Willingness to pay (WTP) is the highest price that a consumer will buy a good or service. WTP is so important because it's the basis of value-based pricing (especially for SaaS, where cost-based pricing doesn't make much sense). But WTP is extremely hard to measure – and nobody can seem to agree on how to do it. Here's a few different methods to calculate WTP.

Source

Surveys: Van Westendorp's Price Sensitivity Meter (PSM) asks consumers 4 questions:

You can use this data to calculate an acceptable price range for your product. This method is probably the standard when doing direct surveys.

One of the best books I've found on WTP (and other pricing topics) is Monetizing Innovation. A great tip from the chapter on WTP is to avoid the "average trap". Look at the distribution of responses – you might have bimodal (or other kinds of non-normal) distributions. You could imagine this would look something like consumer, SMB, and enterprise customer segments who value your product and features very differently.

Experiments and auctions: Google, Facebook, and other ad networks use auctions to determine the optimal price to set for ad inventory. The types and structure of auctions that these companies use could probably fill their own book, but Hal Varian's writings are the definitive source for how they work (Varian is an emeritus economics professor at Berkeley and longtime Chief Economist at Google).

Many startups avoid running pricing experiments, in fear of alienating or angering existing customers (why did I get a different price than so-and-so)? I think that startups have more room than they think to experiment with different pricing schemes. As much as a startup iterates on product-market-fit, they should iterate on pricing (pricing is a critical part of the product!).

Comparing the different WTP methods. The authors of the paper that categorized the different frameworks for calculating WTP also made a matrix of what is the best for each situation.