IBM infoMarket: A place where publishers can cell their wares

Secure superdistribution of proprietary content on the Internet

(This brief description if IBM's infoMarket is assembled by Gisle H. from various sources.)

IBM's infoMarket system consists of three core components:

All three technologies - invented, acquired and licensed - complement each other in the infoMarket service.

IBM infoMarket also consists of an IBM operated clearinghouse which will enable commercial content providers with the management of content access and payment processing.

Finally, there are three so-called Plug-N-Publish toolkits to enable content providers, technology partners, and customers to customize infoMarket service's capabilities to meet their specific needs.


IBM introduces ground-breaking rights management and secure payment technology to bring leading commercial content online

Cryptolopes* enable publishers to Plug-N-Publish* safely on the Internet

Boston, October 31, 1995 (press release)

Continuing to unfold its comprehensive strategy to make the Internet a secure and profitable marketplace for online publishers and digital content providers, IBM today revealed a number of new partnerships and technologies as part of the IBM infoMarket service, a network-based service from IBM.

IBM's infoMarket is an online, searchable general information directory that allows you to search for free and pay only for the information you need. The research service searches through multiple commercial and Internet databases and retrieves relevance-ranked results list.

IBM infoMarket service allows users to simultaneously search, and retrieve content - including still images, text, sound and video - from distributed databases in a wide variety of formats, hosted on a number of popular publishing platforms. This capability gives digital publishers greater control over how their content is presented to potential customers, since they are not limited by the technical requirements and presentation formats of typical commercial online services.

The aim is to create the first secure environment for intellectual property owners to reach a worldwide audience of millions over the Internet. This new service will provide digital publishers with the incentive to create and distribute their content electronically, with the knowledge that they will receive compensation from every person who makes use of their work. Additionally, IBM infoMarket service will supply the tools to provide these content owners with maximum control over how their information is presented to their customers.

The service will serve as a clearinghouse for its commercial content and service providers. It will manage the access to content, services and payment processing directly, on behalf of providers, or by referring the financial transactions to a clearance center, such as the Copyright Clearance Center. All marketing, editing, conditions of use, and pricing for the use of commercial content will be set by the providers themselves, giving them greater control over distribution and profitability than any other electronic publishing outlet offers.

Major sources: 75 news wires, 300 newspapers, 819 newsletters, 6882 journals, 13 000 Usenet newsgroups; and 11 million companies in a variety of areas including: Business & Finance, News, Computers & Telecommunications, Health & Biotechnology, Sciences & Engineering, Environment, Law & Government, and Industries; plus leading Internet directories such as Yahoo!, The McKinley Group's Magellan, and Open Text Index.

Jeff Crigler, vice president and managing director of IBM infoMarket says:

"With today's IBM infoMarket service announcements, we have made great strides in advancing the Internet as a secure and profitable marketplace for both publishers and consumers. We are addressing three persistent problems: how to efficiently find and retrieve precisely what users are looking for; how to protect intellectual property from piracy once it is placed on the Internet; and how to ensure that content owners are paid for the use of their works, even as they are redistributed to others. IBM infoMarket service is the first service to address all of these problems, and provides digital publishers with the incentive to create high-quality content for distribution on the Internet."

IBM infoMarket version 2.1 available

From: Tim Byars
Date: Fri, 24 May 1996

We are pleased to announce IBM infoMarket version 2.1 - available April 29th. The foundation of this new release is a Cryptolope(tm) container - a secure packaging technology allowing users to buy copyright-protected information right from their desktop. Cryptolope containers can include text, digitized images, video, and audio.

IBM infoMarket's powerful search capabilities enable users to search through a rich archive of content that includes previously unavailable commercial content, popular web indices, and USENET newsgroups. Our search service will be free of charge - you pay only for priced content you decide to purchase. Registration is not required to conduct searches, however, you must re-register for the service at the time of your first purchase.

Visit IBM infoMarket today at http://www.infomarket.ibm.com [defunct] and take advantage of Cryptolope(tm) technology - tap into the wealth of published information now available quickly and securely through the Internet.


An ONLINE Interview with Jeff Crigler at IBM infoMarket

ONLINE, July 1996
Copyright © Information Today, Inc.

