I might be wrong, but this seems like you received bad advice in so many ways! I see problems at every step.
What a person claims they’re willing to pay and what they actually do when asked to enter their credit card info are not the same.
Let’s assume optimizing is about maximizing profit. Let’s simplify, as many startups do, by looking to maximize revenue and assuming that costs won’t dramatically change, though people point out that cheap customers have higher costs (support requests, complaints, etc).
Okay, so let’s maximize revenue. Revenue is price multiplied by number of customers.
Let’s replace the two survey questions with a more pointed question: “what is the maximum price you’d be willing to pay?”
Once you conduct your survey, the graph won’t have straight lines but a distribution that might look like one of these curves, where X is price and Y is number of answers. Given that at any price point all customers below that price will buy too, you’ll want to calculate the cumulative distribution curve, which looks like one of these S curves (but flipped around because it’s going to be a decreasing function). You’ll end up drawing rectangles similar to the way they are on this graph and picking the one with the maximum area (price times total customers).