Benne: Turning data into strategy
Industry
Hospitality
Timeline
Jul 2025 - Sept 2025
Organization
Benne - Heritage Bangalore Dosa
My Role
Business Intelligence Lead
Team
Chief Growth Officer, Head of Operations, 1 intern
Type
Live project
Description
Benne's team needed a clearer way to connect their data with operational strategy. Working closely with the Chief Growth Officer and Head of Operations, I developed a BI dashboard that translated fragmented data into actionable insights, helping inform strategy for menu, inventory, wastage management, labour planning and customer experience.
Note: All data displayed is dummy data and for demonstration purposes only.
The problem wasn't a lack of data
Benne had lots of data across sales, inventory, operations and finance, but there was no way to connect the numbers to decisions around what to sell, how much to prepare, when to staff, or where operational effort should be focused.
The opportunity was to move from collecting data → collating data → visualizing data → creating strategy.
Understanding how different stakeholders use information
I spoke with stakeholders across management, operations and kitchen teams to understand what information they needed, how they currently accessed it, and where decision-making was getting blocked. The following are the personas of the stakeholders:
What I found from my research
Product performance was difficult to compare
Patterns of demand fluctuation were buried under raw data
Actual wastage and recorded wastage had discrepancies
There was no system to record and analyse customer feedback
Lack of consolidated visibility to the executives
Framing the stakeholder needs as business questions
After framing the business problems as questions, I plotted a path to go from these questions/problems to informed strategy using data.
Structuring the information
The dashboard needed to serve different levels of decision-making from an executive overview to detailed operational analysis. I structured the information into interconnected areas rather than treating every metric as an independent chart.
Raw datapoints used in each section
Metrics by section
From raw data to usable information
The required information wasn't available in a single clean dataset. I structured and transformed data from multiple sources before it could be visualized.
From transaction data to staffing & preparation insights
Individual POS transactions weren't useful for planning operations on their own. I transformed the timestamps into day-hour groups and aggregated revenue to reveal recurring demand patterns across operating days and hours.
Average revenue per day per hour visualization is showed in the picture below.
This turned transaction-level sales data (which is a proxy for demand) into a planning signal for staffing and preparation.
In another example of transforming raw data, my colleague and I used the n8n automation platform to create a new workflow, fetching the restaurant's google reviews and transforming them into actionable points using Gemini AI.
Dummy review analysis
This turned Google reviews into a concise customer feedback module which provided pin-pointed actionable insights to the operations team.
Choosing the right visual form
I explored different visual forms before selecting the ones that best communicated each business question.
To compare product performance for example, I looked at multiple different visualizations.
I chose the scatter plot.
Why: It allowed two variables to be compared simultaneously while revealing clusters and outliers and since I was already comparing one monetary parameter on an axis, it didn't make sense to make it into a bubble chart with another similar parameter like revenue.
Finding patterns in the data
Question: Which products are driving performance?
Visual: Menu Engineering Matrix- Popularity vs. Gross Profit
Why this visual: Compares popularity and profitability in a single view.
Insight: This matrix showed what actions needed to be taken
regarding different products based on their demand and their contribution.
Strategic implication:
High sales + high profit = Ensure availability and promote
Low sales + high profit = Explore better positioning or promotion
High sales + low profit = Review pricing, portions or costs
Low sales + low profit = Consider discontinuing
Question: When does demand peak and require more staffing?
Visual: Demand Heatmap- Average Orders by Day & Hour
Why this visual: Reveals demand patterns across days and time-slots at a glance.
Insight: The heatmap shows demand clustered around specific day-hour combinations, revealing recurring peak periods.
Strategic implication: This gave the team a basis for planning labour allocation and preparation cycles around actual demand patterns rather than fixed schedules.
Question: How is customer sentiment changing across stores and over time?
Visual: Customer Sentiment- CSAT, NPS & Review Scores
Why this visual: Tracks changes in customer sentiment over time.
Insight: Customer metrics revealed whether satisfaction was improving or declining, while review analysis helped identify the reasons behind the trends.
Strategic implication: Team could use overall trends to guide service strategy, while store-level differences helped identify where targeted interventions were needed.
Note: All data displayed is dummy data & for demonstration purposes only.
Designing the dashboard
The final system brought these signals together so the team could move from monitoring performance to investigating problems and identifying actions.
Dashboard metrics and logic
Dashboard pages
What I would do differently
Looking back after completing the project, there are a few things that I would do differently if given a chance.
Validate insights before expanding the dashboard
I would validate a smaller set of insights with the team before expanding the dashboard. This would help prioritize the metrics that directly influence decisions and reduce the temptation to measure everything that can be measured.
Measure the impact of the system
Although the team found the metrics very useful, I would build a stronger feedback loop between insights and outcomes. Tracking changes after implementing new strategy would help quantify the system's actual business impact.
Impact that we had
Translated raw data into actionable insights that improved decision-making for CXOs and internal stakeholders.
Increased visibility into specific data points that traditional POS systems could not provide.
Helped CXOs get a better view of customer experience by analyzing online reviews.
Enabled the team to create better, informed strategies across different departments.