By Jeff Pemberton

IBM's new infoMarket product offers some new solutions to information retrieval and delivery over the Internet. Jeff Pemberton recently had the opportunity to spend a morning in the Virginia headquarters of the infoMarket operation, where he interviewed Jeff Crigler. Crigler is the father of the infoMarket concept, and here he shares his vision for the future of information on the Internet with Jeff and the ONLINE readers. -NG

JKP: Where did the idea for IBM infoMarket originate?

JC: I started in the information business in a circuitous way. I started out building an expert information system for political professionals - for lobbyists and for others - many, many years ago, after spending seven or eight years working on Capitol Hill. We developed, in effect, an early distributed database application for lobbyists called Lobbyist Systems Corp. We were noticed by LEXIS-NEXIS and acquired in 1987. I spent eight years working at NEXIS as a director in product development.

About three years ago, I had an epiphany about where the information industry was heading, and the impact of the public data network - some people call it the superhighway, some people call it the Internet. I perceived that we were headed into what I would call the disintermediation war. The result of my epiphany was to realize that big, centralized, traditional online services, business online services, were going to have a difficult time competing in the world of the Web, in a world where there were no barriers to entry for publishers to go directly to customers.

I started thinking about what the new business model should be for the next generation of online information retrieval services. It's a marketplace where publishers could come and sell their wares. The marketplace owner would provide the utility services that publishers need for billing, information retrieval, security, etc. I looked around for someone who might be interested in building such a business, and at that fortuitous moment IBM decided it really needed to become active in network applications. So I wrote IBM a letter and the rest is history, as they say.

JKP: Could you tell us about IBM infoMarket? What makes it unique compared with some of the other systems that we're seeing?

JC: Well, first, it is not a centralized repository of information that we intend to resell to a customer. That isn't the model.

In the traditional online business, essentially you are a reseller or a repackager and you have license arrangements with a bunch of different publishers. The value-add these services provide is packaging and allowing a user access to multiple sources so they don't have to go individually to each one to get information. Of course, publishers are only getting 20 or 30 percent of the end-user's dollar, and now that the barriers to entry are down, they want to go directly to those customers. That presents a problem, both for the publishers and the customers. The customer doesn't want to have to go to a hundred individual publishers, so you still have the problem of aggregation. And from the publisher's point of view, it's how do you get noticed with thousands and thousands of Web sites going up every month?

Our answer is a different model - it is to create a three-way relationship. We have a relationship with the publisher that says I'll bring you customers, I will do your billing and collect money on your behalf. I'll provide a set of utility services to help you make the contact and get noticed, available, and purchased on the Web. With the customer, I have a relationship in which you will pay me for these services on behalf of those publishers.

The important difference, and what makes us unique, is that the customer belongs to each individual publisher. So as a publisher on infoMarket, you have a direct customer relationship with every person who buys your information. Each is your customer and I am like the credit card vendor. In a traditional model, I'm helping facilitate the "buy transaction" between the publisher and the customer.

JKP: So publishers, then, will know the identity of their customers?

JC: Absolutely, it's their customers. I'm facilitating a transaction but it is their customer.

JKP: So you will be sending reports that not only give the amount of usage, but the identity of who used it?

JC: When you do a search on infoMarket and retrieve a set of abstracts, maybe some of the documents are from ONLINE magazine. When you choose to purchase a document you have created a contractual relationship between yourself and ONLINE. ONLINE has just acquired a customer, and infoMarket is keeping the billing records and maintaining the back-end office support that the magazine requires to service that customer.

JKP: So it functions like the phone company - I have an ongoing account and when I make a call, I am billed.

JC: That's right, you'll be billed at the end of the month - you don't have to fill out a thousand different registration forms for a thousand different publishers.

JKP: Who are the customers that you expect to reach with infoMarket?

JC: First, we have to distinguish between the architecture and the technology that we're building. The technology is applicable to many, many different markets, and in fact, what we're doing in the business information market, we'll be doing in others. Now we are specifically going after users of business information in high-tech, consulting, and some other fields that we think are particularly ready to use the Web as an information retrieval resource. Not all companies are ready to jump onto the Internet and use it as a vehicle for doing research. Some companies are, some markets are, and so we've focused on those. And you will see a very rich Internet research service that pulls together publishers who cater to those markets in the initial version of infoMarket.

JKP: Who actually maintains the data files?

