Social Site Search

Fast, which may be a company best known for specializing in site search, has launched a product platform that is looking to socialize the ecommerce storefront search function. It’s called FAST Recommendations, and it is based on offering product recommendations similar to those of, but with a social twist.

“The basic idea is that search is very good for helping people find things that they know exist. What it doesn’t do as well is help surface things that might be interesting to you, things that you’re not aware of. That’s where recommendation comes in,” said FAST Search and Transfer’s Vice President of Product Management Joe Lichtman. “What FAST Recommendations leverages are relationships between content. It adds in a strong element of what users are looking at, and sort of a social aspect of how are the other users like you and what they are interested in, and can it find other products that may be interesting to you.”

In order to provide such recommendations, FAST constructs consumer profiles, or personal profiles, for every customer based on user behavior. Profiles are made up of data from recent searches, product views, purchases and reviews submitted or read. The software then generates relationships between like products and content based on interactions shoppers have had with an online storefront. Then end result are recommendations, all in real-time, anywhere on the site.

While the recommendations are like those of Amazon, and include related-item recommendations, FAST is looking to present search results that are much more individually relevant. “The main difference is that our technology for recommendations, and even personalization, is based around a social model where we identify and relate people in the profiles to various products within the content,” Lichtman said. “As a result, we also create ad hoc groups all the time to define demographics within the system. As an example, I don’t know anything about heavy metal music but I need to buy something for my nephew. If a retailer wanted to do something to the effect of getting recommendations from somebody who’s the expert on heavy metal music, the software could give me a recommendation around heavy metal music, or show me the people within the profile list that are in the recommendation service’s mind the most knowledgeable or interested in heavy metal music, and show me what they recently purchased.”

By learning the penchants and interests of each shopper, FAST believes it can assist ecommerce businesses in providing what it calls a personalized “micro store” for each customer, through its personal and social recommendations.

Such personalization, Lichtman believes, is what differentiates FAST Recommendations’ site search feature from any other product that is currently available in the ecommerce space. “Everything you do on the site, when you do a search, the recommendations that come back are personalized, even the search results can be organized based on your activity in the past, what we know about you, and the navigation could even be modified to present to you, for example, the genres within a category that you have shown more interest in the past. We believe that sort of a social search and recommendation service can really improve the site experience.”

Lichtman also said that implementing an advanced site search feature could immediately help small- to medium-sized business level the merchandising playing field “One of the beliefs I have is that nobody has the staff to do what they want to do in terms of merchandising on their site. This includes some of the biggest retailers I’ve worked with. I would imagine any small- to medium-business would be more so, and that their ability to have somebody figure out what merchandising messages should come out for their products is limited by the fact that they are a small company. I would argue that what we’re doing with a personalization and recommendation platform built with search is that we’re giving them one tool that is an automated way to do a lot of what the big guys are doing, and this would allow them to compete directly in that fashion.”

The social version of FAST Recommendations, which will enable users to request suggestions from a specific individual or expert, has been available since mid November.

Prices to integrate FAST’s social enterprise search start at approximately $100,000 for the licensed software. Additional factors that may affect an annual fee include website volume, quires per second and site traffic. Additional information is available at

Pat Callahan
Bio   •   RSS Feed