%402x.svg){kind=icon}
.svg)
The truth about data is that harnessing it correctly can massively impact your organization’s success.
Our client, a leading nutritional supplements retailer, recognized the potential of their data and partnered with Launch to leverage it to boost revenue, enhance customer engagement, and ultimately prepare for an AI-driven future.
They were looking to gain insights into their customer behavior, improve customer retention, and develop marketing campaigns that were not only effective but also measurable – a challenge in any industry.
A healthy challenge
With many retailers, enhancing marketing efforts and increasing customer engagement is at the top of the priority list. With our client, the challenge at hand was clear: how do you measure and improve the return on investment for marketing campaigns, with a particular focus on reactivating inactive customers and attracting new ones.
Paving the way for marketing campaign effectiveness
Launch developed data-driven analytics and machine learning (ML) models. These customized solutions delivered deep business insights, enabling the capture of inactive customers and generate new revenue through increased customer loyalty.
The first step to success was an initial Proof of Concept (PoC) phase, where Launch meticulously evaluated and explored sales, customer, and product data. Exploratory Data Analysis (EDA) uncovered factors related to customer retention and renewals with predictive potential.
These insights served as the foundation for training ML prototype models to predict customer churn and renewals. The PoC approach quickly demonstrated the feasibility and value of implementing ML solutions.
The positive outcomes from the PoC naturally paved the way for a Phase 2 project aimed at implementing a production ML model. Collaborating with our client's senior stakeholders, Launch pinpointed a priority use case: a marketing campaign targeting the re-engagement of inactive customers.
Thinking beyond a single campaign, Launch introduced a comprehensive approach to customer identification, segmentation, and scoring. This enabled the generation of a set of inactive customers to target with a campaign, as well as a method for assessing the campaign's effectiveness across various dimensions.
This data science approach helped understand the impact of different marketing campaigns and make better decisions about how to achieve their goal of higher customer retention.
Partnering for success
A pivotal aspect of the work done was a technical partnership with Snowflake: the integration and utilization of Snowpark at the project's core.
Snowpark, an innovative data science platform, empowered the team to write and execute Python code for training and running machine learning models, all seamlessly within Snowflake.
This integration eliminated the need to rely on external platforms like Azure or AWS, aligning perfectly with our client's preference for leveraging their substantial investment in Snowflake.
Snowpark offered several key benefits that contributed to our success:
- Scalability: Snowpark's ability to leverage UDTFs (user-defined table functions) allowed us to train hundreds of thousands of models in parallel in minutes, significantly enhancing our efficiency and speed in delivering results.
- Ease of Use: With Snowpark, our team could continue to use Snowflake as their preferred Integrated Development Environment (IDE) while utilizing Python as their primary ML language. Snowpark also facilitated the deployment of ML models and inference pipelines, streamlining the development process.
- Security: One of the critical advantages of Snowpark was its robust security features. No data ever left Snowflake, ensuring the confidentiality and integrity of our sensitive information throughout the ML process.
- Transparent Cost: Another significant benefit was the cost transparency offered by Snowpark. Our client incurred no additional fees from cloud providers when engineering with Snowpark, allowing for precise budget management and cost control.
Ultimately, Snowpark equipped Launch with the agility to explore and harness the capabilities of emerging technologies, providing a distinct advantage in our data-driven efforts.
The Data Science and AI group at Launch was instrumental in delivering a streamlined, scalable and most importantly, cost effective solution for our customer. This effort was a great demonstration that, with tight partnership and collaboration, Snowpark enables organizations to deliver valuable machine learning workflows at a very rapid pace, using the language of their choice, within the secure perimeter of Snowflake. I would recommend Launch to any of my customers looking to take full advantage of Snowflake’s capabilities for data science, AI and MLOps.
- Brandon Whited, Account Executive at Snowflake
Becoming AI-Ready
Achieving an 8% return rate in the treatment group, compared to 4.9% in the control group, resulted in a substantial 3.1% campaign-driven lift. However, this success isn’t an isolated instance.
Our client has now established a data-driven segmentation model that reaches beyond optimizing their current processes. This data capture and analysis method serves as a North Star, enabling their teams to make informed decisions in future campaigns.
Launch's partnership wasn't solely about delivering results; it was about fundamentally transforming the way this retail organization operates. With a commitment to enhancing their internal data capabilities, the team is in the process of building a Data Center of Excellence and even recruiting in-house data experts to prepare for the potential impact of new AI technologies on their organization.
Ultimately, this transformation is about more than just data; it's about gearing up for an AI-First future. Clean and well-structured data are the foundation for successful AI implementation and crucial for organizations who want to be prepared for and competitive during the AI Revolution – that’s the AI Imperative.