Clearly, he's not getting enough conversions now to be able to get a statistically significant result in a reasonable time frame. I'm all for A/B testing, but

Clearly, he's not getting enough conversions now to be able to get a statistically significant result in a reasonable time frame. I'm all for A/B testing, but the problems here are at a much more fundamental level--it is about understanding your market, and matching your customer's expectations. There will be plenty of time for A/B testing, once the sales volume is there to support it.