Lead review that ranks itself
A points-based scoring system that ranks every incoming record on fifteen criteria, so an acquisitions team stopped reading spreadsheets by hand and started calling the right people first.
- Client type
- Acquisitions firm
- Problem
- An acquisitions firm reviewed every new lead by hand. The list grew faster than anyone could read it, good prospects sat for days, and the order people got called in depended on who was looking at the spreadsheet.
- What I built
- A scoring system in the database itself: each record is scored on fifteen weighted criteria, ranked automatically, and surfaced with the reasons for its score. Alongside it, data-compliance alignment and trend visualizations for management.
- Result
- The team works a ranked list instead of a raw export. Prioritization is consistent, the best leads get called first, and management can see trends without asking for a report.
- Stack
- SQL · Stored procedures · Views & indexing · Data visualization
Built during a database engineering internship with a consulting firm in Ponte Vedra Beach, Florida, for one of their acquisitions clients. The client isn’t named; the approach is exactly what I’d bring to any business drowning in a spreadsheet.
The situation
New leads arrived faster than the team could read them. Every record was judged by eye, which meant the judgement changed with the person and the hour, and the best opportunities were often buried under newer, worse ones.
What I did
I sat with the people doing the reviewing and wrote down what actually made a lead worth a call — fifteen criteria in the end, each with a weight. Then I put that logic where the data lives: stored procedures score every record as it arrives, a ranked view is what the team opens in the morning, and each score carries its reasons so nobody has to trust a number blindly. I also tightened the data model to line up with the firm’s compliance requirements and built trend visualizations so management could see the pipeline moving.
What changed
Lead review went from reading to calling. The ranking is consistent whoever is on shift, the top of the list is where the value is, and the spreadsheet that used to run the process is now just an export.