M1 Finance Securely Powers 100+ Marketing Audiences with a Composable CDP

Results

Accelerated marketing velocity

• Built 100+ marketing audiences in Hightouch
• Improved targeting capabilities by increasing customer data attributes from 150 to ~750+
• Reduced campaign execution time from 3+ months to mere minutes

Improved operational efficiency

• Saved 3-5 hours weekly in data pipeline maintenance
• Reduced engineering effort by 80-90% for marketing requests

The Challenge

With millions of users leveraging the platform every day, M1’s marketing team is constantly trying to move users through the marketing funnel and drive them to the next best action so they can get more value out of the app and ultimately build wealth.

M1’s marketing team wanted to build custom audiences for personalization, but their existing data architecture was heavily reliant on data engineering teams to make changes. With each change requiring work and coordination across multiple teams, engineering couldn’t support the pace at which the marketing team wanted to iterate.

Coordinating engineering changes wasn’t the only challenge. Much of the data needed for these campaigns resides in M1’s data warehouse, which houses customer data across accounts, users, and products.

To run effective campaigns and experiments across different audience cohorts, M1’s marketing team needed a secure way to build self-serve audiences using their data without waiting on engineering resources. This realization led M1 to adopt a Composable CDP using Hightouch.

The Solution

M1's Composable CDP Architecture

Security First

M1 takes security extremely seriously and applies exceptionally strict privacy and data access controls to uphold the company's privacy policy as well as applicable laws and regulations. All software that the company implements must do the following:

1. Ensure that data is not stored or transmitted to an unapproved party
2. Ensure that data is encrypted and not accessed by any inappropriate internal or external party
3. Maintain a robust and tamper-proof audit trail of all access to sensitive data

The M1 data team considered a variety of internal and external options when looking for ways to improve the marketing workflow. While the internal options maintained the existing level of security, they didn’t reduce the engineering overhead required to make changes, and many of the external vendors that were evaluated required data to leave M1 or connections to M1’s infrastructure that wouldn’t maintain the necessary levels of security.

Hightouch provides several security benefits to the data team. Because Hightouch operates on top of the data warehouse, no data leaves M1’s cloud infrastructure before it is used for marketing activation. Likewise, the ability to set up reverse SSH tunnels to create a virtual private cloud (VPC) ensures that M1 can create and own the tunnel that connects Hightouch to the data warehouse. This gives M1’s data team full control and governance over access to M1 networks.

In addition, Hightouch’s highly customizable role-based access controls (RBAC) enable the data team to take a “least privilege” approach to access management, thus ensuring that any employees using Hightouch are only working with the resources necessary for their role.

Reducing Engineering Overhead

Prior to Hightouch, the M1 data team was using homegrown scripts and pipelines to move data to Salesforce Marketing Cloud (SFMC). Operationally, the solution had reached the limits of its scalability. The data team was spending an increasing amount of time on maintenance and troubleshooting, and consequently, all new marketing requests faced delays.

In addition to the core data ingestion and modeling work that the data team was doing, they maintained a separate set of data models tailor-made for marketing. This meant that as new data was added to the warehouse, additional work was always required to meet marketing’s needs.

Everything in M1’s data warehouse is modeled in dbt, and the adoption of Hightouch provided an opportunity to re-think how data is modeled for marketing. Rather than continue to maintain a custom set of marketing-specific data models, Hightouch allows the data team to expose these models to marketing so they can use them for campaigning purposes. While this hasn’t entirely eliminated new marketing data requests, most of the data that marketing needs is now already planned and scoped, so little to no additional engineering resources are dedicated to marketing-specific asks.

Since adopting Hightouch, all of the in-house pipelines have been replaced with automated Hightouch syncs, and the integration with M1’s core dbt models has all but eliminated marketing-specific engineering work. The data engineering team has seen an 80-90% reduction in marketing-specific tasks, and migrating to all Hightouch has saved the data team an average of 3-5 hours per week on pipeline maintenance.

Unlocking Self-Serve Audiences

With a vastly improved and much more flexible approach to managing and activating data, the final piece of the puzzle was ensuring the marketing team could easily segment audiences of users for various campaigns. With Hightouch Customer Studio, the marketing team can easily self-serve and build audiences using a no-code UI, and the number of audiences the team manages through Hightouch has continued to grow steadily since the initial rollout in 2022.

Audience adoption from 2022-2023

M1 is now powering over 100 different audiences through Hightouch, and the number of properties that the marketing team can use to segment and build new audiences has increased from 150 to over 750 unique attributes.

Hightouch syncs these audiences automatically to SFMC to power a number of important lifecycle marketing use cases to orchestrate customer journeys and touchpoints. This includes email marketing campaigns like:

• The Next best action: Encouraging users to set up recurring deposits after they’ve funded their account
• Brokerage transfers: Offering special promotional incentives for users who migrate savings to M1.
• Personal milestones: Notifying users when market changes impact their investments or when their portfolio value reaches a certain threshold