At any rate, for these design elements, we use Google Website Optimizer. The idea is that instead of saying, “that won’t work” or “well, i don’t agree with you, do it anyway” you say, “hey, that’s an idea that we can test out on the site.” You just set up the variations you want to try, and run until you get enough data to draw conclusions from the experiment. It’s all very scientific, but it’s a lot easier than someone might think, and everyone can feel good about their contribution.

The other thing is promotions. You can’t just come in one day and say, “[competitor] is doing a 20% off sale, so we need to do a 20% off sale right now.” There are profit margins to consider, and whether or not there’s a holiday coming up, when was the last time a sale was run, are the customers going to stop buying without offering a sale if you run too many sales, when is the next e-mail newsletter going out, will we have banners on the site, will we notify our affiliates, are we using a coupon code or are we putting the products on sale for everybody, etc.? Many people don’t realize all the ways that can be explored to make a sale a success, and instead get a banner up on the site, call it a day, and then try to figure out why traffic wasn’t up and sales weren’t as high as they expected. A better route would be to try every available channel to promote the sale, and then analyze what worked best. Combine that with different landing pages tested with Website Optimizer, and you can really start to figure out what your customers like.

The bottom line: everything must be tested scientifically. Buy-in will be higher from the team that makes it happen, and you’ll be able to use what you learn to make the next site change or promotion an even bigger success.