CRM Pipeline Analysis Dashboard
Power BI Development | Data Extraction & Modeling | Sales Performance Analytics
01The challenge
A growing sales organization was sitting on thousands of leads but had no real visibility into their own pipeline. Leadership couldn't answer basic questions: Where are we actually losing deals? Which agents are driving results, and which aren't? Why is our conversion rate stuck? Decisions were being made from manual CRM exports that told a partial, outdated story.
02The data problem
The CRM's built in reporting only offered a flat, point in time snapshot with no historical stage tracking, no defined pipeline order, and no location intelligence. To build something leadership could actually rely on, the data needed to be pulled at the source, stored properly, and rebuilt from scratch.
03My approach
I connected directly to the CRM's REST API to extract leads, deal stages, agent activity, and status change history as live, structured datasets, rather than depending on static manual exports. The extracted data was staged in a SQL Server database, where I handled the heavy cleaning and transformation work before it ever touched the reporting layer, deduplication, null handling, standardizing inconsistent stage and status naming across agents, and writing queries to reshape the data into clean, analysis ready tables.
Set up a scheduled data pipeline connecting the SQL database to Power BI, so the dashboard refreshes automatically from live data instead of static exports
Rebuilt the dataset into a proper star schema, enabling accurate time based comparisons and eliminating data duplication
Engineered custom sequencing logic for pipeline stages, since the CRM had no built in concept of stage order, this was critical to making the funnel visual actually represent reality instead of a random alphabetical sort
Enriched the dataset with geocoded location data to enable an interactive map view of deal activity by country
Built dynamic DAX time intelligence measures for month over month performance tracking, with instant variance and percentage change calculations
04The result
The finished dashboard gave leadership a live, single source of truth for the entire sales pipeline. It revealed that only 11.6% of leads were converting to closed deals, with the sharpest drop off happening between the "Sales Accepted" and "Opportunity" stages, a bottleneck that had been completely invisible in standard CRM reporting. It also exposed major performance gaps between agents (93 closed deals vs. 18), giving sales management a clear, data backed starting point for coaching and pipeline strategy.
Tools & tech stack
- Data Extraction: REST API integration
- Data Storage & Cleaning: SQL Server, T-SQL (deduplication, transformation, staging tables)
- Data Modeling: Star schema design, Power Query (M)
- Reporting & Visualization: Power BI Desktop, DAX (time intelligence, dynamic measures)
- Automation: Scheduled data refresh pipeline
- Geospatial Enrichment: Geocoding integration for map visuals
Have data that isn't telling you anything yet?
Send me what you are working with — even if it is four spreadsheets and a CRM export that don't agree with each other.