A ₹1.2 Cr pilot with a top-3 Indian bank. Nothing invoiced. Then the bank asked the founder for a commercial proposal. He sent it that night, and believed the commercial discussion had finally started.
spent on the pilot
invoiced
The commercial discussion had actually started nine months too late.
A proposal request is not always a price discussion
A proposal request with no agreed buying process behind it can become the document procurement uses to compare you with everyone else. Your scope, commercials, timelines and architecture become inputs for a buying process you have not been invited into.
Then other vendors enter. The team that spent nine months proving the solution discovers that technical preference and commercial preference are not the same thing.
What nine months of pilot looked like
Six people on the pilot
Daily stand-ups, CXO time, dedicated engineering, custom integrations.
Nine months later
Nobody had agreed what a "yes" actually meant.
The pilot was not the risk. The unagreed definition of success was
Fix it before the proof of concept starts
One page, four answers, agreed with the customer in writing before any work begins:
Four Answers, Agreed in Writing
What number tells us the POC worked?
One measurable result, with the threshold that counts as success.
Who owns the commercial decision if it works?
A named person on the customer side, not a committee.
Which budget will fund it?
The line item, and whether it is approved or still to be approved.
What happens at the end of the agreed period?
Commercial discussion, procurement process, another approval, or stop.
Then ask one question before work begins: "What would have to be true for this pilot to become an order?"
Reading the Answer
"We want to explore"
Interest, with little commitment from the customer.
"We will see how it performs"
Evaluation without a decision attached.
"Let us first run the pilot"
The start of a pilot, not the start of a buying decision.
I do not automatically walk away from those answers. But I know I am validating interest, not a buying decision. That distinction matters, because exploration costs you engineering time, sales time, founder time, customisation and opportunity cost, while the customer commits to very little.
Ask what happens after the proposal goes out
Before sending any commercial proposal, ask: who evaluates it, who signs, is the budget approved, are other vendors being invited, and what has to happen for us to get the purchase order.
A proof of concept should not only prove that the technology works. It should also clarify what happens commercially if it does. Otherwise you can win the pilot and still lose the business
Related: Order booking is a sales number. Collections is a commercial number — what happens when the commercial terms are not agreed up front.