🧠 Nonprofit Gradient — Smart Grant Intelligence Dashboard

Built by Smart Grant Solutions
A pragmatic system for surfacing high-priority nonprofit leads and visualizing the philanthropic landscape.

---

🚀 Project Overview

This project combines public IRS Form 990 data with machine learning-driven prioritization to help teams:

• Understand the nonprofit sector at scale.
• Identify ideal prospects by revenue, sector, and mission alignment.
• Visualize momentum, financial capacity, and organizational focus.
• Export filtered, high-signal lead lists for outreach.

What started as a data-cleaning experiment evolved into a purpose-built sales intelligence dashboard. It helps our internal teams (and now, others) find, rank, and connect with high-opportunity nonprofit organizations.

---

🎯 Goals

• Collect + normalize IRS 990 data for thousands of nonprofits.
• Score organizations across multiple axes (e.g., revenue, volatility, program %).
• Expose insights through clean, interpretable visualizations.
• Filter and export high-priority orgs by custom criteria.
• Enable anyone to run the dashboard locally with minimal technical skill.

---

🛠 Tools & Stack

• Python 3.10+
• Pandas for data manipulation
• Plotly for interactive charts
• scikit-learn for clustering + scaling
• Streamlit for dashboard front-end
• Matplotlib + Seaborn for exploratory visuals (now deprecated in favor of Plotly)
• Data source: IRS 990 filings + internal enrichment pipelines

---

📈 Dashboard Highlights

• ✅ Interactive filters: Sector, ZIP, revenue, momentum, etc.
• ✅ Export filtered orgs to CSV
• ✅ Pre-clustered organizations using KMeans
• ✅ Top 10 leaderboard based on priority + financial strength
• ✅ Sector/Flag breakdown visualizations (minimal but insightful)

---

🔍 Understanding the Dashboard

Each element in the dashboard has been built to surface actionable insight while remaining interpretable. Below is a breakdown of how to read and use each one:

📋 Explore the Filtered Organizations

A dynamic data table of nonprofits based on your filter selections. Sorted by PRIORITY_SCORE.

Key Columns:

• ORG_NAME: Name of the nonprofit.
• REVENUE: Reported total revenue (capped view at $200M by default).
• SECTOR: Organizational focus area.
• PRIORITY_SCORE: Our internal ranking metric based on multiple financial & strategic factors.
• MOMENTUM_CLASS: Velocity of growth based on multi-year behavior.

Use this to scan for high-scoring orgs in your sector or ZIP. Use the “Download CSV” button to export a list for outreach.

---

📊 Target Flag Distribution by Sector

Shows how many organizations within each sector are classified under different TARGET_FLAG categories (e.g., “High Potential”, “Monitor”, etc.).
This helps visualize which sectors are underexplored or saturated.

---

🔥 Momentum Watchlist by Sector & Class

Grouped bar chart showing average momentum scores across sectors and growth classes.

• X-axis: Sector
• Y-axis: Momentum Score (clipped between -0.5 and 1.5)
• Color: Momentum class

Most nonprofits have stable or negative momentum; positive surges are rare. This chart helps surface potential growth orgs hidden in underperforming sectors.

---

🚨 Excluded Orgs Over $200M Revenue

A secondary table listing orgs excluded from the main view due to extreme revenue skew. These are still valuable to inspect but were removed to avoid biasing visualizations and metrics.

---

🎛 Filter Logic & Guidance

Each filter works on its own or in combination with others:

• Sector: Choose one or more mission areas.
• ZIP: Enter prefix to filter by geography.
• Revenue: View ranges from $0 to $200M.
• Priority Score: Higher is better.
• Target Flag: Our internal classification schema.
• Momentum Class: Indicates trajectory or trend.

---

🧰 How to Run This Yourself (Even If You’re Not a Developer)

Requirements: Python 3.10+ installed

1. Clone this repo

git clone https://github.com/tsear/nonprofit-gradient.git
cd nonprofit-gradient

2. Set up a virtual environment

python3 -m venv env
source env/bin/activate

3. Install dependencies

pip install -r requirements.txt

4. Run the dashboard

streamlit run app/ui_dashboard.py

This will launch the dashboard in your browser where you can explore, filter, and export nonprofit leads.

---

👀 Results & Insights

We found the most actionable nonprofit leads:

• Operate with high program-to-revenue ratios
• Belong to emerging or underfunded sectors
• Are financially mature (>$10M revenue) but unflagged
• Cluster in key ZIPs indicating outreach potential

Over 25,000 orgs were processed and scored.
All logic is reproducible and transparent.

---

👥 Authors & Contributors

Tyler Sear – Lead engineer, analyst, and author
Dayna Sear – CEO
Smart Grant Solutions – Infrastructure + internal use case

Contact:
📧 tylersear910@gmail.com
📧 dsear@smartgrantsolutions.com

---

🧼 Disclaimer

This tool uses public IRS Form 990 filings and internal heuristics.
It is designed for research and prioritization—not definitive qualification.
Always manually validate organizations before engagement.