On-Demand Document Management
Posted by Max Dunn Fri, 18 Aug 2006 18:05:00 GMT
Here are some of my thoughts regarding on-demand document management.
History
Back in the dot-com heydays, one of the promising market segments was Online Storage Services (OSS). These were the precursors of todays on-demand document management services. Some examples are Driveway, iDrive, MySpace, bigVault, Xdrive and iBackup, and MyDocsOnline. I worked with many of these companies since we provided a product TeamDrive that would turn any Internet file service into a virtual drive by providing drive-letter access.
Most of these were OSSes well funded, for instance, Xdrive raised over $120 million and Driveway raised over $50 million. They also had a lot of users, Xdrive had over 6.5 millions users, and Driveway had over 6 million users. However, none of these companies made it big, and most didn’t even survive. Why not? The big problem was that this was before Google AdWords, and they didn’t have a good way of monetizing their business model. Chris Logan, the CEO of Driveway, confided in me in late 2000 that he wished that all his users would go away, since they were costing more in bandwidth and storage charges than they were making. Tim Craycroft, the CEO of i-drive, told me that they were making some money in the educational market, but were being eaten alive by the the bandwidth costs of the software pirates using their service to distribute bootleg applications.
The OSSes that survived the dot-com crash of 2001 were the small, scrappy ones like bigVAULT, iBackup and MyDocsOnline. These didn’t have VC backing, but they also had much lower overhead. However, none of them have made it big. iBackup and MyDocsOnline continue as small, privately held companies. bigVAULT was just acquired by Digi-Data for an undisclosed amount after being kept afloat for years by the valiant efforts of John Salerno, the CEO and owner.
Market Future
While history has not shown a lot of success in the on-demand document management market, what about the future? There are some signs that this market might be opening up. Certainly the overall document management market is growing, especially with the need to support work-at-home programs and distributed offices, but also to comply with Sarbanes-Oxley (SOX), HIPAA and other compliance regulations. Previously, most enterprises wanted to own the document management systems and run them behind their firewalls, but with the success of SalesForce.com and other on-demand software services, many business are more willing and comfortable to let their non-core IT needs be outsourced. In particular, there is interest in K-12 schools and some medium sized businesses to out-source their document management and collaboration needs.
Open Source Challenges
Open source software is changing everything. When you can get high-quality software for free, and get the source code too so that you can enhance it and fix bugs, why pay for closed, proprietary software? It used to be that open source was limited to infrastructure software like operating systems (Linux), servers (Apache, Tomcat) and databases (MySQL, PostgreSQL). However, with the success of SugarCRM in 2004, high-level application are now being open sourced and the open source model is being embraced by the VC community. In the document management market Alfresco is a solid and rapidly developing document management system, and is continuing to make big waves.
As we see more-and-more high quality application software being distributed as open source, it is going to put a lot of pressure on the traditional, licensed software model. For instance, Documentum has been feeling the heat and several years ago started offering a low-end document management to universities for about $50k, way below their usual prices. Microsoft’s Sharepoint is also contributing to this price pressure since it is usually included at no extra cost for many companies in the overall Microsoft license.
However, even when the software license is free, businesses still recognize that the cost of deployment and maintenance is not. And there is a growing awareness that it makes better sense to try to out-source as much of the non-core business processes as possible. So reasonably priced on-demand software services, whether based on proprietary or open source software, will continue to be attractive.
Product Challenges
Here are some of the product challenges facing many document management systems that have a solid base of underlying library services. For those that don’t, they have many more challenges in bringing their library services up to par.
Upload/Download
The biggest challenge with web-based document management services is getting documents into and out of the repository, or being able to edit them in-place. Without plug-ins or downloads, there is a limit of uploading one file at a time due to HTML security restrictions. WebDAV support helps by providing Explorer based file management and in-place editing, but only for WebDAV enabled applications, of which there are few, most notably Microsoft Office and Adobe products. Products like Xythos Drive and WebDrive Internet enable any application, but require installation on the client computer.
Ajax
None of the current document management web interfaces make good use of the Web 2.0 technique of Ajax. A complete redesign of the web interface with these principles in mind would make it a lot easier to use and much snazzier.
Role Based Access Control
ACL based access control can be a real mess on large repositories. Some document management systems try to simplify access control by allowing only folder level permissions, but this quickly becomes a limitation, for instance, when there is just one file in your sales folder that you would like a certain customer to see. But for systems that allow both folder and file level access control, and maybe for both users and groups, there is usually no auditing capability of access rights so it is impossible to determine exactly who has access to what. And even with access auditing, the access control lists are usually so convoluted that the reports are useless. To solve these problems, document management systems should start to incorporate real role-based access control.
DoD 5015.2
Whenever trying to fill records management needs, or even to help with compliance regulations like SOX and HIPAA, DoD 5015.2 is going to come up. While this standard is not necessarily the best records management system for non-government organizations, it is almost always on the requirement checklist for customers with records management needs, and document management systems without it won’t make it to the short list. However, achieving compliance with 5015.2 and being certified is a very difficult process and will usually require deep changes to the document management system.
Workflow
Workflow is hard. When group hasn’t used workflow before, they will usually believe that their workflow needs are easy but will soon discover there is a lot more complexity in it than they originally though. For instance, maybe a group of 3 people need to sign off on all press releases. Easy right? But what happens when Mary is sick? Can the press release be assigned to someone else for approval, or should it just skip Mary’s approval? Should there be a timeout so that if Mary is out for a long time and the document is waiting for her approval, an alternate action can be taken? Questions like this are only the tip the complexity to workflows but illustrates how the term “simple workflow” is really and oxymoron and document management systems really need a well designed, complete workflow system.
Folksonomy
There is a lot of work in Web 2.0 being done around folksonomy and personal tagging. Document management systems should embrace this informal method of adding metadata to documents to avoid the tediousness of mandatory properties and rigid taxonomies which will encourage more users to participate in the tagging process and improve search results.
Future
Document management systems represent an intermediary solution as we move from a file-based information system to a web-based one. Why use a separate program to create an external file that needs to be uploaded, stored, downloaded and viewed in the separate program when we can simply put the same information on a web page that can be directly viewed and even edited by authorized users? Of course all of the infrastructure of document management systems will still be needed, like access control, versioning, metadata tags, workflow, etc, but the focus will shift from storing the information in externally created files that need to be moved around and managed, to creating and viewing the information directly in the content management system. In the future, we will see a convergence of document management systems with wikis, blogs and forums to create truly useful information management systems.