Data-driven hospitality for a national fine dining chain

For a national fine dining chain, we designed a guest‑insights application. The goal was to help restaurant teams understand their guests better and adapt service accordingly.

Success meant three things:

  1. Restaurant teams had to adopt the tool

  2. Guest experience had to improve

  3. Those improvements had to measurably impact the business

In the program’s first year, it was on track to deliver a 1.5% EBITDA lift, driven by increased guest spending and visit frequency.

Can we provide service experts—managers and wait staff—with actionable insights about their guests, and turn more guests into “regulars”?

CONTEXT

For this client, we were already building a data platform, Guest DNA, which would enable new personalization, marketing, and guest insights.

Our idea was to build an application on top of that platform that restaurant teams could actually use during service. The product would surface insights about guests and help managers and staff tailor the dining experience in ways that encouraged repeat visits.

WHAT WE MEASURED

  • Check size

  • Visit frequency

We delivered a tool that complemented general managers’ existing processes for service prep, team communications, and reservation management. In year one, the program was on track to deliver a 1.5% EBITDA lift based on guests’ increased spending and visit frequency.

OUTCOME

Selling a new approach

For both our team and the client, delivering an enterprise product on top of a new data platform was largely new territory.

I advocated for an accelerated, collaborative approach that involved both corporate stakeholders and restaurant users to meet timeline and budget constraints. This required coordinating across restaurant operations, data science, product teams, and business leadership.

Designing for behavior change

The tool introduced a shift in how managers approached “gifting” for guests—complimentary appetizers, drinks, or other small gestures. Our analysis suggested that increasing this budget strategically could drive higher spending and more frequent visits.

In other words: spend money to make money.

To make that shift possible, the system had to do two things well:

  1. Show managers how they were tracking against their gifting budget

  2. Provide recommendations strong enough to build trust in the data

Without that trust, the behavior change wouldn’t happen.

Feedback loops & change management

The feedback loop for this system was much longer than in typical digital products. Even frequent guests at a restaurant like this might only visit once every month or two. Measuring success wasn’t like tracking daily active users.

The cycle looked like this:

  1. The platform generated insights from guest data

  2. Managers and staff adjusted their service plans

  3. Guests experienced that service

  4. Guest behavior changed—or didn’t

Because of this, we paired quantitative metrics with direct feedback during rollout. We worked with individual restaurants to understand whether the insights felt useful, how comfortable managers were changing their service approach, how those changes landed with guests.

Those conversations helped us refine both the recommendation algorithms and how insights were presented, based on how restaurant teams actually ran their operations throughout the day.

HOW’D WE DO IT

Previous
Previous

AI authoring tool

Next
Next

Deep sea control system