Give feedback that helps an early idea
Early ideas are fragile, but soft feedback wastes everyone’s time. The trick is to be precise about what you saw before you say what you think.
Separate what you saw from what you think
Start with the observation. "The demo showed the invoice being created, but not being paid." Then the reaction. "So I am not sure the hard part is solved." A founder can act on the first sentence even if they disagree with the second. Feedback that skips straight to the verdict leaves them guessing what caused it.
Rate what was pitched, not what you would build
In The Pitch Room you score the problem, the solution, the evidence and the differentiation. Score the pitch in front of you. If you would have chosen a different customer or price, that is a comment, not a lower score. Keeping the scores honest is what makes the community results worth unlocking.
Remember that your scores are hidden until you submit them and stay independent of everyone else’s. Nobody is watching you rate, so there is no reason to hedge.
Be specific about the gap
"Needs more detail" helps nobody. "I wanted to know how long setup takes" is something a founder can answer in the next version. Point at the moment in the video where you had the thought, quote the line if you can, and say what you expected to hear instead.
When the case is not made
Say so. "Not yet" is a valid answer, and an honest one is a gift at this stage. Then give the founder something to work with: the one specific thing that would move your answer up the scale. It might be evidence that anyone has paid, a clearer explanation of the first step, or proof that the problem is as painful as claimed.
Be direct and be kind. Talk about the pitch and the idea, never the person. The founder chose to put this in front of strangers; make it worth their while.