Case study
A lead engine that says less than it knows.
Finding prospects actually worth contacting is slow, manual work. This engine starts from two free public datasets: the US Department of Education's IPEDS, and ProPublica's Form 990 API. No paid data vendor. No scraping. No account needed for either.
The funnel.
The output.
The 11 organisations that made the final cut. Labels and revenue bands are anonymised because this is real client research.
| Organisation | Revenue band | Donor-distress score | Contact confidence | Score |
|---|---|---|---|---|
| Organisation A | $4.7M | 50 | high | 58 |
| Organisation B | $2.9M | 50 | none | 56 |
| Organisation C | $4.2M | 35 | low | 42 |
| Organisation D | $3.9M | 35 | none | 38 |
| Organisation E | $3.6M | 25 | none | 28 |
| Organisation F | $2.2M | 22 | high | 25 |
| Organisation G | $3.4M | 20 | none | 23 |
| Organisation H | $4.1M | 12 | none | 22 |
| Organisation I | $3.9M | 12 | none | 18 |
| Organisation J | $3.7M | 10 | none | 16 |
| Organisation K | $2.7M | 10 | none | 16 |
Data is anonymised because it is real client research. No organisation names, cities, EINs, websites or people's names appear on this page.
Differentiator
It refuses to assert anything it cannot evidence.
The engine grades the confidence of every contact name it finds. If a name cannot be grounded, it returns "unclear" instead of inventing a lead.
Of the 11 organisations above, only 3 had a findable contact name. That number is unflattering and it is included on purpose. A tool that never says "unclear" has simply stopped checking.
Start here
Have something you keep doing by hand? Let us automate it.
Tell me what the task is and how often it eats your week. I will give you a straight answer on whether it is worth automating, and roughly what it would take, before you commit to anything.