Still revamping its content management and archiving (CMA) division to try to become a leader in eDiscovery, EMC Corp. experienced a decline in sales for that group last quarter and analysts said it is late
EMC is looking for its September acquisition of eDiscovery software vendor Kazeon to give its SourceOne platform a boost. Whitney Tidmarsh, chief marketing officer for EMC's CMA division, said EMC's goal is to put together an eDiscovery suite that can take on Autonomy Corp., but the CMA division has been among the weaker parts of EMC, this past quarter posting a year-over-year revenue decline of 6%.
Analysts said the SourceOne family lags behind competitors in support of content types other then email, such as SharePoint and file shares. Tidmarsh said EMC's roadmap calls for support of those content types next year.
Enterprise Strategy Group analyst Brian Babineau said while Autonomy and others have a jump on EMC for supporting more content types, the market is changing.
"Obviously, [EMC's] Documentum has more competitors with things like SharePoint in the enterprise content management space," he said. But checking all the boxes on the list of content types to archive isn't the point anymore, Babineau said.
"Kazeon can get into any file system it wants," including SharePoint, "so EMC is looking to own the discovery application rather than only focusing on customers who want to archive to their back-end stuff."
Tidmarsh said the Documentum group has decided to work with SharePoint rather than fighting it. Last week EMC launched Documentum Repository Services for SharePoint, which would allow SharePoint to use Documentum as the back-end storage repository for enterprise content; My Documentum for SharePoint, for existing customers who want to use SharePoint with a Documentum Interface; Kazeon's collection integration with SharePoint repositories; and Captiva support for SharePoint as a destination for scanned documents.
SourceOne was conceived as a distributed architecture that could centrally manage multiple repositories, including those belonging to third-party vendors.
Adding pieces to SourceOne eDiscovery platform
The SourceOne line replaces EMC's EmailXtender and DiskXtender data archiving products. It consists of Email Management for email archiving, Discovery Manager for legal holds, and Discovery Collector, an appliance that automates data collection in multiple repositories for eDiscovery purposes.
When EMC launched SourceOne last April, it was working with private vendors Clearwell Systems, Kazeon, and StoredIQ Inc. as OEM partners. Now StoredIQ and Clearwell are being phased out, and Kazeon will be integrated into SourceOne as SourceOne eDiscovery by the end of the year.
"We knew we wanted to own the technology and picked Kazeon because of its end-to-end eDiscovery capabilities and architectural compatibility with EMC," Tidmarsh said. "The challenge is that neither Kazeon nor SourceOne is a household name yet."
EMC is also considering combining IP from Kazeon with IP from other software acquisitions such as Tablus to create a stronger, standardized search engine for the entire product portfolio.
A shifting eDiscovery market
Vellante said many first generation email archives were marked by scalability problems, and limited application support for collecting electronic evidence. "It's not all the vendors' fault -- the world asked for a solution and they came up with the best they had," he said. "Three years ago there were some companies that spent $30 million on email archiving and knew it wouldn't scale, but legal was driving the bus."
In general, Vellante and Babineau agreed that data archiving and eDiscovery technology has a long way to go, but were split on the ultimate endgame for these tools.
EMC has talked a good game about federating content repositories rather than "shoving it all in to one big archive," Vellante said, and then using tools like Kazeon to search it. But "search is a blunt instrument," he said. And today's approaches to narrowing down data sets to save on legal review still aren't efficient enough. "There's no such thing as an end-to-end eDiscovery solution until we have automated classification at the point of data creation," he said.
Babineau said this may be a theoretical "nirvana" for eDiscovery but doubted its practicality. "Categorizing and classifying everything at the point of creation is nearly impossible to do -- there's just too much data," he said. In the meantime, the approach of doing federated eDiscovery regardless of how data is stored is the most practical one, he said.