Finding the Whole Buying Committee, Not Just One Person

I thought a contact list was done when every company had one good name next to it.

One name per company feels like progress. You can send to it, track it, and count it.

Except a deal doesn't have one buyer.

The person who owns the number is one person. The person who owns pipeline from marketing is another. The person who owns the tools and the data is a third. If you only have one of them, you have one door, and you're hoping it's the right one.

Some background

I first built a version of this for a client. It found three finance people at each company they cared about.

It worked, and it taught me something I didn't expect. The expensive part wasn't finding people. It was finding people I already had.

Every re-run searched everyone again. Every company that was already covered cost the same as one that was empty.

The question and the theory

Based on this, we asked: what's the cheapest way to fill a buying committee when you already know some of the people?

And then we made a theory: sort what you have first. Search only for the roles that are still short.

What I built

Give it a company. It does this:

1. Sorts the people you already have into three roles: a sales leader, a marketing leader, and a RevOps person (the one who runs the CRM and the tools).
2. Counts how many of each you still need. I use 2 per role.
3. If nobody is short, it stops. That run costs nothing.
4. If some roles are short, it runs one Clay people search per missing role. Only those.
5. An AI step ranks the candidates for the open spots.
6. If a role is still short, one web research pass looks for the gap. Every person it finds comes with a source link and a confidence level. Medium-confidence people go on a review list. They never get added on their own.
7. You get a final list, with a cap on how many new people it will add (I use 5).

What I found

The part I like most is step 3. A company where you already have enough people costs nothing to re-run. So you can run the whole list every month and only pay for the holes.

The part that surprised me is that Clay's people data is thin at some big companies. At one with about 2,000 people, there was no record of a CRO at all. A search by title just gave me regional VPs. That's the reason the web step is there.

Here's what I haven't done. I tested the free path from start to finish. I tested the paid steps one at a time. I haven't run the whole thing live on a real company yet, because I ran out of Clay actions first. If you run it before I do, tell me what broke.

What it costs: every search, every AI step and every write is one Clay action. Check your balance before you run it on a real list.

Here's how to get it running.

Build it with your AI