Brick-and-mortar businesses suffer from a chronic problem: they often lack mechanisms to identify and understand all their in-store customers. Unlike online transactions, there is no immediate record of who made what purchases, making it more difficult for these companies to launch personalized, targeted marketing campaigns.
Additionally, the gradual deprecation of the third-party cookie has placed a growing premium on first-party data, or data collected directly from customers, to support effective digital targeting strategies. Many businesses are turning to loyalty programs to source this data, but since loyalty penetration rates rarely reach 100% and often achieve only sporadic participation, they fail to form a reliable basis for truly understanding customer behavior.
These shortcomings represent millions in potential lost revenue, particularly as retail media networks continue to soar in popularity. Brands are spending upwards of $100B globally to advertise on these networks, with ads delivering a 70-90% sales margin. But retailers need a strong, scalable first-party foundation to deliver the granular customer insights, precision targeting, and accretive sales that brands and advertisers expect when they pay a premium to advertise on these networks.
Expanding first-party audiences with pipeline automation
Purpose-built for brick-and-mortar stores, the Bridg data and audience platform helps marketers expand their first-party audience for personalization and targeting by using market-leading offline identity resolution to recognize the individual behind the in-store transaction in a privacy-safe way. The Bridg platform, powered by Snowflake, delivers a single source of verified shopper truth in the form of SKU-level insights and longitudinal customer profiles to drive direct, privacy-safe engagement with target audiences.
With timely insights into their unknown consumers’ purchase behavior, Bridg clients can:
- Launch and scale retail media networks
- Get advanced analytics and insights
- Activate digital media and/or promotions
- Scale loyalty programs and drive conversions
Powered by Snowflake automated pipelines
Bridg uses an automated pipeline enabled by Snowflake to process and analyze customer data to increase speed and deliver a range of granular insights. Snowpark, Snowflake’s developer framework, empowers Bridg to pull in common data science and Python packages, unavailable through SQL, to develop proprietary models. As a result, Bridg no longer has to pull customer data from Snowflake to different environments to create effective machine learning (ML) models, instead relying on Snowflake as a central location for all data science activities.
Snowpark provides a way for Bridg to access data directly, train models, and execute actions all on a Snowflake cluster—making the process fully contained, automated, and efficient and enabling new targeting and personalization opportunities for clients. With its dynamic segmentation model, Bridg uses industry-leading customer attributes and lifecycle properties to identify customers that are likely to react well to a tailored marketing campaign, providing brands a way to target the right customers with the right message.
Dynamic segmentation
Traditionally, marketers have taken a “follow your gut” approach to identifying customer preferences and interests with limited demographic information. A marketing group might assume, for instance, that families with young kids would be the ideal target for prepackaged desserts—until you uncover that their purchases mostly consist of healthful, organic food products.
Instead of relying on inferences, the Bridg ML model uses clustering methods to analyze Bridg’s entire identity spine to pinpoint significant segments across a wide variety of demographics and affinities. Using the groups created by the model, Bridg clients can narrow their targeted segments to increase accuracy and optimize ad spend. One CPG client, for example, identified a previously unknown customer segment that only purchased newly released products, allowing them to only target this group during new product launches.
Extended customer attributes
Working with multiple partners and data sources, it’s critical to unite, process, and clarify discrepancies across all data sets. Using Snowpark, Bridg developed a statistical model that consolidates data sources and makes decisions. If, for example, four of their five data partners identify an annual salary range of $100,000-$150,000 for a shopper, while the fifth says $75,000, their statistical model determines the outlier and applies the most likely outcome—in this case, a salary range of $100,000-$150,000.
Additionally, Bridg collects and processes a wide array of customer interest data from surveys, previous shopping behaviors, life events, and other relevant individual information to track attributes for marketing. As a result, Bridg clients can access reliable and comprehensive shopper insights—from interests and demographics to home ownership status, car ownership, purchasing behavior, and beyond—on demand using Bridg’s analytical dashboards.
Customer lifecycle
Bridg clients can also use the platform to accurately track a customer’s lifecycle using recency, frequency, and spending habits. Leveraging Snowflake and specifically Snowpark, Bridg’s automated model identifies shopper trends so that clients can target customers to a greater degree of granularity. Lifecycle metrics allow brands to identify new, recurring, re-engaged, and lost customers to identify candidates based on campaign goals. Customer types like “active,” “lapsing,” “lapsed,” and “lost” help clients focus their campaigns on the right targets.
For example, clients could distinguish between shoppers who visit the store once every three months versus those that visit once a week, or track those who are exhibiting abnormal purchasing behaviors to anticipate their churn from the brand.
New levels of customer intelligence and efficiency with Snowflake
Bridg is helping retailers create new, high-margin revenue streams by expanding their high-value, first-party audiences and deepening intelligence across those audiences. With the support of Snowflake and Snowpark, Bridg data scientists are unlocking a greater depth and breadth of reliable, comprehensive insights at a more efficient pace—all while protecting consumer privacy in a secure and compliant manner.
“Together, Snowflake and Snowpark have allowed us to develop and automate proprietary machine learning models at a much faster pace,” said Dylan Sager, Lead Data Scientist at Bridg. “As a result, we are able to deliver a scale and depth of insight previously unavailable to brick-and-mortar marketers and their brand partners.”
