Quote Originally Posted by Lissou View Post
The problem with your suggestion is that people say they will buy something, and then they don't. I've been there. You can't really rely on what people say they'll spend money on and invest based on that. Might be different with an established franchise like OOTS, I guess.
Surveys are very difficult, however, they can give you ranges. That is why I was noting ways to measure data quality.
Give more credibility to those who claim to have made past purchases. Perhaps discard the others.
An account that is tied to the survey (even a free one) means people are less likely to lie about their past purchases, but you have a higher percent of buyers. This gives you an upper bracket (a ceiling).
If you can ask one or two questions based on claimed purchasing habits (i.e.: something that appears in the books only), you have a measure of true past purchasers (inaccurate, but within 20%). Think of it like a pop quiz (insert maniacal laughter from team evil). You also get to measure the % of liars misleading respondents. If the sample is large enough, you may consider applying this % to the overall survey.

Even if the PDF+print option is not possible, a survey can give a measure of how profitable the options are. However, as I noted, a decent survey is a lot of work and a bad one will leave you with misleading answers. Tricky stuff. At times it is a lot of work and it ends up being no better than an educated guess.