SunButter Case Study: Building a Snowflake Data

Warehouse on AWS to Forecast Business Performance

SunButter’s existing data program needed to evolve to meet the demands of their growing business and introduce a “single source of truth” for data across a variety of internal and external sources. North Labs helped SunButter design a best-in-class Snowflake deployment on AWS to automate data ingestion, preparation and reporting.

About SunButter

SunButter is a brand of sunflower seed butter, promoted for use by people with nut allergies. It is an edible food paste similar to peanut butter, mainly used as a sandwich spread by people with peanut allergies and/or tree nut allergies.

Overview

Our cloud experts evolved SunButter’s data culture by centralizing their data in a Snowflake data warehouse hosted on AWS and migrating their reporting infrastructure from spreadsheets to PowerBI.

The Problem

  • Sales leadership spent 21+ hours a month combining sales reports, industry trends, retail data, and wholesale inventory spreadsheets for analysis via pivot tables and manual extraction
  • Inventory monitoring at the store and wholesale levels was based on “gut feeling” and trailing monthly reports

The Goal

  • Eliminate manual intervention and data prep
  • Create a source of truth with internal and external data for reporting & ad-hoc analysis
  • Leverage data sharing for industry trends, retail data, wholesale inventory, & sales performance
  • Leverage AWS ML Insights and Forecast for future inventory and sales predictions instead of relying on “gut feeling”

The Solution

We utilized the following technologies

  • AWS (Route 53, SES, Lambda, Glue, Athena, S3, Forecast, Eventbridge)
  • Snowflake
  • Matillion ETL for Snowflake
  • Power BI

The Outcome

  • Files are automatically uploaded and processed using email extraction on a weekly basis, transformed into a usable form for querying, and mirrored in Snowflake to be leveraged in the BI tool of preference
  • With all data in one place (“single source of truth”), business stakeholders are able to make informed decisions regarding future order fulfillment and region/store-specific sales while eliminating days of work formatting and combining spreadsheets for analysis
  • ML Insights and Forecast provide trend analysis and predictions across horizontal (changes over a period of time) and vertical (changes to a particular column or set of data) analytics

SunButter’s existing data program needed to evolve to meet the demands of their growing business and introduce a “single source of truth” for data across a variety of internal and external sources. North Labs helped SunButter design a best-in-class Snowflake deployment on AWS to automate data ingestion, preparation and reporting.

About SunButter

SunButter is a brand of sunflower seed butter, promoted for use by people with nut allergies. It is an edible food paste similar to peanut butter, mainly used as a sandwich spread by people with peanut allergies and/or tree nut allergies.

Overview

Our cloud experts evolved SunButter’s data culture by centralizing their data in a Snowflake data warehouse hosted on AWS and migrating their reporting infrastructure from spreadsheets to PowerBI.

The Problem

  • Sales leadership spent 21+ hours a month combining sales reports, industry trends, retail data, and wholesale inventory spreadsheets for analysis via pivot tables and manual extraction
  • Inventory monitoring at the store and wholesale levels was based on “gut feeling” and trailing monthly reports

The Goal

  • Eliminate manual intervention and data prep
  • Create a source of truth with internal and external data for reporting & ad-hoc analysis
  • Leverage data sharing for industry trends, retail data, wholesale inventory, & sales performance
  • Leverage AWS ML Insights and Forecast for future inventory and sales predictions instead of relying on “gut feeling”

The Solution

We utilized the following technologies

  • AWS (Route 53, SES, Lambda, Glue, Athena, S3, Forecast, Eventbridge)
  • Snowflake
  • Matillion ETL for Snowflake
  • Power BI

The Outcome

  • Files are automatically uploaded and processed using email extraction on a weekly basis, transformed into a usable form for querying, and mirrored in Snowflake to be leveraged in the BI tool of preference
  • With all data in one place (“single source of truth”), business stakeholders are able to make informed decisions regarding future order fulfillment and region/store-specific sales while eliminating days of work formatting and combining spreadsheets for analysis
  • ML Insights and Forecast provide trend analysis and predictions across horizontal (changes over a period of time) and vertical (changes to a particular column or set of data) analytics