Understanding the satisfaction of your existing customer base isn’t just a smart move, but a survivorship skill. Your existing customers are the key to your success. They follow your brand; they’ve chosen your products and trust your expertise. It’s no wonder why they’re so crucial to your bottom line.
What if customers are about to churn and you don’t have the right tools in place? Or the right insights readily accessible? And the right action framework to be proactive? These questions don’t have an easy answer.
In today’s business landscape, your CRM data is not enough to have the full picture. From website interactions to social media, from point of sale to fulfillment systems, from call centers to the Internet of Things (IoT), customer data is scattered across multiple touchpoints. But here’s the catch: missing the right insight at the right moment can be the difference between retaining a loyal customer and losing them to the churn abyss.
Bringing your data together with Salesforce + Snowflake
Admittedly, the journey to having readily accessible insights and the right action framework hasn’t been all sunshine and rainbows. But what if we told you that Salesforce and Snowflake have already implemented the technology to address these needs and much more?
Salesforce Data Cloud now enables you to access your Snowflake data via a Zero-ETL integration, so insights from disparate sources can be centralized where it matters most — in front of your sales, service, and marketing teams.
Why is this important? Let’s step back to our churn/retention discussion. Aside from the core capabilities of Data Cloud (integration, harmonization, insights, and activation), Salesforce has incorporated Einstein Studio, a powerful AI tool that enables predictive insights and generative AI in your business workflows.
With Einstein Studio, we can train an AI model to predict customer churn. However, training the model is just one part of this journey. With Data Cloud, you can ingest, harmonize, unify, and transform the data that will feed your model. Here’s what that process looks like for a wealth management firm, for example:
3 steps to getting data ready for consumption
Initially, we ingest the data from Salesforce CRM and Snowflake into Data Cloud. Then we follow a three-step process to make the data ready for consumption:
- Transform. An important step when training a model is feature engineering, which is creating the right variables that will help us make the most accurate predictions (e.g., last store visit, open cases, satisfaction score, CLV, etc.).
- Harmonize. Through this process, we map our objects from disparate sources into the Salesforce canonical data model. This way we harmonize the data so it aligns with industry best practices.
- Unify. Since our customer base exists on multiple systems, it’s crucial to create a consolidated dataset prior to training the model, to feed it with unified data and increase its accuracy.
Finally, we enter the model training phase. Salesforce provides the capability to train various types of models, including regression and classification models. It also allows for Bring Your Own Model (BYOM) from services like Amazon SageMaker or integration with major GenAI vendors. For our purpose, we aim to classify customers who are likely to churn, which makes a classification model the clear winner.
With a sophisticated AI model now at our disposal, we can predict customer churn with greater accuracy. However, the true value of this predictive power lies in the actions we take based on these new insights.
With tools like Flow Builder, Next Best Actions, or Journey Builder, you can empower your service reps to make proactive calls, improve your AEs’ visibility to make the right offer at the right time, or assist marketers in identifying and segmenting customers at high risk of churn — allowing for the creation of targeted journeys that cater to their specific needs.
Want to see Zero Copy/ETL in action?
Watch our on-demand webinar Salesforce + Snowflake: Unlocking the Power of Data and AI with Zero Copy Capabilities to learn how Salesforce and Snowflake’s Zero Copy/ETL data-sharing approach will give you more control to securely manage your data and build richer customer 360 experiences in Data Cloud.
Our data experts demonstrate how to connect your data in Snowflake to Salesforce and apply generative AI to your workflows. They also show you how to use Salesforce Data Cloud and Einstein to predict customer churn — and take quick, meaningful action on your accounts with Copilot.
