Search vendor EasyAsk Inc. has decided to put its money where its mouth is. In the wake of a hugely successful deployment for leather-goods retailer Coach Inc., EasyAsk is introducing a new pricing structure in which it will charge customers based solely on the results EasyAsk's technology generates. Of course, the danger is that not every customer will realize the 200% gain in online sales Coach registered after switching to the EasyAsk search engine.

"We've decided to take the risk ourselves," says the vendor's founder and CTO Larry Harris. "The customer can feel confident of achieving the ROI they were looking for." EasyAsk would establish a period--probably 30 to 60 days--during which a random comparison would be conducted, routing customers to either EasyAsk or the customer's old search engine. After that test period expired, the "uplift" in the site's sales would be calculated, and a percentage of that figure would be used to establish EasyAsk's first-year fee. If no uplift occurs, then the software is free.

It's easy to see why EasyAsk found this model so tempting. In its test deployment for Coach, the tripling of sales translated to the product paying for itself in three months, and a one-year return on investment of 900%. Meanwhile, Lands' End, which consistently posts one of the highest conversion rates among online retailers, saw its online sales rise $4 million--or 10%--after it switched to EasyAsk. Compare that with EasyAsk's license-based pricing that ranges from $100,000 to $500,000, and it appears the growth-based pricing could be potentially lucrative. EasyAsk will continue to make its old pricing model available as an option.

Forrester Research analyst Harley Manning says tying pricing to bottom-line impact is a brave move by EasyAsk, but one that could backfire. Manning met with Coach, which refuses to do media interviews, and verified the impact of EasyAsk's system, but he also points out that the retailer's old search engine "was a horror show." He says EasyAsk's new pricing could attract companies that don't have a lot of money to spend on search engines at first, but the new model could hurt EasyAsk if customers fail to maintain the search engine properly and thus don't achieve desired results. "Long term, this is a dangerous strategy," says Manning. "Your product can be great, and people can still break it."