Map the whole buying committee from public evidence, then write a different message to each person on it. Ten prompts, one terminal, any market.
Built by Prospectology®. Free to use, copy and adapt. No sign-up, nothing gated.
A B2B deal takes 5 to 7 people to sign off. Everyone knows that. Yet outbound still targets one persona and hopes they carry it internally.
That is not ignorance, it is cost. Mapping a committee properly used to mean hours of manual research per market, so people guessed a title that sounded senior and built a campaign on it.
The research was always the bottleneck, never the concept. That bottleneck is what has gone.
No job title, frustration or number enters your map unless you can point at the page it came from. If it cannot be sourced, it is a guess, and you never build a sequence on a guess.
That single rule is what separates this from asking a chatbot which titles matter.
This runs upstream of Clay, it does not replace it.
Clay is excellent at finding and enriching people once you know who you are looking for. The weak link was always the input: which titles go in the filter? Most people guess that input, then enrich the guess flawlessly. Beautifully enriched guesses are still guesses.
The order is: MPH decides who → Clay finds and enriches them → your sequencer sends. Every credit then lands on a title that has actually been seen buying.
You need Claude Code running in a terminal. That is the only requirement. No Clay, no credits, no other subscriptions to run the research.
Fill these three in, then swap them into the prompts wherever you see square brackets.
[WHAT] what you sell, in a few words
e.g. open banking payments
[WHO] the market you are selling into
e.g. UK ecommerce retailers doing £50m to £200m
[COMPETITORS] three competitors who publish case studies
e.g. competitor-one.com, competitor-two.com, competitor-three.com