Brings data-driven decision-making to an industry
Establishes a unified data platform to build on
Simplifies data ingestion, analytics, and distribution
Enables straightforward AI adoption
Saves about $250,000 in one app's development costs
Tiny but mighty, blueberries continue to grow in popularity. According to the most recent National Economic Impact Study by Dr. Dennis Tootelian, in 2024 the blueberry industry's value hit $9.1 billion and was growing by more than 6.0% year over year. To sustain this rapid industry momentum, growers and suppliers — including fruit brokers, processors, and food manufacturers — need accurate insights into production, pricing, sales, and demand trends to improve forecasting and decision-making. However, accessing industrywide data remains challenging. Farming and wholesale operations traditionally rely on manual tracking, simple spreadsheets, and intuition rather than integrated, data-driven systems. Suppliers have also struggled to shift to data-driven business operations because it's been challenging to get accurate industry insights that include comprehensive data from growers, other suppliers, retailers, and government agencies that oversee imports, exports, and tariffs.
Recognizing that no one organization could overcome the industry's data and insight challenges, the North American Blueberry Council (NABC), established the United Blueberry Task Force. It brings together domestic growers, global suppliers, and industry leaders such as the U.S. Highbush Blueberry Council (USHBC) to collectively address issues and opportunities, and increase domestic growth by improving industry insights and forecasting. After constructive, solutions-focused conversations, the NABC decided to join forces with the USHBC to fund and develop a data and insights platform called BerrySmart Insights Fresh. It would be accessible to the USHBC members who provide data for the solution. “Blueberry importers came together to align on a clear direction,” says Joe Vargas, director of business intelligence for NABC. “We recognized the need to better understand what the data reveals about domestic supply and demand.”
Building a trusted data and BI solution for the blueberry industry
To make the most of the project's limited budget, stakeholders decided to initially use Snowflake as its solution database because they already used and liked the technology. For similar reasons, they decided to design and host their website with WordPress. Instead of going with their existing analytics tool, they chose Looker and engaged Bytecode IO to help with their solution design and implementation. “We had been using Tableau for analytics, but it had a lot of shortcomings,” says Vargas. These included inconsistent report data and difficulties publishing and sharing insights.
When we met with Bytecode IO, I thought what they were doing with Looker was fantastic. They showed us how we could use LookML to build a semantic data layer to aggregate data and deliver deeper insights at the row level.
Joe Vargas
Director of Business Intelligence, North American Blueberry Council
To build a data platform for BerrySmart Insights Fresh, the collaborative NABC and Bytecode engineering team developed data pipelines with Fivetran. The pipelines ingest crop and production sales data from industry suppliers Berry Fresh, California Giant Berry Farms, Driscoll's, Family Tree Farms, HBA Farms, North Bay Produce, and Wish Farms — as well as the USDA and SPINS, a retail data and analytics company. “We're bringing different datasets to understand supply and demand, including consumers' buying habits, shifts in preferences, shifts in demand by region, preferences for different pack styles, and factors like price points and household income sizes,” Vargas says.
Delivering data controls and personalized insights
Sitting on top of the data platform is a semantic data layer that the engineering team built using LookML in Looker. It provides one data model including definitions and KPI formulas, and it governs which users can access which data, down to the row level. This strategy ensures accurate, consistent insights from a single data platform even though it supports different users and BI types. Having one model also helps reduce duplicated data and simplify IT operations.
Being able to control and curate that data narrative, create data dictionaries, and immediately push through any changes to field names and calculations to all users with LookML is really impactful. I can't do that in Tableau.
Joe Vargas
Director of Business Intelligence, North American Blueberry Council
To authenticate users and provide them with a customized experience when they log in to the BerrySmart Insights Fresh platform, the engineering team built a front-end app with React and Looker Embedded. Based on the user's ID, the app delivers a personalized portal experience, including customized dashboards. “Essentially we have different portals for users,” Vargas says. “When users log in, they see their logo and their information against the aggregated data they're authorized to view.” As a result, shippers now have the data necessary to better understand the market, make informed decisions, and facilitate fact-based sales meetings.
Simplifying ongoing app development with Looker
The team also used Looker Embedded to create six apps that deliver specific types of information requested by particular BerrySmart Insights Fresh users. Today, authorized users access the apps from their portal view of the platform. One app, called The X, provides blueberry export data including a color-coded map and information about topics like tariffs and the chemical residue levels that different countries deem acceptable. Because many of the target users for The X app prefer to work with spreadsheets, the app supports Connected Sheets for Looker so users can explore export data in a Google Sheet.
The collective team built three other apps with Looker Embedded. Two provide information about members and suppliers. A third provides self-service data and insights using Looker Studio and Looker Explores, which are pre-built data queries. “People can filter on things like geography, taxes, growth methods, and other factors to understand how the market's performing,” Vargas says.
Rebuilding with Google Cloud to cultivate greater success
The NABC is now rebuilding the website for BerrySmart Insights Fresh to increase user insights, improve their experiences, strengthen authentication, and simplify development. The new site will employ the organizations' existing LookML model but will use BigQuery as its data warehouse instead of Snowflake. The site will also provide self-service tools such as Looker Explore Assistant so users can generate their own Looker Explore queries using natural language prompts and the data they're authorized to access. “We're using BigQuery to become more vertically integrated with the Google stack,” says Vargas. “With Google Cloud, we can consolidate and take advantage of the versatility of being on one platform for analytics. We can also use tools like Looker Explore Assistant to provide assistance to users when someone is not available to help them.”
Adding AI and other integrations to improve data usage and insights
To make better use of the data platform that's supporting the BerrySmart Insights Fresh platform data, the USHBC asked the NABC to develop an automated analytics solution using Orion from Gravity Foundation, which is an AI-powered data analyst built for Looker. The NABC will also use the solution. “I'm the only employee on our team that's in the data and analytics department, so it doesn't make sense to spend $100,000 on retail data if we don't have time to analyze it,” explains Vargas. “On top of that, I'm trying to elevate myself out of the day-to-day and analytics to focus on what's next. When I learned about Orion, I thought, ‘Oh my gosh, yes. How do we get more of that?’ We can use it to generate a weekly report and use that to create an article that we post in LinkedIn or other communication channels to improve how we push out information.”
The NABC is now looking at expanding the forecasting capabilities of BerrySmart Insights Fresh by ingesting weather and map data from Google Earth Engine as well as crop data collected with a solution from Bitwise Agronomy that uses remote IoT sensors and an autonomous four-wheeler equipped with a GoPro camera. “When you have a core foundational technology like Looker, it's much easier to adopt tools like Orion or Bitwise because it's just easier to move forward,” Vargas explains. “It takes away some of that fear around where to start. We have our one scalable source of truth. Now we can just bring in more technologies like AI to help blueberry growers and suppliers use that data to better understand the market and know where to invest so they can drive steady growth for years to come.”
We built this entire system with Looker, which was super cool. And we kept project costs below $20,000. Another vendor quoted this project at nearly $250,000, so we realized huge savings with Bytecode and Looker.
Joe Vargas
Director of Business Intelligence, North American Blueberry Council
The North American Blueberry Council (NABC) is a nonprofit organization that helps highbush blueberry growers, brokers, processors, food manufacturers, and other industry stakeholders address issues, establish standards, and drive growth and profitability. The NABC works closely with other industry organizations including the U.S. Highbush Blueberry Council (USHBC).
Industry: Technology
Location: United States
Products: Google Cloud, BigQuery, Connected Sheets for Looker, Google Earth Engine, Looker Embedded, Looker Studio
