Solve E-Discovery Challenges Posed By SharePoint

Monday, May 3, 2010 - 01:00

Editor: Let me ask you about the scope of your business and your clients. Does most of what you do relate to e-discovery?

Kiersted: Yes, our primary business is e-discovery and various related activities. We also have a very strong consulting group that helps corporate legal departments effectively use technology in matter management systems, legal holds and other areas. Even before e-discovery was an industry, my focus was to help people with discovery and tracking. I started my company in 1985, and now we are very fortunate to serve many smaller clients as well as very large corporations at the top of the Fortune 500.

Editor: Tell us about SharePoint and its impact on electronic discovery.

Kiersted: More and more of our clients are adopting SharePoint, and it is changing the way enterprises manage data. Multiple parties can edit and track versions of documents and collaborate more effectively. SharePoint also brings smoother workflow to processes, and its search and integration with other Microsoft Office applications means faster access to information. However, it presents complicated challenges during electronic discovery, particularly in the collection and review phases, and this is becoming a major issue for our clients.

Editor: Explain the issues behind producing SharePoint content for review.

Kiersted: The challenge of electronic discovery in general is to make sure nothing is lost that could be relevant in litigation. This includes metadata, which is data about data. Take email. The obvious content is the subject line and the body. The metadata is when it was sent and when it was received and who exactly received it, and there is a ton of information beyond that. For documents, there is a lot of metadata such as author and title and date last accessed.

With SharePoint, the metadata is multiplied quite significantly, and its richness rises to an even higher level. The metadata must be preserved so the review team can mine it for meaning using assessment tools and search capabilities. Failure to properly harvest the data will deprive you of the ability to produce that metadata in a proper fashion, which is a concern.

Editor: What are the issues with respect to new content types, like wikis, blogs and custom lists?

Kiersted: It's becoming a huge challenge to make sure that those items can be reviewed appropriately. For example, blogs are a continuing dialog, and an initial blog posting may be followed by a number of replies. Each one of those would be treated separately during a traditional document review. However, a reviewer needs to see all the documents within a blog in context and understand how the blog posting and the replies relate to each other.

Editor: Are there any other electronic discovery issues unique to SharePoint?

Kiersted: Another subtle, but important, aspect is how SharePoint works and the way companies deploy it. Now, people are more likely to include in their emails a link to a document as opposed to the more traditional practice of attaching a document. Using traditional e-discovery collection techniques to capture a mailbox, you might end up with an email that says, "See document below" with only a link and no actual document. Our techniques enable clients to collect the appropriate related documents, and make sure that any links take the reviewer to the actual linked document.

Editor: How do software and service providers help law firms and corporations cope with these additional complexities?

Kiersted: Companies are definitely recognizing that this is a challenge and are responding in different ways. Many are looking at very specific parts of the equation rather than the whole problem. Although the information is searchable based on keywords, most collection tools in the market today focus on libraries of documents that are stored within a SharePoint site. These tools help determine how much content might be responsive, but they are not fully capable of forensic collection. They grab the embedded files from the SharePoint document library but lose the context and metadata that is critical for both review and authentication as evidence. Other products may be limited to getting a single version, or only Word documents or Excel spreadsheets - nothing more sophisticated than that.

Editor: So what is Kiersted doing to solve this problem?

Kiersted: With blogs, wikis, documents and versioning, the challenge goes beyond how to find and collect it. SharePoint to us is just one of several different kinds of data that need to be managed together and provided to the review team in a way that permits reviewers to understand the meaning in context. By doing this, we maximize the effectiveness of the review team.

Editor: Describe the specific challenges that surround the versioning aspect of SharePoint.

Kiersted: SharePoint is a document management system that maintains multiple versions of documents. You may create a document and I may modify it, and then someone else may create a new version and you may go back and revise it. If you are not careful, you might capture only the most recent version, whereas previous versions could be crucial in understanding how relevant that document might be.

Editor: From an analysis standpoint, how is dealing with SharePoint different from more familiar applications?

Kiersted: A project for us that incorporates SharePoint data in the end is really no different from any other project where you have a mixture of email and documents. Clients can apply the full power of our integrated platform and analysis toolset to everything that is being discovered, whether it's paper, email, or server files.

