How to achieve personalization and customer retention with Segment and Snowflake

Companies who desire customer growth require personalization, but it's difficult to obtain accurate data. Real-time identity resolution through Snowflake and Segment leads to effective personalized marketing and customer insights.

By Segment

Personalization is one of the most effective strategies to grow a customer base and increase sales. In our annual State of Personalization Report, we found that 56% of consumers will become repeat customers after just one personalized experience. In fact, they expect personalization. 

Yet, 50% of brands find it difficult to get accurate data to power their personalization strategies. And 69% of companies plan to increase their investment in personalization, even though this challenge of collecting and ensuring high-quality data at scale remains unsolved. 

Luckily, with the right tactics and technology, businesses of all stripes and sizes can effectively implement personalization to strengthen customer loyalty. 

The challenge: Real-time identity resolution

Identity resolution is the process of stitching together individual data points into a unified customer profile. It’s an instrumental aspect of understanding your customers, especially as their interactions with a business become more multi-faceted (spanning multiple channels and touchpoints). 

For example, an average customer journey may look like this: 

1. The customer sees an ad on Facebook. They click the ad to visit the website and sign up for a mailing list for future promotions.

2. They get an email newsletter announcing a fantastic in-store deal and drive to the store to buy the product.

3. After using the product for a while, they see reviews for accessories and buy an accessory or two from the store's mobile shopping app.

There are several customer touchpoints in this journey, each with its own actionable data. The customer gives their email address on a website, their phone number on an app, and their customer loyalty card number in a brick-and-mortar store. 

Ordinarily, tying this data together to show that it all came from a single person would be cumbersome. Setting up a data infrastructure to crack the code is costly and complex, and stitching together customer profiles across multiple sources can take years.

And it’s never-ending. What happens when a customer creates another data event by calling customer service for help or sending a message to a social media chatbot? This new data must be instantly attributed to that customer for continued high-level service and engagement.

The dangers of not resolving identities

What do companies stand to lose by not having the complete picture of their customers?

Turns out, a lot: 

  • Small companies can’t stay competitive with large brands that can afford to create and host their own in-house data centers and solutions.

  • Without accurate, real-time insights, marketing spend is used in unproven ways. Science and data don't lead to decisions, so money gets wasted on campaigns that may not bring in any true ROI.

  • Customers feel unseen and unserved as they are increasingly accustomed to tailored digital experiences. They turn to alternative shopping methods for competing brands that use these highly engaging experiences.

  • Increasing data and marketing regulations make data governance difficult. Attempts to market with data without proper safeguards leave a company open to the risk of fines and reputational damage.

Without the big science data modeling available through tools like Segment and Snowflake, too many opportunities get left on the table. 

How to use Snowflake and Segment to drive personalization

Segment and Snowflake work bi-directionally; information can flow from one tool to another to be processed and optimized by each platform's technology. Here is an example of how it works:

1. Customer event data from your e-commerce and marketing tools gets cleaned and unified in Segment, which connects each data event to a Customer 360 profile for each shopper.


2. This profile can be sent from Segment to Snowflake, a business warehouse.


3. Once in Snowflake, additional matching and filtering occurs through the Profile Sync capabilities. This unites offline and custom data events with the data events that are already connected to the customer in Segment. A more comprehensive customer view forms.


4. Additional datasets found in the Snowflake Marketplace create more value for each customer profile. Weather data, historical industry trends, and demographic reports add more context to customer data and create truly actionable business insights.


5. With reverse ETL, teams can send data from Snowflake back to Segment for additional building of the customer profile. Teams can also send data directly to downstream tools through Segment's connection to over 450+ apps, marketing tools, and dashboard integrations.

6. Newly-enriched customer profiles drive personalized multichannel marketing campaigns.

7. As new data points are introduced into the pool, customer profiles (and the resulting campaigns) are optimized in real time. Instead of waiting months to update batches of data sets, the Segment-to-Snowflake and ETL cycles can be set to run as quickly as every 15 minutes for rapid results.


Use cases

Now that you have access to all this enriched customer data, you can use Segment and Snowflake to better understand your customers in unique ways. For example, you could:

  • Sync lead scores generated in Snowflake with sales engagement tools like Salesforce or Hubspot.

  • Connect audiences based on their demographics, past buying behavior, or product interest with marketing automation tools or CRMs.

  • Pass offline data to APIs on Google, TikTok, Snapchat, and Facebook, so ads get optimized quickly, and the return on ad spend is realized much sooner.

Brands can also use nearly complete customer data to fill in gaps in their historical sales data or customer interactions, a process known as operationalization. By using reverse ETL, these holes in a customer journey get resolved, and a holistic understanding of how and why a customer buys a product takes form.

