Avoiding Over-Reacting
From my experience with iPrioritize and now with SportsLizard's Price Guide, I've learned that when you initially launch and get your first 50-100 users you'll get quite a bit of feedback - especially if what you're doing is a bit different than the status quo. It takes every bit of mental power you have to resist a knee-jerk reaction to that feedback.
In Getting Real, one of the most valuable lessons you learn is to listen to feedback but not necessarily react to it. With iPrioritize I scrambled around trying to tweak everything to make everyone happy - big mistake. They suggest a 30-day revamp, and I'm holding strong with that for the price guide.
But the price guide has been a bit different. The IDEA alone gets a lot of "holy crap that's an awesome idea" feedback. Where there are (and will continue to be) problems is with the pricing algorithm. It's hard to tell EXACTLY what item someone is trying to price out based upon what they enter in a search box. I don't want people to get a few wacky values and never come back.
My first thought was "I better step up this algorithm ASAP". Then I realized we will always be fighting a losing battle with the algorithm for just that reason. When I search for stuff, I can usually find what I'm looking for because I'm an experienced user. I know how to query the system and go through a few iterations to fine tune the results, but a first time user doesn't.
So the solution is really simple. Today I'm going to add a "training" section to the FAQ, and link from it from below the search box so everyone can see it. It'll say something like"learn how to get more accurate prices."
Would I prefer to have the perfect algorithm? Of course. But even Google doesn't have a perfect algorithm, the difference being that if Google slips in a few bad results you'll just ignore them. If we slip in a few wrong items, the price is all sorts of inaccurate. So I think there's going to be a bit of a learning curve for people to work the price guide correctly and I'm going to have to do a good job of presenting and teaching them so that the learning curve is minimalized.
In Getting Real, one of the most valuable lessons you learn is to listen to feedback but not necessarily react to it. With iPrioritize I scrambled around trying to tweak everything to make everyone happy - big mistake. They suggest a 30-day revamp, and I'm holding strong with that for the price guide.
But the price guide has been a bit different. The IDEA alone gets a lot of "holy crap that's an awesome idea" feedback. Where there are (and will continue to be) problems is with the pricing algorithm. It's hard to tell EXACTLY what item someone is trying to price out based upon what they enter in a search box. I don't want people to get a few wacky values and never come back.
My first thought was "I better step up this algorithm ASAP". Then I realized we will always be fighting a losing battle with the algorithm for just that reason. When I search for stuff, I can usually find what I'm looking for because I'm an experienced user. I know how to query the system and go through a few iterations to fine tune the results, but a first time user doesn't.
So the solution is really simple. Today I'm going to add a "training" section to the FAQ, and link from it from below the search box so everyone can see it. It'll say something like"learn how to get more accurate prices."
Would I prefer to have the perfect algorithm? Of course. But even Google doesn't have a perfect algorithm, the difference being that if Google slips in a few bad results you'll just ignore them. If we slip in a few wrong items, the price is all sorts of inaccurate. So I think there's going to be a bit of a learning curve for people to work the price guide correctly and I'm going to have to do a good job of presenting and teaching them so that the learning curve is minimalized.
posted by Adam McFarland at
5/07/2007 02:34:00 PM
Links to this post:
Create a Link
<< Home