JC: Most publishers are going to host their own content. Some smaller ones may ask us to do that for them. But for the most part, they're going to host their own information, in their own Web sites, on their own servers. So they manage their own information, they control what it looks like, and how it's priced. That's very important, the publisher sets the price, not infoMarket. We tariff the rates against their price at 35 percent or so, depending on the model, sometimes lower or sometimes higher, depending on the services that they want us to perform for them. But it's their customer, they set their price, and we base our charges to them on their volume.

JKP: Do you do any hosting yourself?

JC: We do, and for that we charge a little more. We can also work with third-party people to do hosting. If you as a publisher already have a database through some third party, we can connect to it. We'll come to you, you don't have to come to us. You don't have to take your material, stick it on a tape and ship it to us, we don't want to be in that business. We want to create a distributed, decentralized service that links publishers, rather than creating a big warehouse in the sky.

JKP: Can you see a time when you might deal with individuals?

JC: Absolutely, absolutely. We have a service that we're working on right now that would be directly targeted at small publishers and author/publishers, who'd like a simple way of making their wares available on the Internet and getting paid for them. It'll be something very simple, where, for example, you could email your newsletter to infoMarket. com on the Web and we would pack it into a cryptolope container and send it back to you. Now you, the author of that newsletter, can distribute it throughout the Web - and anytime anybody opens it up to read your newsletter, IBM will handle all the back office billing and send you a check at the end of the month.

JKP: What is a cryptolope container?

JC: IBM cryptolope technology is going to change the fundamental nature of publishing. It's an example of something called a secure container. It's a technology that really alters the nature of information management.

When I'm finished authoring a document, the software that I used allows me to save the file as a cryptolope container. Among the options that I can choose to invoke as I save the document are things like its price and what clearinghouses will be used to collect the money. I can insert discounts as to who can use it and for how much. I can have different prices for different usage - so I can have one price to just read it on the screen, another price to print it. I can have special discounts for book reviewers or librarians. I can also set specific usage limitations so that you can make so many copies or you can redistribute it under certain circumstances.

So there's a rich set of permissions that I can electronically glue on to my document as I stuff it in this secure container. Then I can distribute the container without worrying about how it is encrypted. That's the secret of the success - it will allow you to distribute information freely across the Web. The point in our service is not copy protection, it's payment assurance.

JKP: Let's say I've done a search and I've got some output in a cryptolope container that's been downloaded to my computer. Now I want to look at it. What happens at that point? What if I want to share it with a couple of colleagues?

JC: Two things will happen. First, there are the Type 1 cryptolope containers that we're coming out with very soon in infoMarket. The only key you can buy in this first implementation is the unlock key. It will unlock information and make it available to you with the same copyright restrictions that apply at LEXIS-NEXIS.

In a Type 2 cryptolope container which we'll have available later this year, you can purchase a view key which will be much cheaper than an unlock key. With this you can view or read a document - and you could pass the cryptolope along to your six closest friends. When they open it, they will chose whether they want to use a view key, a print key, or an export key.

Now it sounds very complicated, but imagine that the application will be a helper app that's part of Netscape and is transparent in the background. When I click it will open and ask do you want to purchase this, print it, or just view it? If I say "view," then behind the scenes the helper app will come back to the clearinghouse and retrieve the appropriate key, cause the money to be transferred from your account to the publisher's account, and so forth. The user will not notice any of that stuff.

JKP: Wait a minute, who's collecting the money at that point?

JC: InfoMarket debited your account and credited the publisher's account.

JKP: Suppose I want to put that document in my company-wide database?

JC: That would be one of the options. Essentially you could buy a subscription to that information for your users. We've created something we call our Rights Management Language (RML). It's a way for publishers to describe what rights they wish to convey with the document. It is language to describe those rights that is glued on the top, into the content inside the cryptolope container. The applications that read RML will know how to behave appropriately - to allow or not allow you to do certain things with that information.

JKP: What do you see as future applications for cryptolope containers?

JC: I think they will change the way business information is retrieved on the Net. But think about cryptolope containers carrying all sorts of other information - advertising, music, video. I can put an advertisement in the cryptolope container and pay you to view the advertisement. Direct marketing applications are potentially very good uses for the technology. I think it will have applications in every sphere of information consumption.

JKP: Is the technology proprietary to IBM?

JC: No, it is available to any one who wishes to use it. We will be making the cryptolope technology available to the Internet community. We think it will be good for the industry if these technologies are adopted. We hope to make it a standard.

JKP: What's your immediate goal?

