A little check up on this. I've mailed probably close to 400 customers and heard back from 50. Assuming the number of customers I've mailed is accurate (I don't record it as I find it to be a redundant exercise) to about +/- 100 I'm hearing back from about 1/6 to 1/10 people. This sounds pretty low but I'm sure compared to the realworld phone solicitor, mail in surveys it's a fair bit higher. I haven't done any quantitative analysis yet but some observations and things I've mulled over are:
*A number of the people I'm hearing from are commenting this is the first time they've been contacted and that they're glad to be involved. It makes them feel special, this probably can't hurt.
*Not a ton of people have book marked my shop but more than I expected have. I think my mail is, at the least, encouraging people to look at my shop again and some of the respondees have commented No, I haven't bookmarked it, but I have now
*A lot of people say that they would visit my shop again, but not all of them are bookmarking it
*The vast majority of the respondees, and likely my customers, are female
*The price feedback has been, generally, that it's "Just right". However, surveying people who have bought from my shop already may not be the most effective way to judge this. That said, even people who say that the price was too high say they would return
*There's a large smattering of who buys from malls for their quests at least. Everything from newbies to 8 1/2 year accounts seem to, with a fairly even distribution between <3 months, 3-12 Months and >12 months. These arbitrarily defined categories are based on my judgement of how long people would stick with the game for and that active users would probably be distributed within a 10% variance between those categories.
I've also taken to mailing the respondents saying that I appreciate their feedback and that we have a new special (of late an expanded stock) going on.I hear back from 1/6 of these people, I don't know if its made a dent on my sales yet but it hasn't hurt as no one has responded in an upset manner.
I think I'll send off a last batch of mail then look into a qualitative analysis of these numbers. I wonder, like Zoro suggested, what the difference is in character between someone who responds and someone who doesn't...I guess that just a source of error for the survey.
I've recorded no usernames throughout this process and haven't correlated purchases with respondents, which, in hindsight, may have been helpful but on such a large scale probably not worthwhile.
I'd be curious to compare the entire mall's shop history and see how many users buy from multiple shops and if there is any interesting correlation between who buys what where...that's a lot of work and a huge group collaborative effort that is beginning to feel like a nightmare just suggesting it.