Fornova Enables You To Trawl And Aggregate The Deep Web

“We aggregate the Web”, declares web content aggregation service provider Fornova (www.fornova.net), which provides B2B turnkey solutions for classified advertisers, e-commerce and other specific content verticals. The company claims to be the only one in the world that has a solution that can automatically aggregate content from an unlimited number of websites without the need for site-specific customization.

Fornova was established in the beginning of 2008 by a group of three entrepreneurs headed by Michael Rubanovich, who were previously working at Intel leading development of chip design tools. The founders soon joined a small private incubator led by Amir Freund, who had previous entrepreneurial experience and later joined the company as its CEO. Amir brought in an additional experienced entrepreneur, Dori Stein, to handle the marketing and business development. The company is backed by a group of angel investors and to date has raised more than US$1 million.

Aggregating The Deep Web

The idea for Fornova came about from understanding the challenges involved in searching the ‘Deep Web’, says Michael Rubanovich, Fornova‘s chief operating officer. The first version of the technology was originally developed for monetizing arbitrage situations in betting, but they decided to shift their focus after realizing the difficulty in trying to scale the company in that market, he says.

“With the ever-expanding amount of content available on the World Wide Web, many surfers are finding it harder than ever to quickly and accurately pinpoint the information and products they need,” says Michael, adding that the founders identified a real need that some companies needed to provide quality content for their customers, such as e-commerce affiliate networks and price comparison engines. So they set about creating a solution that allows the aggregation of information from the ‘long tail‘ of classifieds sites, providing the consumer with the highest coverage of all offerings available, that removed duplicate listings and an algorithm that allows the ranking of listings based on quality.

“Publishers in e-commerce, classifieds and local market segments want to have the information normalized across source sites with unified structure and same set of data items,” Michael says. “They want to reach the long tail of merchants, agencies, dealerships source sites without the need to pay high price to establish aggregation from each one of them.” Poor content, he says, will reduces the chances that consumers will click on a listing and buy, and also creates a bad user experience.

Fornova’s platform uses artificial intelligence technology that imitates the human understanding of web pages. “By using visual analysis algorithms combined with a robust semantic engine, the platform can accurately scan thousands of sites, without the need for a per site adaptation,” explains Michael. “This reduces significantly the amount of efforts and costs involved in providing a high coverage solution. The platform is capable to extract all data items much more accurately, using much less computing resources. It enables us to provide a very high level of content quality and freshness.”

Understanding And Targeting Market Segments

Michael says that one of the key challenges were to identify main market segments to target, considering that their technology applies to so many verticals and applications. It was also critical to establish the right business model when approaching customers in those segments, he emphasizes.

Fornova‘s key markets are the e-commerce sector, such as affiliate networks, price comparison engines and product search portals, as well as the classifieds industry, where they target large media and newspaper companies, online car and real estate aggregators, and online job solutions. “Our revenue model is a licensing fee, based on the amount of sites we cover or amount of content we provide,” Michael shares. “In some cases, we create partnerships to establish a new leading solution and share the revenues and ownership of the new solution.”

Michael highlights the example of how the company once helped an online jobs solution provider that was previously using manual labor to develop scraping scripts for job boards and job agencies. “The customer was using manual labor to ‘tune’ the received information, paying more than 25 cents per listing! We enabled the customer to scale the coverage of its service, getting information of required jobs directly from companies’ sites career sections, removing the high cost and time required to establish scraping,” he explains. After implementation, the new solution automating the process completely and reduced per-listing costs by more than 50-percent.

One of the key concerns with such powerful aggregation technologies is how it can be open to abuse, for example, the infringement of intellectual property rights through the illegal scraping of content from a content owner’s site. To that, Michael says Fornova pays close attention to the legal aspects of content aggregation and search. “We try to be involved only in projects that have synergy between the source sites and the aggregation based applications,” Michael elaborates. “In e-commerce, this is usually the case, as there is an agreement with the merchant.”

“For classifieds, we omit the need to scan competing aggregation sites by addressing directly the agencies, dealerships and companies. They all want to get traffic and are thus interested in participating in the service.” Michael adds that they support opt-out mechanisms, and do not crack security measures such as captcha codes.

Control Your Success

For the founders of Fornova, being an entrepreneur means that they can take control of their future and success. “We want to create a company that makes a change and provides a home and a place to grow for our employees,” says Michael.

“We have learned to always aspire for high targets and to persist to achieve them.” For inspiration, Michael says they look towards this reference from Alice In Wonderland: