Yesterday I had a friend reach out who told me they’ve been using gpt3 for lead validation. I thought it was super interesting and solved a massive problem, so I wanted to share it with you.
Every b2b software company should have an owned database of all their target customers sourced from anywhere they can collect them. Think directories, scraping google maps etc.
(Side note, if you’re an early-stage startup and have a list of all your target customers when you’re raising you A, investors will be stoked.)
But say you collect 50K leads, which do you prioritize your sales team working?
Traditionally, SDRs would have to comb this database and manually validate leads.
Most companies have keywords or language that signal they’re a good fit for your product. But keywords are super limited because people talk about the same thing differently pretty often.
Until now.
The process is pretty straight forward:
- Use python to request a potential leads URL
- Ingest all the text on the persons site
- Then use a gpt3 prompt like “Does this company have multiple mental health professionals or are they a solo practitioner?”
- For loop this through all your leads
And you just validated your list.
Final note, I don’t think AI tools like this will replace people. I think it will make 10x people 100x people. We’re tryihng to make our teams super human, so that one really good person can work like they they’re 100 people.
