Challenge
For restaurant partners, reporting metrics are essential not only to track performance but to interpret trends and identify opportunities. They rely on this information to make decisions that directly affect their business.
But our existing dashboard wasn’t meeting those needs. Key metrics were missing, the data lacked meaningful context, and most users spent less than a minute on the page.
These gaps in our reporting were costing Grubhub millions, contributing to the loss of a $2M account, putting more than $300M in existing contracts at risk, and blocking $150M in new opportunities.
Existing dashboard
Objective
To establish a shared vision for a new Insights experience that helps restaurant partners make informed decisions and grow their business on our platform.
Role
I was the lead product designer on this project, partnering closely with a product manager and collaborating with a junior designer.
I was responsible for shaping the end-to-end vision, defining the reporting framework and component patterns, creating prototypes, and guiding cross-functional alignment across product, engineering, research, and operations.
Approach
I began by understanding the questions restaurant partners were trying to answer and where our existing tools fell short. From there, I mapped the key scenarios and metrics that mattered most and explored ways to present the data with clearer context and guidance.
To ensure the experience could scale, I created flexible frameworks, components, and interaction patterns that could support different types of data, and partnered with our systems design team to develop our first data visualization guidelines.
I brought these explorations together into a high-fidelity prototype, which we validated with restaurant partners and used to build alignment across teams on a shared vision for future reporting.
Result
The structure, component framework, and data visualization guidelines I developed during this project became the foundation for the Insights experience that exists today and continue to guide reporting work across the platform.
Testing with restaurant partners strongly validated the direction, with users finding the new experience noticeably clearer, more intuitive, and more actionable than the existing dashboard.
STEP 1
Before diving into the problem space, we made sure the right people were involved. Because Insights touched so many parts of the business, we pulled together a cross-functional group from product, engineering, sales, and restaurant operations.
We set up a weekly meeting to share updates, ask questions, and get feedback as the work evolved. Keeping everyone in the loop from the start helped build a shared understanding of the project outcomes and made it easier for the group to align on the final direction.
STEP 2
With the team in place, I shifted my focus to understanding what information our partners actually needed and why they needed it. I reviewed past research and previous initiatives around reporting and insights, and also spoke with internal subject matter experts to uncover themes such as:
Shared metrics, different goals
Most restaurants want the same info, but the insights and recommendations need to adapt to the restaurant based on context
Customer service matters most
Restaurants care deeply about customer satisfaction and want visibility into reviews and rates of return, not just sales.
Guidance should be balanced
Users want help interpreting trends, but they also need the autonomy to make decisions that fit their specific operations.
Proactive communication is essential
Partners don’t have time to dig for answers. They want timely alerts and insights embedded throughout the product experience.
These findings clarified not only what data was needed, but how it should be communicated.
STEP 3
With a clearer understanding of our restaurant partners’ needs, I began to identify the scenarios where Insights could meaningfully support restaurant decision-making. I mapped the fundamental questions partners were trying to answer about their business, such as:
How are my sales trending compared to last week?
Why are refunds increasing?
How well am I retaining customers?
For each scenario, I identified the metrics required to answer the underlying questions and determined how those metrics should relate to each other to form a clearer narrative.
Scenario A
Scenario B
STEP 4
With the scenarios defined, the next step was figuring out how to group and present the data in a way that felt clear to restaurant partners. I explored different page layouts, narrative patterns, and chart arrangements to understand how best to organize information.
How prominent should opportunities and recommendations be?
How should users be able to drill down into the details?
Explorations
At the same time, I partnered with our systems design team to create guidelines for displaying data visualizations. We standardized chart types, color use, labeling, and interaction patterns to ensure Insights could scale cleanly across modules and support future reporting work.
Data Viz Guidelines
STEP 5
Once the framework was defined, I stitched it all together into a medium-fidelity, clickable prototype that demonstrated how the experience could surface the right metrics, add the necessary context, and guide our partners toward meaningful actions.
Starting with a point of view
Right away, the user sees a summary of their performance that tells a story at a glance. Within each category, they're given a 'Top opportunities' carousel that highlights the most important areas to focus on, each with a short explanation and a link to a deeper page.
Telling the story and providing actionable suggestions
If the user digs into one of the opportunity areas, they're taken to a more detailed view. The experience automatically filters the data to match the opportunity being surfaced, then provides tailored recommendations and calls to action, supported by context that explains why those steps can make a difference.
Celebrating successes tied to behaviors
Once Grubhub has enough data to determine how any recommended actions have impacted metrics, the user will be alerted and guided back into the Insights experience.
Positive reinforcement next to the recommended actions helps clarify what’s driving the changes, and users can see a chart comparing data from before and after the actions were taken. They can also filter the data to drill down and see exactly how and where the impact was felt most.
Proactively alerting partners of issues
If Grubhub recognizes a new pattern or area of opportunity, the user is notified and directed to log in to the restaurant tool. They're taken directly to the detailed Insights page containing additional context and customized recommendations.
STEP 6
When the prototype was ready, I worked with one of our UX research partners to create a research test plan and discussion guide. We then invited a handful of restaurant partners who were familiar with our current reporting dashboard to review the concepts and provide feedback.
Overall, our restaurant partners found the concepts far more actionable and easier to understand than the current experience. Most of the feedback centered on labeling and clarity, meaning the overall structure and approach were strong and resonated clearly with users.
What worked:
There was a lot of enthusiasm for the data-driven recommendations, with partners saying they would be eager to follow the suggestions
Partners appreciated the new vs. returning customer breakdowns
The added detail around adjusted orders was seen as a major improvement, giving partners a more in-depth understanding of operational issues
What needed refinement:
Some terms, like “downtime” and “order channel”, weren't universally understood
The opportunities and recommended actions at the top of the detail pages needed to be more enticing, as some partners skipped those areas initially
Comparative statements needed even more clarity around who the restaurant was being compared to (e.g., similar cuisine types, same area)
I made updates to terminology, layout, and contextual explanations to address areas of confusion and ensure partners could easily understand and act on the insights provided.
Improving the visibility of the recommendations
Adding clarity to comparisons
Updated designs
Impact and key learnings
Although this was an exploratory project, the outcomes extended well beyond the prototype and influenced the product in tangible ways by:
Providing the foundation for the Insights experience that exists today
Reducing future design and engineering time through documented data visualization standards
Influencing the 2022 reporting roadmap and securing support for Insights investment
It also reinforced several important lessons about designing for data-heavy experiences.
Data needs context to be meaningful
Our partners felt much more confident in taking action when there were clear explanations of what the numbers meant and why they mattered
Clear narratives improve comprehension
Organizing metrics into scenarios and performance stories helped partners understand the bigger picture rather than interpreting charts in isolation
Alignment matters as much as the solution
Regular cross-functional check-ins and collaborative scenario-building brought everyone into the process, and the prototype became the final representation of our shared ideas and direction
Focusing on scalability drives long-term impact
Creating reusable components, a consistent page structure, and baseline data visualization guidelines ensured the work could live beyond the prototype, and it ultimately became the foundation for the Insights experience that exists today
Other featured work
