Transforming Workplace Food Programs Through Data

By Jess Legge + Cody Legge

Workplace catering has traditionally relied on headcounts, standard portions, and broad assumptions about what employees will eat.

Data changes that.

By tracking what is prepared, what is consumed, and how those patterns change over time, workplace food teams can make better decisions about menus, portions, production, and waste.

At Sifted, analytics have become part of how we run every food program.

“We’re in the third wave of hospitality at work. We’ve moved beyond endless options and bulk portioning to more refined, conscientious programming.”

Jess Legge, CEO and Co-founder of Sifted

Start With What People Actually Eat

The foundation of Sifted’s analytics program is straightforward: understand how much food should be prepared and how much is actually consumed.

Rather than looking only at total food production, Sifted tracks individual dishes, or “elements,” within each menu.

That allows us to see how much of each protein, starch, vegetable, sauce, or other menu element was prepared and how much was consumed.

Cody Legge, Sifted’s Director of Finance and Analytics, explains:

“Our system doesn’t just track how much food we prepare in total; it tracks how much of each element, by type, is made and then consumed. This allows us to refine every aspect of our service, from production quantities to portion sizes.”

That level of detail gives culinary and operations teams information they can use to make specific changes rather than relying on assumptions.

Four Metrics That Improve Workplace Food Programs

1. Consumption by Element

Knowing which foods employees consistently eat, and which they leave behind, is one of the most useful signals in a workplace food program.

Instead of measuring only overall food consumption, Sifted tracks individual components of a meal, down to different sauce options.

Over time, patterns emerge.

An office may consistently consume more protein than expected. Another may prefer vegetables over starches. Certain dishes may perform well in one office and poorly in another.

That information can be used to adjust both menus and production quantities.

“Consumption by element allows us to cater to the preferences of each team or office. It’s about more than offering variety. It’s about creating menus that match how the people we serve actually eat.”

Jess Legge

2. Runout Risk

Running out of food is a problem. Consistently preparing too much food to prevent it is also a problem.

Historical data gives us a better way to manage the tradeoff.

Rather than relying on a static office headcount, Sifted looks at previous attendance, consumption patterns, and variability by office and day.

“We don’t rely on static headcounts. We calculate the risk of runouts by analyzing previous data, such as variability in attendance and the eating patterns of different teams.”

Cody Legge

That allows production teams to build an appropriate buffer based on the individual program instead of simply overproducing.

The goal is enough food without unnecessary waste.

3. Customized Portioning

Different offices eat differently.

Traditional catering models often use standardized portions regardless of the team being served. Sifted adjusts portions by menu element based on historical consumption.

For example, a sales-heavy office may have inconsistent attendance but larger portions when employees do eat together. A health-conscious office may consistently consume less starch.

Those patterns can be reflected in future production.

This allows clients to spend their food budget on what employees actually consume rather than paying for standardized quantities that regularly go uneaten.

“By tailoring portions to the specific needs of each office, we provide a better experience for employees while helping clients avoid unnecessary costs.”

Jess Legge

4. Menu Performance Over Time

One lunch tells you very little. Months of service tell you a lot.

Sifted tracks menu performance over time so culinary teams can identify dishes that consistently perform well, those that underperform, and how preferences change.

That history informs future menu development.

Feedback comes from multiple sources.

Host feedback: Sifted’s on-site teams capture more than 90 data points during service, including timing, portion sizes, equipment needs, and feedback from employees and program managers.

Eater behavior: Surveys can be useful, but they don’t tell the whole story. Consumption provides another signal. If employees consistently eat more of one menu element and leave another behind, that behavior becomes part of the data used to evaluate the program.

The result is a feedback loop: each service provides information that can improve the next one.

From Data Collection to Daily Operations

Analytics only work when the underlying data is reliable.

That means data collection has to be part of daily operations, from portioning in the kitchen to observations made during service.

Chef feedback, host reporting, consumption tracking, and employee feedback all contribute to the system.

As Cody explains:

“Data collection is mandatory at every level. Host surveys, chef feedback, and consumption tracking are all integral parts of our daily operations.”

The objective isn’t to collect data for its own sake. It is to give culinary teams and clients information they can use to make better decisions.

Technology Built Around Workplace Food

Sifted owns and develops the technology behind its analytics program.

The system has been built around 10 years of workplace food operations and connects information from production, service, consumption, and employee feedback.

Because Sifted controls the technology, the system can also evolve with the programs it measures.

“If we identify a new data point that we need to track, we don’t have to wait for an external vendor to make changes. We can build it into our system immediately. Clients can ask us to track new metrics, too. If you’re curious, let’s track it and see what the data say.”

Jess Legge

The reporting illustrated on page 5 of the original paper spans four areas: client-facing reporting, portions and waste, eater feedback, and service execution.

Better Data, Better Food Programs

The value of analytics is practical.

It can help answer questions workplace teams deal with every day:

Are we preparing the right amount of food?

What are employees actually eating?

Which dishes consistently perform well?

Where are we overproducing?

How likely are we to run out?

Is the program improving over time?

Instead of answering those questions based primarily on assumptions, historical data provides evidence.

That makes it possible to build workplace food programs that are more responsive to employees, more efficient to operate, and less wasteful.

About the Authors

Jess Legge is CEO and co-founder of Sifted. A serial entrepreneur, Sifted is the third company she has co-founded. Her experience spans high-growth businesses, brand transformation, and data-informed decision-making. She holds a degree in Business with a minor in Art & Design from Purdue University.

Cody Legge is Director of Finance and Analytics at Sifted, where he leads the company’s analytics work around portioning, forecasting, and waste. Before Sifted, he worked in business analytics at Skyscanner and built large-scale optimization models at Booz Allen Hamilton. He holds an MBA from Carnegie Mellon University and a bachelor’s degree in mechanical engineering from Virginia Tech.