Inventa is a wholesale marketplace based in Brazil, connecting independent retailers with brands to empower small businesses across the country. Their mission is to serve as a platform for business owners to expand their reach and sell more products. Inventa has been in operation for only two years but has already been providing services to thousands of retailers and brands.
Inventa needed to provide their 1000s of suppliers insights into how their businesses were performing. They wanted to show how Inventa was positively impacting their sales and give the suppliers guidance on how to be more successful on the platform.
Their first solution was to query their database, extract CSVs of data, use Python scripts to create individualized PDFs, and then manually put those PDF reports in Google Drive folders for each supplier.
This process was highly inefficient, prone to errors, and not scalable. It took an entire day each week to generate the reports so they were not able to provide daily insights. Therefore it was difficult for Suppliers to track trends and understand their customers' behavior.
Inventa realized they could use Hex and dbt to automate, scale, and improve the analytics experience by creating an external facing dashboard with unique supplier links. They were excited to leverage dbt’s sematic layer to better model their sales data and create rich metrics and then use Hex to create an interactive dashboard.
Step 1: Building a Proof of Concept
Initially, Inventa built a proof of concept in Hex that replicated the information available in their PDF reports. They leveraged the same data from their warehouse, pulled it into a Hex project, and then created the corresponding visualizations. However, they still needed a workflow that separated the information out by supplier and was easy for them to access.
Step 2: Using Hex to Generate Links
Inventa used input parameters in Hex to generate unique links so that suppliers could only see their own data. They created hash IDs and prefilled input fields with unique links they could share with suppliers, ensuring they would only see their own information. This allowed Inventa to give suppliers full time access to their data.
Step 3: Building Their dbt Semantic Layer
To make the metrics in the dashboard more maintainable, Inventa built them in the dbt semantic layer. This allows Inventa to define what the metrics are, how they can be aggregated, what dimensions they can be sliced by, and what time grain they can be evaluated in. This makes it easier to evolve metrics over time and simplifies how to query and visualize the data in Hex. It also ensured that the metrics they were showing suppliers was the same as the ones they looked at internally.
Step 4: Creating Two Main Dashboard Sections
Now that metrics were configured, Inventa decided to provide additional insights for their report.
They created two sections on the dashboard: the commercial section and the Reputation Program Score section. The commercial section provides insights into business performance, sales, and customer behavior. The Reputation Program Score section helps suppliers understand how to improve their products on the platform and sell more.
Check it out here!
Step 5: Adding Feedback Form
Lastly, Inventa added a form on the dashboard so suppliers could provide feedback about what analytics are provided. This proved to be a great way to find accounts who were struggling and Inventa could reach out and get connected with Suppliers to help them.
Hex and dbt has not only allowed Inventa to automate and scale their analytics reporting process, but it has also provided a better experience for their suppliers.
Suppliers are now able to see daily insights instead of weekly, and the interactive charts make it much easier for them to understand their performance on the Inventa platform. Additionally, suppliers appreciate the new UI, the ability to access the data through a link, and being able to provide feedback.
Overall, the new analytics system has been a success for Inventa. They were able to build something completely new and more efficient in just a few weeks, without any dependency on engineering, design, or product. The system has improved the experience for their suppliers, while also helping Inventa achieve their mission of empowering small businesses across Brazil and Latin America.
Daniel McAuley, Data Lead and Gabriel Marinho, Analytics Engineering Lead