Great food brands don’t guess what customers want—they know. Conagra’s demand science team knew it would take more than dashboards to predict the science of food cravings. 

The team of analysts and behavioral scientists is led by Manav Purohit, Head of Business Intelligence. Before Sigma, reporting was scattered across static dashboards and disconnected tools, forcing brand managers into tedious processes just to get answers. Today, with Sigma’s data apps, Conagra has replaced manual reporting with a single, collaborative platform that puts real-time insights and self-serve data entry in every brand manager’s hands.

Life Before Sigma

Before adopting Sigma, the demand science team operated in a fragmented, manual environment. Reporting and collaboration were spread across static dashboards, spreadsheets, and disconnected tools. Analysts and stakeholders were forced into inefficient workflows that created friction at every turn—reviewing reports in one system, logging updates or corrections in another, and relying on IT for even small data changes.

A lot of the ETL work was falling on the IT department or ourselves to go in and enter a script manually.
_{—Matthew Henkel, Senior Analyst}

“A lot of the ETL work was falling on the IT department or ourselves to go in and enter a script manually,” said Senior Analyst Matthew Henkel.

This created recurring bottlenecks, especially during key reporting periods. The team worked with a variety of complex data types—including search trends, reviews, and menu data—much of it semi-structured or unstructured. But without a unified environment, integrating those sources and applying business context required repeated manual effort.

Brand managers, meanwhile, had limited ability to interact with data directly. When they needed to input marketing events, annotate results, or override assumptions, the requests had to be funneled through the BI team. That limited agility, slowed decision-making, and increased the risk of data loss, especially when projects changed hands or team members rolled off.

Life with Sigma

By building a data app in Sigma, everything changed. With the ability to write data back to the warehouse, a unified app could truly answer every question.

Sigma data apps are eliminating the boundary between data entry and data consumption.
_{—Manav Purohit, Head of Business Intelligence}

“Sigma data apps are eliminating the boundary between data entry and data consumption,” said Purohit. Instead of toggling between reports and offline tools, users can now view data, provide feedback, and input new events—all within a single, governed environment.

Centralized, Interactive Reporting

One of the most impactful applications of Sigma has been the creation of a brand-centric reporting tool. Designed to surface micro and macro-level trends, this dashboard allows teams to compare brand-specific promotions with broader external factors like weather patterns, mortgage rates, and CPI shifts.

Users can toggle between brands and sub-brands, change timeframes, and immediately see the results reflected in the dashboard—making trend analysis faster and more intuitive.

We are moving a lot faster than we were before.
_{—Manav Purohit, Head of Business Intelligence}

“We are moving a lot faster than we were before—everything is being built within the construct of a report or dashboard, which frees up time and increases developer productivity,” said Purohit.

The tool also detects volume anomalies and flags periods where sales performance deviates from historical expectations, helping brand leaders zero in on what’s working—and what’s not.

Empowering Brand Teams with Self-Service Input

Conagra’s Input Tracker, built as a Sigma Data App, gave Conagra’s brand managers direct control over their data. Instead of routing requests through BI or IT, they now log marketing events—like limited-time promotions or distribution changes—right into the dashboard.

This change eliminated a major workflow bottleneck and gave business users more ownership over the data they rely on.

It really takes away that middleman piece and ultimately saves both sides time.
_{—Matthew Henkel, Senior Analyst}

“With Sigma and specifically the input tracker, it allows the brands to own that space and own that data—it really takes away that middleman piece and ultimately saves both our side time,” said Henkel.

The new setup also ensures better continuity. Inputs are now stored in a centralized, governed location—reducing knowledge loss and preserving historical context as people move between roles or projects.

Increased Developer Productivity

For BI and data engineering teams, Sigma also reduced the overhead associated with transforming and maintaining input-heavy datasets. The team can now handle input directly in the reporting layer, cutting down on time spent writing and managing scripts or manual ETL processes.

Being able to take unstructured data and create views or procedures has saved us a lot of pain.
_{—Matthew Henkel, Senior Analyst}

“From a developer standpoint, being able to take that unstructured data from brand teams and create views or procedures off of input tables has saved us a lot of pain moving forward,” said Henkel.

Smooth Enterprise Adoption

Rolling out a new BI tool across multiple departments can be a heavy lift—but Sigma’s intuitive, user-friendly interface helped Conagra accelerate onboarding and minimize friction. Business users who were previously reliant on static reports or spreadsheet-based workflows were able to transition into Sigma with minimal training.

It’s been very easy for users to get onboarded.
_{—Matthew Henkel, Senior Analyst}

“From a UX experience, it’s been very easy for users to get onboarded,” said Henkel. “A lot of pain points we run into are around adoption, but with the new modal functionalities and the input tracker, it’s been very seamless and helped us roll out this tool more enterprise-wide.”

Because Sigma combines reporting and data entry in the same environment, it eliminates the need for separate tools or technical handoffs. That simplification helped boost confidence and self-sufficiency among brand teams and other non-technical stakeholders. Because the data is written back to Snowflake, all inputs and interactions remain secure, governed, and up to date—further supporting trust and scalability.

Future Plans with Sigma

Looking ahead, Conagra plans to expand Sigma’s footprint beyond the demand science team by embedding data apps and self-service workflows across more departments, including retail finance, sales, and marketing operations.

As part of that expansion, the team is particularly excited about integrating AI functionality using Snowflake’s Cortex. By embedding large language models directly into reports, Conagra aims to provide dynamic, context-aware recommendations that can be modified or overridden by users in real time—blending machine insight with human expertise.

This vision aligns with Conagra’s broader goal: to make analytics more flexible, more accessible, and more actionable across the organization.

With Sigma, they’ve laid the foundation—not just for better dashboards, but for a more agile and innovative enterprise.

Read more about building data apps in Sigma.