I have one client who spends quite a bit with AdWords (mid five figures per month), because it works for them - and their widgets (or I should say, widget services) run $5900-$7500. So yea, it can certainly be done with expensive widgets.
Most of what I would suggest has been mentioned here. Do you know what your bounce rate is? Average time on site from the people who come in via AdWords? Are you running Google Analytics, or any other kind of analytics? If not, I'd strongly suggest it. Find out exactly what the people who click on your ad are doing once they get to your site.
I'd get another pair of eyeballs to look over the landing page and the ad text. I just last week had to cut a landing page to ribbons, because it was so long - and I hated to do it, because I felt the information on it was vital, but I think it was scaring off the civilians; sure enough when we introduced the newer, shorter page (with links to more information if it was wanted) and moved the call-to-action even higher, the conversions went up. The thing is, I made the first page, and I felt it was great, but when my partner took a look at it (and he wasn't emotionally or timewise vested in it like I was) the first thing he said was "WAY too long" - and he was right. I'd lost the forest for the trees for a minute there.