JC: My immediate goal is to collect a compelling, critical mass of information for business users. We're well on the way to doing that. We're going to launch with many thousands of sources of business information. IBM InfoMarket will be the only place on the Net where you can do a comprehensive search for commercial information - and also for the free information that's available on the Web.

JKP: What advantage does infoMarket have over the traditional proprietary systems?

JC: Well I think the traditional proprietary systems are in a mad dash to reengineer their businesses for the Internet world of tomorrow. Some have chosen infoMarket as one of their strategies. You're familiar with IAC, for example. These and others are actively working with us. Some of the larger traditional online services have their own strategies for how they're going to re-engineer their business.

JKP: Do you see a substantial decrease in the cost of information?

JC: I believe in most areas prices will come down and fairly significantly - both because of super distribution and because the barriers of entry are gone. Anyone can get into the space now, all you need is a Web site. The cost to play has gone down so much that you can now reduce the prices.

The other thing that I think is going to change is the notion of charging for a search or access. That is a model that the Web doesn't understand and that users don't understand. I think people do a head scratch and say why are you charging me to walk into Macy's? I'm very eager to buy something, but you want to charge admission to get in the door?

Also, what we're headed for is very, very fierce competitive price pressure for information. I think it will be very difficult to sustain higher prices for substantial value-added information because it will be so easy to get it from the lowest price performer. An example is government information. You really have to provide some very substantial added value when users can go directly to GPO's bulletin board and download information.

Cryptolope containers, for example, will allow publishers to keep track of how their information is used. That, in itself, is a value to the consumer. Who liked this article, who read it, how many times has it been read? Has this article been rated by any number of several different rating services? Those are things the publisher can put on their information to make it even more valuable and we'll help them do that.

JKP: IBM has several products - Lotus Notes Newsstand, infoSage. Can you differentiate between them?

JC: It's all a family of content services within IBM. It's all part of a strategy at IBM to unleash the power of content, generally, and to increase the value. So this is part of the overall networked computing strategy, it's not just scatter shock. Right now, infoSage is a current awareness product. InfoMarket will be the archival capability behind infoSage, making the two very complementary.

Lotus Newsstand is a little bit different, that I would characterize more as an Internet opportunity for publishers to create subscription pricing for their customers within a given enterprise. Our model is more pay-per-view, you can buy just a document.

JKP: What other content are you looking for?

JC: We are in discussions with several large book publishers, we think there's a tremendous opportunity for using cryptolope containers to distribute electronic books of various kinds. We're looking at various markets like the education market.

JKP: Will you use cryptolope containers for things like books?

JC: Or what we might do is help publishers use the cryptolope technology to start their own electronic book publishing services on the Web. I might not do it directly but I might help, I might work with publishers, because they're important customers of IBM.

I'm not sure that we want to be in that business, we're looking at it. We don't want to compete with publishers - they're really our customers so we don't want to compete with them. If book publishers come to us and say we'd like to use what you've got, then that's something we'd look at very seriously.

JKP: What search engine will infoMarket use?

JC: What we have is called "plug and publish" technology, which allows us to use any search engine, or almost any search engine. We are creating a standard around something called GQL. Your readers will be familiar with SQL in the database world - we're working on something called GQL for text and multimedia.

We've announced agreements with almost a dozen search engine providers who will support our "plug and publish" interface. Also behind the scenes we will support the search engines Folio, BRS, BASIS, Verity, and PLS. So as a publisher it's really up to you, you decide how you want to store the information, what is your preferred search engine, and we'll plug into you.

JKP: If you host the content, what search engine do you use?

JC: Well we use several different search engines - they all have advantages and disadvantages. We take a look at the content, we look at how it's being stored now by the publisher, and decide what the optimal search engine is, including IBM's own search engine which is Search Manager.

JKP: If I do a search on infoMarket who decides which search engine, what happens?

JC: What we use is Minerva - technology that we buy from Booz Allen Hamilton. Essentially it works by creating what is called middleware. This talks to a server that talks to the World Wide Web. So what we're really doing is a distributed service - multiple databases that don't have to be on the same platform, so they don't have to be in the same search engine. We bought Minerva, we bought the technology from Booz Allen and it's the basis for our service - it's the thing on top of which we've built infoMarket.

JKP: When can we expect to see really significant increases in bandwidth - to allow full-motion video, audio, and so forth?

