|
Interesting. A couple of observations, from grappling with some of the same issues for our negotiation site. Definitely agree that you wouldn't want to share the seller's bottom line in the event of a match. This would lead to built-in buyer's remorse. I gather this may have been for initial proof of concept, not long-term functionality. But you also need to think about the impact of masking the algorithm (see below) so it can't be reverse engineered (on the flip-side) and consider what information can and should be given about the process. Agree with one of the comments below that slowing down the result may make for better user experience. This was (surprising) feedback we received for impassebreaker.com -- people felt "let down" when the match was revealed too fast. There's a game component here, and a big splash may be important. Then again, it does seem a bit deceptive to show a progress bar when the result is already calculated. Subjective observation, but based on a decent amount of feedback. Users need to trust that the process is fair. Trust is hard. If you place the bottom line in context -- e.g., show your user what the comparative price is on other sites -- would that context make the bargaining process smoother and build more trust? Might it also make the site stickier? If your provide the comparison shopping data, then people won't need to leave the site to do comparison shopping. That may or may not make you more appealing to sellers, so perhaps it's not on point. As it stands, the process seems arbitrary -- I totally understand that this is an early sketch, but if/as you build it, I'd keep this in mind. I can't remember the cites, but I believe there's some research out there that suggests that people are more inclined to trust computers in a negotiation than other people. So you have a potential advantage if you make the playing field seem a bit more level. At the end of the day, what you may be building is a trust machine. Maybe it's different, but that's how I've thought of our platform. Negotiation style is culture and language dependent. The verbal dynamic that works in the Suk won't necessarily work in Omaha. One isn't better than the other, of course. But if the goal is closing sales, you need to keep context and language in mind, and at a very granular level. Words and how they are presented are key. Maybe that's part of what the system learns. Best of luck. There's some potential here, long term, for something very cool. I'll be interested to see where you take this. |