Somewhere in the world right now, there’s an email marketer who can't build the customer segment she needs for the campaign she wants to run. She’s asking a data analyst to pull the user list manually. That analyst will likely write an ad hoc SQL query, export the results to a CSV file, and send the file back to the marketer. The marketer will then upload the CSV into her ESP or marketing automation tool, double and triple-check to make sure the data is “just so”, before firing off the email to tens or hundreds of thousands of users.
List pulls are tedious for everyone involved. Analysts find running ad hoc SQL queries mind-numbing, and marketers would much rather pull data themselves than wait around for someone else. Most modern ESPs, CDPs, and MAPs claim to solve this problem with no-code segment builders, but I've yet to meet a marketer who can truly self-serve and is super happy with their tools.
The real question is, why are so many segment builders failing on their promises?
The biggest reason existing segment builders fall short is that they don’t have access to all the customer data. While many products ship with client SDKs that capture customer events, cloud data warehouses have become the center of gravity for all customer data: this includes transactional data, behavioral data, 3rd party SaaS data, and predicted traits. By contrast, existing products only capture a subset of customer data, typically user events and attributes. Without access to all customer touch-points and a single view of the customer, your segments are at best incomplete and at worst detrimental to the metrics you're trying to drive.
The other problem is the data you do have isn’t reliable. For example, you may want to target active users based on revenue within the last 60 days, but both active users and revenue definitions can be complex and require combining multiple data sources. The source of truth for these metrics is the data warehouse, where data teams actively and continuously clean, enrich, and manage customer data and its downstream metrics.
In both cases, you have to fall back to the data team to pull a list of users from the data warehouse directly.
Another limitation of today's segment builders is they rely on very limited, user-centric data models. All underlying data must be stored as either a user attribute or a user event.
In practice, every business has its own unique data model, represented by business entities and the relationships between them. For example, e-commerce stores have product catalogs, and online marketplaces have buyers and sellers. Most traditional tools aren’t able to support non-user-based objects, let alone the complex set of joins required for more targeted segmentation.
A data representation of a simple e-commerce store, not supported by today's segment builders.
As an example, consider a retail store who might want find all users who had a shopping session in the last 90 days and added an item to their cart that was out of inventory. It’s difficult, if not impossible, to achieve this in today’s segment builders without completely remodeling your data or asking engineering to do a bunch of work instrumenting more custom events.
The same problem shows up all the time in the lack of support for account hierarchies for B2B SaaS businesses. A common setup for enterprise software has users belonging to multiple accounts, and/or workspaces. But it’s often difficult or impossible to model these types of relationships in CDPs, MAPs, and ESPs. As a marketer, you’re forced to do one of two things: associate users with only one account (and throw away information), or create duplicate user profiles for each unique user account combination (which creates complications managing things like user profiles and unsubscribes).
Lastly, sometimes marketers resort to list pulls because marketing platforms just don’t store the data for as long as you need it. Many products have relatively short data retention policies and will delete certain high-volume datasets after a certain amount of time.
For example, Braze drops custom event properties after 30 days, and Marketo deletes many email campaign events after 90 days unless you pay more for a longer retention policy. It's true that for vendors, it’s expensive to store and process ever-increasing volumes of customer data. But for marketers, it means any customers segments that depend on historical data beyond a few months cannot be built through the segment builders and instead require ad hoc lists.
The common thread between all these challenges is the data management limitations of 3rd party systems. These 3rd party databases cannot access all your customer data, are not the source of truth for the data they do have, cannot support modern businesses models, and are limited in how much data they can store and how long they can store it. Compared to your existing data warehouse, they are less flexible, less scalable, and less trustworthy!
Fortunately, there's a better way.
Supergrain’s segment builder connects directly to your data warehouse, which means it queries your source of truth customer data. It’s schema agnostic, which means it natively supports all the entities and relationships that are unique to your business (even really complex account hierarchies!). And because Supergrain doesn’t make a copy of your data, the only limit to how far back in time your segments can look is the amount of historical data you’ve stored yourself.
It’s time we ship segment builders that actually deliver on their promise of enabling marketers to create flexible user segments for targeting, without the frustration of list pulls, manual CSV uploads, and wasted time.
And while analysts everywhere can get behind “death to list pulls”, the real winner when teams can easily build the customer segments of theirs dreams is marketing creativity. When time from idea to segment (to campaign) is reduced from days or weeks to minutes or hours, marketers can finally focus on building the strategy, creating the campaigns, and breathing life into potentially game-changing ideas that previously only lived in drafts and customer journey slides.
Supergrain’s segment builder makes this world possible today, and we'd love to show you. If you’d like to see how it works, contact us to get a demo.