JC: Don't worry about the bandwidth, it'll be there long before you need it. The bandwidth will be free - this is almost a religious thing, people seem to have different opinions about the bandwidth problem. It's a little bit like being in 1983 and worrying about when CPUs are going to be fast enough to have GUI front ends. Believe me, it'll get there.

I think we need to focus not on the bandwidth, but on useability. The problem, as Gerstner keeps saying, is that somehow I have never had to get up in the morning and reboot my refrigerator. But for some reason I have to do that with computers - so what's the problem here? I think we should be focused on ease of use and making this stuff friendly - that's our mission with infoMarket.

Communications should be addressed to Jeffery Pemberton, Information Today, Inc., 462 Danbury Road, Wilton, CT 06897-2126; 203/761-1466; jeffp@infotoday.com.

Copyright © 1996, Information Today, Inc. All rights reserved.


IBM Retreats On Cryptolope

Ambitious market plan abandoned; software to become part of Notes

Reprinted from Web Week, December 15, 1997 © Mecklermedia Corp.

By Whit Andrews

The IBM Databolts product group will close, and the technologies it developed, including its Cryptolope superdistribution software, will be parceled out to Lotus and to IBM's Internet commerce division.

The decision closes a chapter on IBM's attempt to develop the superdistribution technology as a standalone product and opens one on its potential integration into existing product suites. Lotus groupware may benefit from the security-on-demand aspects of Cryptolope technology, for instance, said Karl Salnoske, IBM's general manager of Internet applications technology.

But any future for the technologies will be as they are incorporated into existing software systems - a step back from the ambitious plans IBM had for them, in which the company would have built and maintained transaction clearinghouses for countless pieces of Internet content. That clearinghouse approach was part of IBM's InfoMarket initiative.

The decision to deliver the technologies into other divisions comes three months after IBM announced it would build a Java-operated version of the system. That product, Cryptolope Live, will be abandoned.

"Trying to launch that as a standalone product wasn't the right way," Salnoske said. Instead, it may be worked into not only Lotus software but also the IBM Net.Commerce line of electronic commerce software.

"If, a year from now, we can make a case for it as a standalone toolkit, we'll go back to that," Salnoske added.

Meanwhile, other technologies developed by the Databolts group, including early search, push, and filter software components, will be turned over to Lotus, Salnoske said. Their developers will also go to Lotus, he said, and some developers who worked on Cryptolopes will go with them. The remainder will probably join the commerce software development teams, he said, adding that no firm decisions have been made.

A technical executive who previously worked in the InfoMarket division said numerous engineers from the team have left or been laid off. No independent confirmation of that could be obtained. An IBM spokeswoman denied layoffs will occur.

Analysts characterized IBM's decision as a sensible response to a technology whose prospects had seemed murky from the beginning. Big Blue's experience with complex technology had served the company well in developing it, but turning it into a cohesive product line had proven elusive.

Lotus will likely be able to use the technology better, said Eric Brown, a groupware analyst at Forrester Research. "The encryption technology, messaging and delivery, and document creation and viewing are all things the Lotus guys are good at," Brown said. "It's an extremely good idea to fold the Cryptolopes into the Lotus team. It shows good vision on IBM's part."

But the delay between the initial public release of Cryptolope technology 18 months ago and its current circumstances may have helped IBM to decide it would be better folded into other products.

Not First Project Abandoned

The dropping of the division involved in Cryptolopes comes almost exactly a year after IBM abandoned another content-oriented initiative, InfoSage. The personalized news delivery service had been another facet of the InfoMarket initiative that used filter technology. IBM decided InfoSage was pitting it against some of its own customers in the content business, and chairman Louis Gerstner announced the abandonment of the service in a speech at a conference last year.

IBM and Lotus last week were just starting to discuss how Cryptolopes, which have to do with securing documents, might best be integrated into the document-centric universe of Notes.

While the technology's future may be brighter as a part of larger software suites, it will certainly be diluted and seems unlikely to play a sweeping role of revolutionizing information distribution in the fashion that Jeff Crigler, chief of the technology group, had envisioned.

"Rights management is highly complex and requires a great deal of focus, and packaging it as a component would likely be defocusing for them," commented Victor Shear, chief executive of InterTrust, which markets a somewhat competitive product. To succeed with it, Shear said, IBM would need to have a sizable team of engineers focused wholly on the technologies and the business models around it.

"I think Cryptolopes, and InfoMarket in general, was a very good idea," said an executive who had been involved in both. "But Cryptolopes were never a product - I think that may have been the problem."