Sales data work, from a few days a month to a few hours
At one of the world's largest wine and spirits sellers, the sales team did the data work by hand, from collection through analysis. We automated collection, designed a simpler foundation, built the BI dashboards, and stayed through production.
- One of the world's largest wine and spirits sellers
- Data Intelligence
- Days to hourssales-team data work, each month
- 15+POS systems, collected automatically
- 9Power BI dashboards
The challenge
The sales team was stuck in manual data work.
One of the world's largest wine and spirits sellers was struggling with data management, from acquisition through to business analysis.
Most of that work was done manually by the sales team. Productivity suffered, and transformation errors and inconsistent data were common.
The data was not granular enough to show what the business was actually doing.
- Data from 15+ POS systems was collected by hand.
- Excel macros introduced errors and inconsistencies.
- Only aggregated data was available, so product, day and outlet were invisible.
What we did
How the work ran
From a design that removes the manual work, through production and handover.
- 1
Audit the data management.
We audited acquisition from 15+ POS systems and the Excel-macro transformations, then recommended how to automate acquisition from each POS and how to transform the data in code.
- 2
Design a business-driven schema.
A simpler database architecture kept granularity at outlet and day.
- 3
Prove it, then put it in production.
Four proofs of concept used RPA tools, Selenium and UiPath, and integrated with Azure Data Lake, Databricks and Snowflake. The solutions then went to production across all 15 retailers.
- 4
Show the business, and hand it over.
Nine Power BI dashboards sit on Snowflake. Local and regional teams took the knowledge, so the client can run the new architecture.
Before and after
What changed
| Topic | Before | After |
|---|---|---|
| Sales-team effort | A few days a month | A few hours a month |
| Granularity | Aggregated data only | Sales at product, day and location |
| Quality | Transformation errors and inconsistencies | Fewer errors, and less maintenance |
| When a POS changes | The work stops | A setup that holds up when a POS changes |
The results
Time that went into collecting data can go into selling.
- Sales-team data work fell from a few days a month to a few hours.
- Sales data that used to be aggregated is now available at product, day and location.
- Errors dropped, maintenance is lighter, and the data is faster to reach.
- The setup holds up better when a POS changes.
Services on this project
The same work, for your team.
Technology used
- RPA
- UiPath
- Selenium
- Azure Data Lake
- Databricks
- Snowflake
- Power BI
Service: Data Intelligence
Facing a similar problem?
Tell us where you are today.
We reply within 2 business days.