More than just a profile resolution

Creating a complete customer profile is one of the most valuable ways to get the most out of Segment and Snowflake's partnership. Showing how each customer identity was built from separate data sources also holds unique promise.

The profile traits table set up in Snowflake and powered by the data from Segment's customer profiles does more than tell you what customers have done. It's the engine for highly actionable business insights you won't get from either Segment or Snowflake alone.

With the platforms up and running, you can see when and where each piece of event-level data was created, as well as any associated trigger events. This level of detail adds more context to the customer journey and helps you build segments far beyond traditional demographic data.


It goes beyond just direct customer-created data to offer information like:

  • Most frequent product category views

  • Total in-store spend

  • Lifetime value (LTV)

  • Revenue per visit (RPV)

  • Preferred buying method

High ROI on time and resources

Data, while incredibly valuable, also comes with its own costs. You have to pay to collect, store, govern, and share the data. Segment and Snowflake work together to minimize those costs and free up your teams to do other work. 

Segment and Snowflake offer data management capabilities to ensure clean and consistent data across your infrastructure so you avoid costly compliance risks. That means you can keep tabs on all customers with the incredibly accurate ID resolution models and manage their profiles to stay compliant with GDPR and CAN-SPAM regulations.

For companies that want to share some data (but not all) with partners and publishers, the Segment and Snowflake partnership offers a way to anonymize data through custom data access permissions. This is less labor-intensive than manually selecting the data fields you want them to see or creating new reports for each level of access. The data pipelines and cleaning tools offered hide data from those you don’t need to see it.

All of these solutions save engineers time and resources. Instead of building, onboarding, and launching, they can develop new and innovative ways to use data. Maintaining compliance with internal data governance policies is no longer tedious and slow. Setting up new data pipelines is as simple as using Segment Connectors and creating new data models within the Snowflake interface. This significantly reduces cost and labor. 

Segment and Snowflake: Better together

Using a standalone ETL, warehouse, and CDP can seem functionally similar. Your engineering team can set up rigorous data processes to connect the three and even get some insights. However, Snowflake and Segment are more effective.

1. Quickly add or remove sources

You may collect a lot of data, but do you want all that data stored in Snowflake immediately? By picking and choosing what data sources you sync to the data warehouse, you manage data volume, keep the data sets clean, and ensure you only store the data you need. It's a self-controlled way to manage costs and workloads. Setting up or removing new data sources takes minutes, not months.

2. Get true unification and customer profile stitching

As new customer data points are discovered, they can be integrated into the profile, so you continually build on what you know about each customer segment. It's an asset that grows in value over time, so your understanding of your customer gets deeper and more relevant to your day-to-day marketing activities.

3. Score your customers

Segment and Snowflake allow you to automate customer scoring into your marketing plan. Customer scoring models use first and third-party data in tandem with Segment's behavioral insights to determine a customer's next likely move. You'll be able to determine which channels work best for each cohort and if those channels are worth it to invest in.

For example, you’ll be able to learn if a customer is just as likely to engage with an email as a Facebook ad and which one is more cost-efficient.

Data insights tell you all of this and help ensure you don’t count all channels the same. A proper scoring model reduces wasted marketing budget and drives better ROI with every dollar spent.


4. Fill in gaps

Your business doesn't know what it doesn't know. Segment and Snowflake are built to help you figure this out. By operationalizing your data warehouse, data helps answer questions you didn't know you needed to ask. Backfill historical data gaps, use Snowflake Marketplace datasets to discover new-to-you trends, and test out new SaaS tools at a relatively low cost. You can afford to experiment with data, build new queries, and send insights to more partners and advertisers. Once you find a trend or insight worth exploring, you can monitor it in real time as customer events create new data sets. Even the smallest companies get access to these powerful insights.

Balancing customer and business needs

With technology creating more and more data to sort through, it’s tempting to take a strictly “data science” approach to marketing. It’s a valuable asset offering incredible ROI when used properly.

Thankfully, great marketers also understand the human element of sales. They are listening to the 56% of consumers who want to connect meaningfully with the brands they buy from.

With tools like Snowflake and Segment finely tuned to work together, there's no need to choose between technology and humanity. Personalized marketing is profitable, so when a customer needs change, you'll have the insights to change how you serve them.

Are you looking to get the most out of your customer data? Join the 25,000+ companies on the Segment platform today. Request a demo.

The state of personalization 2023

The State of Personalization 2023

Our annual look at how attitudes, preferences, and experiences with personalization have evolved over the past year.

Recommended articles


Want to keep updated on Segment launches, events, and updates?