Predictive Governance: Making The Most Of Your Information

Tuesday, June 25, 2013 - 15:34

The Editor interviews Dean Gonsowski, Associate General Counsel and Senior Director, Business Development, Recommind.

Editor: Welcome back, Dean. Recommind has announced its Information Governance Suite (IGS). Please tell us about this new offering.

Gonsowski: IGS is a combination of practical software solutions designed to help organizations manage their information governance challenges. Historically, organizations have considered disciplines like e-discovery, privacy and information management to be discrete, and they have attacked the corresponding data challenges in a siloed and reactive fashion. Organizations increasingly recognize the commonality among tasks associated with unstructured information, such as data handling, management, searching and collection. As a result, they are looking for a holistic solution in which information governance serves as the umbrella concept.

IGS leverages Recommind’s legacy capabilities around search, categorization, predictive coding and e-discovery. It provides a comprehensive solution for addressing the risk and cost of keeping and managing information, and for optimizing the business value of that information.

Editor: When considering information in terms of business value, risk management and cost, is it helpful to have a clear understanding of the difference between records management and true information governance?

Gonsowski: Yes, and that’s a good point. Information governance is not a new concept; however, it has traditionally been treated from a records-management standpoint. The old idea was to take unstructured information and structure it by archiving and then manually categorizing it and applying record-retention policies. While this strategy makes sense in theory, we now know that it doesn’t work.

Archives have two big flaws. First, it is often very hard to access information for ingestion into the archive. Email data is a possible exception, but what about blogs, tweets, file shares and all the other unstructured content? If you can’t archive this information, then you can’t manage it all holistically.

Second, the governance pertaining to the use of archives is very end-user driven; meaning it relies on custodians to manually categorize information and file it according to a file plan. Then, assuming those tasks are done properly, the software can take over and make decisions about retention and disposition. This process is flawed because it depends on a level of energy, enthusiasm and consistency that we humans, even the very best corporate citizens, cannot offer. This is not the result of capriciousness, but simply a reflection of the reality that people cannot match the retrieval, search and management capabilities of today’s machine-learning systems.

In response to these historical flaws, Recommind developed a suite of different, but coordinated solutions, which rely on a single interface and platform. And our “secret sauce” is the machine-learning aspect, which goes beyond the usual tasks of ingesting or managing data and applies functions, such as predictive coding, to solve the problems of manual categorization.

Editor: Is IGS a predictive governance solution?

Gonsowski: Yes. Fundamentally, IGS is a predictive information governance platform. Our clients have been leveraging many of these capabilities for a while, so in that sense, we are not so much launching a new product as formalizing what we already do into a consolidated platform that clients can customize as they desire.

We have seen a number of use cases at this higher level, including compliance monitoring, data migration, defensible deletion, records identification and the application of policies, plus various knowledge-management and business-analytics functions. We find that once clients have gone through the hard work of locating and ingesting their data – and then indexing and enriching it, such as by harnessing metadata – they realize they have concurrently built the capacity to do e-discovery and a dozen other tasks.

Put simply, IGS provides an “index once, use many times” opportunity, and interestingly, it was some of our very early clients who pushed us to this next horizon. They were already committed to Recommind’s software and were successfully using it, and they expressed great interest in deploying their existing solution and now-familiar skill sets to subsequent use cases.

Editor: That sounds like the holistic approach you mentioned. How does this tie in with the idea of governance/policies that companies establish at a very high level?

Gonsowski: This is a critical point. Our software is very good, but there is no “Easy Button.” The organization has to decide what it is trying to accomplish as a whole. What are its records-management and records-retention goals? Does it operate in international jurisdictions with stringent rules about privacy and information security? Is the company also trying to accommodate other policies, such as those related to HIPAA, Graham-Leach-Bliley or FINRA?

IGS provides the tools to implement company policies. Organizations have unstructured information that may be helpful to them for business purposes but that also may present liability issues, let’s say, in the HR context, if any content might be considered discriminatory, or in the FCPA context, if information is suggestive of bribes. Therefore, the question becomes how to harness data that serves a positive business goal and possibly delete or file away data that presents risks or other challenges. Right now, this is a very difficult task, so organizations tend to fall back on familiar solutions: storage and a “keep-everything” data policy. Contrary to popular wisdom, enterprise-class storage is very expensive, particularly after adding in the costs of procurement, ongoing management, backup and security.

There is also the misperception that by keeping data forever, you preserve the option to use it beneficially in the future, but this is true only if the data had some semblance of value in the first place. We know that a lot of unstructured information is duplicative; perhaps it consists of music videos, or it’s SPAM, irrational or outdated. In other words, there is almost no foreseeable scenario in which it will be useful; therefore, there is no reason to keep it, and further, it may actually pose risks and drive up costs.

