^{This content was originally published by Hightouch. See here for the original article.}

Results

• With Snowflake, Imperfect Foods now has a centralized data platform for analytics where it can build robust data models to power the business.
• Thanks to Fivetran, Imperfect Foods has fully automated and optimized data pipelines for every data source to continuously feed data into Snowflake.
• Since implementing Hightouch the marketing team no longer has to download ad-hoc data sets out of Snowflake to run marketing campaigns across different channels.
• Using Hightouch to sync custom audiences powered by Snowflake to different ad platforms, marketing has reduced the cost of acquisition (CAC) by 15% and achieved a 53% increase in customer reactivations quarter over quarter.
  ‍

About Imperfect Foods

Founded in 2015, Imperfect Foods is the leading online grocer at the forefront of eliminating food waste and building a better food system. As a Certified B Corporation™, Imperfect Foods works directly with farmers and producers to rescue food and deliver it right to customers’ doors. To date, Imperfect Foods has rescued more than 159 million pounds of food from lesser outcomes. Customers enjoy a fully customizable service that is more affordable, and more environmentally friendly, than the average trip to the grocery store.
‍

The Challenge: Increase Signups and Product Usage

With hundreds of thousands of customers and an ever-increasing number of data sources, Imperfect Foods has a ton of customer data. However, acting on all of this information and understanding what traits lead to high-value customers, or what factors cause customers to order can be challenging. With no centralized view of the customer, Imperfect Foods needed a way to consolidate all of its customer data across its entire data stack to leverage it for marketing activation to increase signups and product usage.
‍

The Solution: Creating a Continuous Flow of Data with Snowflake, Fivetran, and Hightouch

Limited on engineering resources, Imperfect Foods was in search of technologies that were easy to use, fully managed, and self-serve. This realization quickly led the company to Snowflake, Fivetran, and Hightouch.
‍

Snowflake as a Centralized Data Platform

Having numerous different data sources across the business, Imperfect Foods turned to the Snowflake Data Cloud to consolidate all of its customer data into a centralized platform. 

“Snowflake has made it very easy to bring our data together from all of our different systems so we can report on key metrics and have a solid understanding of how our business is performing.” - Adam Smith, Analytics Manager

There were multiple factors that led to Imperfect choosing Snowflake, but ultimately, the simplicity and ease of use of the platform played a crucial role in the final decision.

“Snowflake is simple enough to use that a data analyst can set it up and it makes writing code so easy.” - Adam Smith, Analytics Manager

With Snowflake in place, Imperfect Foods is able to leverage dbt to build robust data models about the customer around order data, marketing spend, marketing attribution, etc. This data empowers the marketing team to deliver more personalized engagements, greater return on marketing campaigns, and deep insights to understand the impact of their efforts on business growth.
‍

Fivetran for Automated Data Pipelines
‍

Before any data can be accessed in Snowflake, Imperfect Foods first has to ingest all of its customer data across its many different disparate data sources and this is the exact problem Fivetran solves as a native ELT platform. This includes information from backend databases around orders and users, customer survey info from Qualtrics, accounting and inventory data from Netsuite, and spreadsheet data from various Google Sheets.

“Originally we were using Stitch to handle our data pipelining , but we decided to look at alternatives and Fivetran met our needs.” - Adam Smith, Analytics Manager

Fivetran is underpinning Imperfect Foods' entire data stack, and it’s secretly powering every use case across analytics and data activation. At its core, Fivetran increases the power of Snowflake for Imperfect Foods.

“Fivetran is a set it and forget it tool. After a quick initial setup, it continues to quickly and reliably transfer data from the source to Snowflake.” - Adam Smith, Analytics Manager

Hightouch for Data Activation

With Fivetran and Snowflake working in unison to create a continuous flow of data, Imperfect wanted to leverage the unique customer data in Snowflake to increase the number of signups and encourage more usage in the product. 

“We have all of this great data in Snowflake, but in order to get to that next level of performance we need to activate it across our marketing and ads platforms.” - Adam Smith, Analytics Manager

Accessing the customer data in Snowflake from a marketing standpoint was difficult though. Ad-hoc data sets were manually downloaded from Snowflake every time Imperfect Foods wanted to launch a marketing campaign. These problems quickly led the company to Hightouch for Data Activation.

“Our lifecycle marketing team would download reports out of Snowflake so they could send that data to Iterable and our ad platforms, but this wasn’t scalable and our team had to do this every single week.” - Adam Smith Analytics Manager

With Hightouch running in parallel to Snowflake and Fivetran, Imperfect Foods is able to automatically sync customer data in near real-time to Iterable for lifecycle marketing, and various ad platforms for lookalike audiences and retargeting.

This is extremely beneficial because customers have limited windows when they can order, so “time to shop” notifications are extremely important. By coupon testing with its most engaged customers, Imperfect Foods saw a promising reactivation lift of 53% compared to the previous quarter. With Hightouch, Imperfect Foods is able to build custom audience segments and sync them directly to ad platforms like Google, Facebook, and Tiktok leading to a 15% reduction in CAC (customer acquisition cost).

“Hightouch has made it really easy to sync our data in Snowflake to our various marketing platforms so we can optimize around our most successful customers. With Hightouch we can duplicate our audiences and send data to any destination of our choosing immediately.” - Adam Smith Analytics Manager

‍

What’s Next 

Since Imperfect Foods’ business model is closely correlated to the orders of its customers, identifying which trends and audience segments are more likely to lead to higher-order values is extremely important. With Snowflake, Fivetran, and Hightouch working together, Imperfect Foods’ optimization in this area will only continue to improve with time as the company is able to iterate and experiment with different audience segments and marketing campaigns.

‍