For example, experts estimate that it costs buying organizations $150 or more to process every paper based purchase-order, regardless of whether it’s for a $10,000 computer or, $10 worth of pencils. Self service purchasing environment could slash this cost to as little as $25. The overwhelming point is that these savings have direct bottom line impact. According to a recent report by Killen & Associates (a leading market research firm in Palo Alto, Ca. specializing in quantifying IT markets), a typical company will realize a 28% gain in profit when it reduces its total purchasing costs by 5%!

E-Commerce purchasing relies heavily on the use of supplier on-line catalogs - built for buyers to browse through, in search for requested items. However, on-line catalogs, exactly like their printed versions, are built with a ‘concept’ in mind. This concept dominates the look & feel of the catalog, the way it presents and organizes information and the techniques it provides for navigating and accessing the information. Moreover, this concept often dictates the nomenclature used to describe the products and their merits.

On the buying organization side, Internet-based purchasing systems use special servers to provide end-users with browser access to their services. The purchasing server typically resides inside a company's firewall and manages the purchasing process, providing end-users with the required self-service environment. Regardless of the exact purchasing scenario, the purchasing server is required to provide access to all the relevant and only the relevant product information - residing in disparate on-line catalogs, and accessible through extremely different conceptualization systems.

That is, some sort of mechanism is required to provide powerful and uniform access to different supplier’s product information. We call this very mechanism a Catalog-Integration solution.

2. Existing Catalog Integration Solutions
Currently, the only available solution to the Catalog Integration need is the least of all evils - namely, catalog aggregation. In this solution, supplier’s product databases (not catalogs as a whole) are duplicated onto the buying organization purchasing server. Then, they are exported into an aggregated database using a common format. In purchasing sessions, end-users access this centralized aggregated database rather than the supplier original on-line catalogs. This solution suffer numerous flaws. Here are the most obvious ones:

First, its setup and maintenance costs are tremendous. Usually, a special content aggregator (service provider) is hired to setup the initial content of the aggregated database. Then, a dedicated IT and content management staff is required in order to maintain this ‘mega’ catalog viability. Actually, according to the Killen & Associates report (mentioned in the overview section), some 30% of the overall costs of an Internet-based purchasing solution are in ‘catalog services’ (projected to reach $200M a year through the end of the century).

Secondly, it’s not flexible since changing the approved supplier list means starting the whole process of manual database incorporation and maintenance again. Not only is this process complex and resource intensive, it requires a great deal of cooperation and resources from the supplier side. This limits the scope of such a scheme to high-profile suppliers trading with high-volume buyer accounts.

Thirdly, product information is up-to-date only at the last database duplication-aggregation point. However, information items like product availability are not relevant unless up-to-the-minute accurate. Furthermore, keeping the product information up-to-date (literally, a supplier problem) becomes the buyer’s tough on-going problem.

Finally, the suppliers themselves dislike the very idea of aggregating their product and pricing data into a ‘common format’ database - thus, stripping it from its unique added value product presentation.

Obviously, the aggregation solution is very problematic. Yet, until recently, simply because no other convincing solutions existed, it was the least of all evils. Not any more. AGENTics Internet-based virtual catalog integration technology - presented in the next section, is a breakthrough that changes the Internet-based purchasing applications map and enables a whole set of purchasing applications and services to take place.

AGENTics Virtual-catalog technology, presented in the next section, is a breakthrough that accomplishes better results without the burden necessity of either maintaining an uneconomical aggregated on-line catalog at the buy-side or incurring the expense of a third party catalog aggregator.

3. AGENTics Virtual-catalog System
AGENTics Virtual-catalog Integration technology provides unified view, navigation and seamless access to product information residing in disparate on-line catalogs without aggregating the catalogs. That is, remote on-line catalogs are not exported, imported or duplicated. In fact, suppliers are not even required to provide any special or dedicated interface to their catalogs! Rather, the Virtual-catalog maintains no product information of its own. It is linked by means of AGENTics patent pending software technology to any set of supplier Internet catalogs.

In each purchasing session, end-users describe and identify their product requirements using a powerful product identification wizard. Then, the cutting edge Virtual Catalog architecture allows product information to be retrieved at a phenomenal speed from the suppliers’ original on-line catalogs. Advanced Natural Language Processing techniques are used in order to provide unmatched accuracy and relevancy of the retrieved product information. Overall, the process effectiveness is exceptional, allowing requisitioners to focus on the products they want to buy rather than on the ‘sourcing’ process itself.

3.1. Technology
AGENTics Virtual-catalog technology is based on a powerful server and thin-client model. Using a standard Internet browser (thin client), end-users access the catalog-integration server which manages a ‘virtual’ catalog. The catalog is ‘virtual’ in the sense that it contains NO product information whatsoever. Rather, it contains a patent pending mechanism, UNIClass™ (UNified Information Classification), which enables real time access to the relevant products in the remote suppliers catalogs.

UNIClass unified product classification resides at the catalog-integration server. It contains an entry for each product type available in any of the connected supplier catalogs. In case of identical product types from multiple suppliers it contains only one entry for the type. Each entry is ‘dynamically connected’ (by means of the patent pending technology) to as many product offerings as there are in the original supplier catalogs. Whenever a user specifies a product type, UNIClass is able to ‘dynamically’ retrieve all the relevant products from the supplier catalogs, in real time.

However, the most intriguing aspect of UNIClass is its creation. A specific instance of UNIClass (per a buying organization and its multiple suppliers) is created in a two phase process which consist of a short semi-automatic phase (30 minuets per catalog) and an automatic integration phase. The result is a straightforward system which is maintained effortlessly.

3.2. Benefits
AGENTics Virtual-Catalog technology delivers whatever is required from a catalog integration solution, and more. First, in the search process, end-users experience unified view and access to product information, no-matter what its format is in the target on-line catalog. Moreover, the technicalities of ‘where’ and ‘how’ the product information was gathered are intentionally hidden in order to provide seamless access.

Secondly, product information is obtained from right within the supplier catalogs. The most up-to-date information source available.

Thirdly, this is the most cost-effective and convenient scheme for both suppliers and buyers. Buyers are relieved from the tremendous effort of aggregating numerous heterogeneous catalogs onto their site. Also, supplier-connection setup costs are minimal since this scheme leverages their uniform e-commerce interface of Internet catalogs (which actually enables them to tap into a much broader customer base).

Finally, the related maintenance costs are also minimal since both sides are relieved from the burden of managing duplicated aggregated databases which require a dedicated IT team to manage.