Sales Intelligence Hub

Inspiration

Sales teams are often rich in data but poor in actionable insights. Standard reports show past performance, but they don't guide a sales manager on where to focus their team's efforts right now. I was inspired by this disconnect—the gap between having data and having a clear, prioritized action plan. I wanted to create a tool that acts as an intelligent co-pilot, automatically surfacing the most critical opportunities and risks within a sales pipeline, transforming a reactive reporting process into a proactive, data-driven strategy.

What it does

The Predictive Sales Intelligence Hub is an interactive dashboard that provides a multi-layered view of sales performance. At the highest level, it presents key performance indicators like win rate and pipeline value. It then allows users to drill down into historical performance with a "Win/Loss DNA" analysis to understand what drives success. Critically, it also analyzes the active pipeline to identify bottlenecks where deals are stalling. Finally, it uses a custom, AI-driven "Opportunity Score" to rank every open deal, presenting a prioritized list of at-risk opportunities that require immediate attention.

How we built it

This project was built entirely within Salesforce's Tableau Next platform. The foundation is a robust semantic model that connects Account and Opportunity data objects. I created a series of multi-level calculated fields to power the analytics:

1. Row-Level Calculations: Fields like Is Won, Is Lost, and ICP Fit Score were created to evaluate each record individually.
2. Aggregate Calculations: The Win Rate metric was built as an aggregate calculation (SUM/SUM) to measure performance across different dimensions.
3. Level of Detail (LOD) Expressions: A key filter, Has Open Opps, uses an LOD expression to perform complex segmentation at the account level.

These calculations feed a series of visualizations, which were then assembled into a single dashboard. The key feature, interactivity, was enabled using Dashboard Actions, allowing any visualization to act as a filter for the others, creating a seamless analytical experience.

Challenges we ran into

The primary challenge was debugging the nuances of the data and calculations within a new platform. We encountered and resolved several issues, including:

• Data Mismatches: Initial visualizations were blank due to subtle inconsistencies between the text values in our formulas (e.g., 'Closed Won') and the actual source data.
• Aggregation Errors: I faced errors when mixing row-level and aggregate calculations in the same view, which required us to redesign certain formulas for stability. For example, our "At-Risk Opportunities" count was refactored to a SUM of a 1/0 field instead of a COUNTD on a text field.
• UI Evolution: Adapting to the modern interface of Tableau Next, particularly for features like disaggregating measures to enable AI clustering, required troubleshooting beyond standard documentation.

Accomplishments that we're proud of

I am happy to have learned Tableau's interface such that I was able to create a dashboard that is more than just a report—it's a connected analytical application. The successful implementation of cross-visualization filtering allows a user to fluidly move from a high-level strategic insight (e.g., "we are weak in the manufacturing industry") to a specific, tactical action list (e.g., "these three manufacturing deals are at high risk and need a follow-up today"). The fact that this entire workflow is powered by a handful of robust, multi-level calculations within a single semantic model is a significant accomplishment.

What we learned

Throughout this project, I learned three key lessons. First, the power of a well-structured semantic model cannot be overstated; investing time in creating clean, reliable calculated fields is the foundation of any good analysis. Second, I learned that the most effective dashboards are not just informative but also intuitive, using interactivity to guide the user's discovery process. Finally, I gained a deep appreciation for the different levels of calculation in Tableau, understanding when to use a simple row-level formula versus a complex LOD expression to answer a specific business question.

What's next for Sales Intelligence Hub

The current dashboard is a powerful tool, but this is just the beginning. The next steps would be to further enhance its predictive capabilities:

• AI-Powered Clustering: Implement a scatter plot that uses AI to automatically segment opportunities into groups like "quick wins" and "large, complex deals."
• Automated Alerts: Configure data-driven alerts that proactively notify managers when a high-value opportunity becomes at-risk.
• Einstein Discovery Integration: Embed Einstein's prescriptive recommendations directly into the dashboard to provide "Next Best Action" suggestions for sales reps.
• Full Customer Lifecycle View: Integrate post-sales and service data to expand the analysis from winning a deal to managing customer health and identifying upsell opportunities.

Built With

• salesforce