As your nonprofit invests in collecting first-party digital data to drive your fundraising, issue advocacy, and marketing program, you may find yourself overwhelmed by multiple data repositories, dashboards, and indicators. The customer data platform (CDP) technology being used by large corporations is equally useful for nonprofits that operate substantial and complex fundraising and engagement programs.
Deploying a CDP aggregates all the data you’re collecting into holistic user profiles that show a customer’s entire constellation of interactions with you in one place, with consolidated reporting and metrics you can use to make more precise marketing planning decisions.
New privacy standards are encouraging first-party data collection ― but making your first-party data actionable is a challenge.
Running a targeted and personalized action-oriented marketing program ― whether you’re fundraising, driving advocacy, or selling products or services ― depends on having access to reliable data about individual customers or prospects and their individual interests. But as companies like Apple, Google, and Facebook implement new privacy standards in response to consumer pressure, the sources of third-party customer data (like tracking pixels) that organizations have come to depend on are being curtailed.
In this more restrictive privacy environment, companies and nonprofits are developing first-party data strategies, so that they can collect customer data that they own and control, and use it for robust targeting and personalization. But building a rich and detailed first-party data set is a long-term project, and turning data into usable information is a whole separate challenge.
If you manage a marketing program for a nonprofit or advocacy organization that’s just setting out on the path towards a first-party-data-driven strategy, you are probably finding yourself drowning in data of highly variable quality from multiple sources, making it hard to convert that data into actionable metrics to steer the program by.
Even if you have a lot of first-party data, it’s probably not well integrated. Your CRM (customer relationship management) system stores supporter information and some transactional data, but doesn’t track every engagement with every email, text message, or other messaging touchpoint. You may have a marketing automation platform that tracks email opens and clicks. Your fundraising and e-commerce platforms track donation and purchase transactions, your advocacy platform tracks online actions, your events platform tracks who shows up in person, your website analytics show pageviews and content interactions, and your ad buys are spread across a range of display, search, and social networks.
When viewing all these data sources separately, not only can you not see all your reporting consolidated in one place, most of the data can’t even be easily mapped back to unique individuals across channels and platforms.
You may start out by designating one or two of those data sources as core or canonical, and build custom integrations to import data from other sources into those, in order to approximate an aggregate picture of a given customer and their interactions and interests. For example, many organizations that fundraise will build or deploy an integration to flow online fundraising transactions into the CRM. But custom integrations like those are only as good as the logic you build into them, they need to be built and maintained, and this method scales poorly as you add more and more sources of data.
With a CDP, you see everything in one place.
A CDP (customer data platform) is a centralized repository of overlay data with built-in logic that turns the paradigm around. Conceptually, it’s built around a single, central, unified user profile for each unique individual who interacts with your ecosystem, no matter how many of your channels they touch. It pulls in data from all your data sets about the individuals you interact with and creates a “super-profile” for each of them ― a 360-degree view ― which shows you the totality of their attributes, behavioral characteristics, and transactions and interactions all in one place.
That unified view of a supporter makes it easier for you to draw holistic judgments about how your communications and promotions affect their behavior, and what strategic decisions you can make across channels that will most likely raise their lifetime value, whether you measure that in funds raised, advocacy impact, or sales revenue.
And by monitoring aggregate patterns in behavior involving multiple channels, CDPs can even start to draw insights from anonymous visitors ― people who haven’t even identified themselves by signing up to receive emails or making a donation or purchase ― that you can use to confirm or improve the fundamental assumptions about behavior that drive your acquisition and conversion programs.
By applying matching logic, a CDP combines data from multiple sources.
Organizations just getting started will typically begin building out a first-party data program by matching records from multiple data sets based on email address. But even when a person has used the same email address across all your systems (which happens less often than you think), pulling in and combining that person’s multiple records from a dozen or more data sets is still tedious and error-prone. And it leaves out all the records that can’t be matched on email or phone number to an existing record in your CRM.
When you outsource the matching, identity-resolution, and aggregation logic to your CDP, all that labor happens behind the scenes. Your CDP can ingest existing profiles from all your existing data sets, and apply cleanup and matching algorithms based on shared identifiers (like email or phone number) to bring together the user’s multiple records into a single master profile. It can also collect data from and about site visitors and app users directly, and match it back to master profiles, thus itself serving as an additional source of first-party data.
A CDP generates inferences right away that you can use to drive personalization, even for new users.
If you’re attempting to build unified user profiles by matching secondary data sets (e.g., website pageviews) back to your CRM, it won’t work unless you can conclusively match records from multiple data sets. That’s because a CRM is meant to be an authoritative system of record ― a set of known facts about the people it houses: their names and addresses, their specific transactions, their known and confirmed topic areas of interest, etc. If we’re not sure that this browsing history belongs to this user in the CRM, we won’t put it in (and we shouldn’t!).
A CDP, on the other hand, is perfectly happy making loose inferences, so it delivers value even when direct matching isn’t possible ― and it starts delivering that value immediately.
Your CDP can apply machine learning to make probabilistic matches between profiles that share attributes and behavioral histories, which (because people with similar attributes and histories tend to behave similarly in the future) provides you with useful directional data even when the CDP can’t conclusively confirm a match. (And because a CDP is an overlay database ― it doesn’t actually merge records in the data sets underneath ― there is no damage to your underlying data if it makes a “wrong” inference.)
As you gather more information from individual users, you can refine the inferences and connections delivered by a CDP. But even based only on its initial AI-driven inferences, your CDP can help you provide a quasi-personalized experience in your users’ early interactions, before they’ve identified themselves or provided any information directly. This gives you a greater chance of reeling in prospective supporters sooner, while they are still in the initial stages of excitement about your cause.
Information about a user’s topic browsing history, for instance, can be used to infer their interests and deliver them more content in the interest areas they are already viewing, even before you know who they are. Or, you can use your inferences about their interests to deliver them a topically targeted signup opportunity that’s highly likely to successfully collect an email address. And when you do learn who they are, all the data you collected about their behavior while they were anonymous is instantly added to their user profile, as a retroactive interaction history ― not a single click is lost.