I worked for a while on dynamic packaging models (user clicks on "flight+hotel", books the flight, and then is presented with a list of hotels at their destination). We used time between booking and date of travel as one of the signals, and saw that there was a pretty clear correlation between planning ahead and being more price conscious. I don't think it was a huge effect but it probably bumped down the target price of our offered hotels by a buck or two.
Unrelated side story about this: we also ran some experiments with bumping higher margin (usually meaning higher price) hotels into the top ranking slots to see how it affected conversion percentages. Normally the top slot has the highest conversion and it decays sharply from there. With these experiments running, for most countries you could see a clear change where the peak would move down to slot 2 or 3 or wherever the "natural" #1 hotel was (score one for our models, I guess). This effect was most pronounced with German travelers, and almost nonexistent for French ones.
Unrelated side story about this: we also ran some experiments with bumping higher margin (usually meaning higher price) hotels into the top ranking slots to see how it affected conversion percentages. Normally the top slot has the highest conversion and it decays sharply from there. With these experiments running, for most countries you could see a clear change where the peak would move down to slot 2 or 3 or wherever the "natural" #1 hotel was (score one for our models, I guess). This effect was most pronounced with German travelers, and almost nonexistent for French ones.