Naming the category: proposal intelligence

We call PursuitAgent a proposal intelligence platform. The phrase is not standard. The standard phrases are “RFP response software,” “response management platform,” “proposal automation,” and a handful of variants. I want to write down why we did not pick one of those and what the phrase we chose is supposed to earn.

Three reasons we did not use the standard labels.

“RFP software” is too narrow. Half of the documents teams respond to are not RFPs in the strict sense. They are DDQs, security questionnaires, RFIs, procurement portals, and enterprise vendor reviews. A “RFP software” label tells the buyer we do RFPs and is silent on the rest. The actual product handles all of them. The label needs to handle all of them too.

“Response management” is too administrative. Response management is project management — assignments, deadlines, status. We do that, but it is the floor, not the ceiling. The work that matters is the substance of the response: the win themes, the citations, the verification, the post-mortem feedback loop into the KB. Response management as a category leaves the substantive work invisible.

“Proposal automation” is wrong. 1up’s post made the point cleanly: most questionnaires are similar but different enough that you can’t copy-paste, and re-writing takes as long as starting from scratch. The work isn’t automation. It’s retrieval, drafting, and verification — three distinct steps that require human judgment at each handoff. “Automation” implies a button you press; the actual product is closer to an editor with citations.

What we mean by proposal intelligence: a system that learns from every bid that runs through it and that compounds across bids. Win themes that worked promote. Win themes that didn’t retire. KB content blocks update from post-mortem feedback. The next bid starts with the accumulated intelligence of every prior bid. We covered the mechanic in the eight-stage RFP pipeline and in the discussion of post-mortems and KB updates that closes that post.

What the phrase has to earn: the compounding has to be visible to the user. A buyer who tries the product and sees their second bid run no faster than their first is right to call the phrase marketing. The phrase is defensible only if the post-mortem stage writes its output back into the corpus the drafting stage draws from. We have built that loop. We will keep writing about how it actually performs in production.

What we are explicitly not claiming with the phrase: that the system replaces human proposal work, that it generates proposals end-to-end without review, or that “intelligence” implies any kind of autonomous decision-making. The system drafts, retrieves, verifies, and tracks. Humans approve, edit, and ship. The intelligence is in the corpus and the loop, not in any one drafted sentence.

A label is a promise to the reader that the rest of the work has to keep. We picked one we think we can keep.