating the data so that you can act on the data with an appropriate message via the appropriate contact details at the right time, and not multiple times and in multiple ways to what is ultimately one real-life recipient. I’m sure you’ve all experienced this before, when a company that you have purchased from or donated to in the past sends you multiple texts, emails and social media ads on the same topic. Or, if you’re like me, and you have changed your email address and your last name, then you might get identical email messages at multiple email addresses. It’s annoying. And for the company, all those contact points might be inflating their expectations of returns on all of their efforts. They thought they were emailing 1,000 unique donors, but maybe it was only 750.

This sounds complex, right? It sounded that way to me too, at least at first. Then I realized that the basic path for Data Cloud is fundamentally the same as the process that I and many others have used for years in genealogy research. (When I became a genealogist years ago, I recognized that the genealogical research platforms are just customized databases. But that’s a different story.)

In this article, I have summarized Data Cloud’s standard path, and then I have provide a few specific ways in which the steps line up between database management and genealogical research. The intent isn’t to teach you Data Cloud; there are plenty of other resources for that. The point is to provide a way of thinking about the process that is different from any of the other examples I have seen, in case this helps to deepen your understanding too.

Data Path

• Connect: gather / ingest data
• Harmonize: arrange and map data onto one standardized data model
• Unify: select the appropriate value for the appropriate use
• Analyze and Predict: use analytical tools to gain insights
• Activate: act on data with automated processes

Gathering Data

Data Cloud starts with ingesting data from multiple sources, mapping it to appropriate fields on a theoretically universal profile, then deciding whether different bits of info belong to the same person, and adding snippets of knowledge in layers over time. That time horizon can be years for more static data like names and mailing addresses, or days and weeks for more active data sources like email clicks and purchases. It’s the same idea in genealogy: researchers and family tree software providers start with reviewing documents like birth certificates and census records, and normalizing what information is provided so birthdate on one document and census date on another both are mapped to a universal field like Event Date. The analogy continues with figuring out whether the person listed on the 1920 census as Mary Elizabeth, age 18 is the same as the person listed on a marriage certificate from 1925 as Elisabeth Maria, age 24. This process of grouping similar data together via matching, and then choosing the best value of several available options, is known as identity resolution in Data Cloud.

Credit: Salesforce Trailhead on creating unified individual records 

Arranging Data

Data Cloud allows you to arrange your data into data spaces, so information ingested into one data space is not used to add to unified profiles in another data space. This is useful for keeping branding or regions clearly separated. In genealogy, you might want to have a data space known as a family tree that is separate for one branch of your family vs another, or for a family tree you are researching for an unrelated friend or historical figure.

Segmentation is Data Cloud’s process of grouping individual profiles that share common criteria into one data payload. You might want to look for all people who donated at least $10,000 over the last three years to a specific set of campaigns. The genealogy equivalent would be to look for people who lived to an age of at least 50 and had at least one child.

Data Cloud segments can consider direct attributes like age and birthdate, just as genealogists may want to know specific data points like age at first marriage. Data Cloud can also segment on related attributes, such as a series of social media account handles, while genealogists might want to know a list of a man’s military service muster rolls, or number of census residence locations. I have a segment in my family tree for women in my matrilineal line who have had twins. (This propensity skips a generation or two, because my mother and great-great-grandmother had twins, but neither I nor my grandmother did.)

Acting on Data

The payoff for all of the work of mapping data sources and identifying segments in Data Cloud comes when you can take action on the data, and send personalized, targeted messages to individuals. This could be useful for letting customers know when a new version of the shoe they have purchased at least twice in the past has become available. Or Data Cloud could be set up to notice that a client has purchased a refrigerator in the past and then reviewed several related pages in the Help section of the manufacturer’s website more than once in the past three days, and automatically open a service case on their behalf. In a similar way, I receive alerts within my Ancestry.com family tree when a new record has been added that may match to an existing family member’s profile. Maybe in the future there would be a way to have records be automatically added to the profile if certain criteria are met.

Conclusion

All of these analogies and similarities have helped me quickly gain an understanding of the power and possibilities of Data Cloud. In fact, I am pleased to say that I was recently certified by Salesforce as a Data Cloud Consultant. So if these examples have sparked your interest, or you have further questions about how Data Cloud could work for your organization, please use our Contact Us page to let me know how to reach you!

Here’s a photo from about 1905 of my great-grandmother Martha (recorded variously as Martha, Mary, Marta, and M A on census records) and her younger sister Elizabeth (known as Elizabeth, Elisa, Liz and Lizzie).