We lost our first RFP by 4 points out of 100 on the scorecard. We lost the second one without ever finding out the score. By the third loss I did what a lot of founders do after getting beat twice: I hired a proposal consultant to make our next response sharper. We still lost. That's when I stopped looking at the document and started looking at the six weeks before it ever landed in our inbox.
The mistake I made three times before I saw it
Every time an RFP showed up, I treated it as the start of the sale. Buyer sends questions, we write great answers, best answers win. That's how I'd have designed a fair process too, so I assumed procurement worked the same way. It doesn't, not for the vendor who shows up cold. By the time an RFP is formatted into a document and emailed to a distribution list, the buying team has usually already talked to the vendor they expect to win, sometimes for months, and the RFP exists to satisfy procurement's requirement that they "considered options," not because the outcome is undecided.
The number that reframed it for me
I went looking for data instead of trusting my own bruised ego, and the split was stark: vendors who already had a relationship with the buyer before the RFP existed were winning at roughly 60-80%, while vendors responding cold, with no prior contact, were winning closer to 10-20%. Same document, same questions, same scoring rubric, wildly different odds. The variable wasn't proposal quality. It was whether we'd talked to a human on the buying team before the document existed. We hadn't, three times in a row, and no amount of consultant polish on the writing was ever going to close a 4-to-1 gap that got set before we knew the deal existed.
What changed on the fourth one
A prospect's ops lead had emailed us a casual product question two months before their formal RFP went out. Old habits would have had someone answer the email and move on. Instead I got on a call with her, not to sell, just to understand what was actually broken on her end. Forty minutes in, she mentioned a reporting requirement that wasn't in any public materials yet because their team hadn't finished drafting the RFP. I asked if I could send a one-page note on how we'd typically handle that requirement, framed as "things worth asking vendors about" rather than a pitch. She used almost that exact language in the RFP six weeks later. We weren't guessing what they wanted anymore. We'd helped write the test.
The meeting that mattered more than the document
We still spent two weeks on that RFP response, same as the losing ones. But the forty-minute call before the RFP existed did more to win the deal than any paragraph in the final submission. That's the part that's hard to accept as a founder who likes to believe good writing wins deals: the writing was never the lever. Access was. The RFPs I'd lost weren't lost because our answers were worse. They were lost in a conversation I wasn't in the room for, weeks before I knew a decision was being made.
How we qualify before we write a single answer now
Before committing real hours to any RFP now, someone on our team has to answer one question honestly: did a human from our company talk to someone on the buying committee before this document existed, in a way that wasn't just an inbound demo request? If the answer is yes, we go all in, because the odds are genuinely good. If the answer is no, we still respond if it's cheap to do so, but we stop pretending we're playing to win. We're playing to stay on the list for next time, and we budget the hours accordingly instead of burning a week's focus chasing a coin flip that was closer to a 1-in-6 shot from the start.
What I'd tell a founder about to answer their first cold RFP
If an RFP lands in your inbox from a company you've never spoken to, don't ask your team to write a better proposal. Ask whether anyone can get a real conversation with someone on the buying committee before you submit anything. If you can get that meeting, even a short one, take it before you write a single answer. If you can't get it, respond, but respond lean, and put your real hours into the next prospect where you can still get in the room early. The document was never the deal. The six weeks before it existed were.