Editor: Describe why your approach inspires greater confidence in the results.

Kiersted: It starts with really sound collection methodology. Other solutions that rely on exporting documents can lose metadata and context. Our simple but very effective way of using existing back-up techniques means that clients don't need to adopt or install any new technology to preserve and capture the content of specific sites within a SharePoint network or system. When we get that data, we restore it to a protected, read-only nonpublic stance that makes it part of our review platform where we preserve the richness of all of the content. This is then used in the context of other collected items to help reviewers do what they need to do.

Editor: I would assume that is essential in terms of a preservation hold where you are not permitted to affect the original documents.

Kiersted: When documents are stored within another system and not just as files, one of the real challenges is capturing and preserving them in a forensically defensible way. To ensure that collection techniques haven't altered them or caused any data to be lost, we created a process to restore the client's SharePoint content at a site level. That virtual site is locked down, and then we extract pages with attachments as individual web page archive files (MHT files) for processing into our K4 Review Platform. We extract all the electronically stored information (ESI) metadata and make sure that it is indexed, searchable and ultimately producible. Reviewers can go into the secured system to follow links and experience the embedded files in their actual context. Furthermore, our approach avoids any disruption to the client's use of the SharePoint application during normal business operations.

Editor: SharePoint sites can be massive. Don't review costs explode when you backup such large volumes of information? How do you avoid bringing in so much information that it becomes cost-prohibitive?

Kiersted: We are careful to avoid over-collecting. For one thing, we encourage clients to identify specific sites so we don't have to back up the entire network. Another way we contain the cost is through the culling mechanism at the heart of our platform. The very first step in the process is to start whittling the data down. You look for duplicates and relevance through search terms and various techniques. This allows you to take what potentially might be a large collection and narrow it down to a relatively small review set.

As you know, the review phase of electronic discovery is the most expensive because it requires attorneys' time and expertise to evaluate ESI and determine what is relevant and privileged. Our native review platform lets reviewers plough through a huge amount of material in a very small amount of time. It accomplishes this not just by clever user interface design, but also by some really powerful workflow and rules-based processes that offer confidence that the review process is being done properly. This avoids such things as coding inconsistencies, incorrect privilege calls and other error issues that force you to go back and make changes when your deadline is approaching - thus costing a lot of time and energy. Using these workflow steps makes review teams very efficient, and that is a feature that people really like.

Editor: How could this solution help the legal hold process?

Kiersted: When an issue comes in and a client does a quick search, they can put sites on temporary hold. They can harvest that content, cull it quickly and proceed to the next level of looking at it for relevancy before the hold expires. In other words, they can use this as a staging area for content that will flow through for processing and review. This tool lets them get down to the bottom of the funnel quickly even though they've poured a lot into the top of the funnel initially. This way, they no longer lose potential content and blow the preservation.

Editor: How does early case assessment fit into the picture?

Kiersted: It is a key feature. We call it dynamic case assessment because the same tools in our platform let you assess a case both early and over time as it evolves. The richness of metadata that SharePoint collects makes the early case assessment process even better because of the ability to slice and dice across new dimensions. Even though we are culling down the set that should be reviewed, you can assess your case based on a whole corpus of data. Rather than using just a standard set of fields of metadata, clients can also use their own systems and their own metadata to characterize their documents. Our process recognizes those custom fields and ingests them into the toolset we provide for early case assessment.

Editor: Kiersted's review platform seems to solve a lot of problems. Specifically regarding SharePoint, do the benefits of your overall approach extend into other areas of electronic discovery such as preservation?

Kiersted: Yes, SharePoint is a dynamic live system that constantly changes. Collection requests are ongoing, and our clients need to remain nimble. Our K4 Review Platform is flexible and has satisfied clients with many different challenges. It's rock solid and easy. We capture SharePoint content without losing context, provide it to the review team, and use the same tools we use for everything else. We help reduce the body of material in a defensible way. What happens when the information gets to legal and they want to know where this came from? We put the whole picture together, sophisticated yet simple. Nobody else is doing that.

Please email the interviewee at with questions about this interview.

Please email the interviewee at with questions about this interview.