Editor: Let’s shift focus and talk about the process of adopting a solution like IGS from the ground up. Do you suggest a compartmentalized approach?

Gonsowski: That’s a great question. E-discovery has been tremendously successful as a newer discipline, but there was a lot of pain in the process, including pressures relating to judicial deadlines, the possibility of sanctions and the fact that everything was mandatory. There was no room for “analysis paralysis,” but there was substantial, even cathartic, motivation for legal, IT and security/risk people to start working together.

The best way to implement an information governance initiative is to form similar cooperative groups. In many ways, this process can rely upon relationships that were created in order to deal with issues like e-discovery, i.e., as a discrete challenge, only now the issue may be something like data migration to the cloud. While these compartmentalized situations are substantial, they don’t reach the daunting proportions of a wholesale, multinational adoption. Obviously if you adopt the information governance philosophy, it leads to significant organizational changes; however, you need to walk before you run, build up experiences that deliver a good ROI, and develop an appetite for taking on bigger projects.

Editor: There are other governance solutions on the market. Please talk about some of the exceptional qualities of Recommind’s solution.

Gonsowski: First, not many solutions offer true information governance, and many that use this label are focused merely on transforming an existing, reactive e-discovery process into something proactive. While this solution may be helpful in limited situations like defensible disposition, it is not really offering the full value of information governance.

Our clients want to embrace a longer-term, holistic and mature vision of information governance. A model adoption plan might run in three phases: cost minimization, risk reduction and then the actual performance of tasks like moving, deleting and filing data. The important aspect is reaching beyond the risk/cost processes and into activities that really derive value for the business unit. There, the idea is about putting the right information into the right hands at the right time. It gets people excited about spending time and money on an information governance solution.

This is Recommind’s long-term vision. We focus on predictive analytics, and the infusion of machine learning is an important differentiator for IGS. We view the governance area as a bridge between a reactive e-discovery process and our forward-looking tools, which enable information governance that derives business value under the even larger umbrella of big data.

Editor: Is Recommind’s solution also unique in being modular?

Gonsowski: It is. Once clients install CORE, which is our indexing and data-enrichment platform, they can deploy modules pertaining to e-discovery, defensible disposition, categorization and other tasks. But, that’s just the tip of the iceberg. Over time, we will build additional modules, such as might allow customized access to CORE through an application programming interface (API) or other capabilities our clients request.

In short, offering a modular product goes to scalability as a company grows, and it reflects the long-term vision of always meeting the full roster of client needs.

Editor: Would you expand on the concept of scalability?

Gonsowski: I mentioned earlier that humans cannot match the information capabilities of software systems. In today’s world, information exists in terms of volume, variety and velocity – all three in large and increasing proportions. In addition to problems relating to the inconsistencies of manual data assessment and categorization, we face literally impossible odds in trying to manage big data without the help of machine-learning tools. And, let’s remember that these tools don’t function without guidance and training from humans.

Savvy institutions like multinational banks and pharmaceutical companies understand their exposure as they move from the cloud to social media and to a bring your own device (BYOD) environments. Data is getting out of hand, and, for example, a healthcare institution cannot place sensitive patient records at risk of being breached. So, to the point about scalability, these organizations will not be prepared for operations, even just a few years forward, unless they adopt a scalable solution for dealing with ever-increasing volumes and complexity of data. And really, it’s a matter of best practice to develop a sustainable information culture rather than bootstrapping in the midst of a crisis.

Editor: Tell us about a specific use case that Recommind handled.

Gonsowski: The U.S. Department of Energy’s office of Energy Efficiency and Renewable Energy (EERE) was struggling with how to make sense of its data, which was located in unstructured repositories like SharePoint. EERE used Recommind’s CORE engine and the IGS platform to categorize information into buckets that, once established, formed the basis for applying defined policies around retention, migration and legal holds. In essence, the categorization process delivered a framework for understanding the data and then managing it strategically, which were EERE’s intended goals.

Editor: Where can our readers find out more information about the Predictive Information Governance suite?

Gonsowski: Recommind’s website,, is a great resource, not only for information about our solutions, but also for visitors to learn more about information governance – from white papers written by third-party experts, analysts, case studies and our dedicated blog ( Our first and most critical step is to ensure that people have a consistent understanding of what defines and distinguishes true information governance, and we believe that educated clients will naturally look for a solution like IGS.

Please email the interviewee at with questions about this interview.