top of page

How do Corporate Legal Organizations “get” to Business Intelligence … and Beyond?

The combination of many factors over the past 20 years, including the global debt crisis

and subsequent “great recession,” has left many legal departments around the world confronted with the omnipresent “do more with less” budget mantra. That has often resulted in a “bring more in-house” approach. Combine this reality with the ever-increasing data flows into legal departments and we are confronted with, not only an urgent demand for better data security, management, and storage capabilities but also a tremendous pressure to capitalize on this data with Business Intelligence (BI) capabilities.

Presumably, you are convinced that BI is relevant to your organization and quite possibly have embarked on a BI development project. If your organization’s BI experience is like most companies, the results have been less than satisfying. Don’t be dismayed. You are in good company, and there is a reason to be optimistic.

So, how do we successfully convert this plethora of data into a robust Business Intelligence platform with the limited IT resources and data science talent available?

Process Improvement (PI)

A BI platform relies heavily on integrated data. Most data is created and stored across a wide variety of systems. For example, Matter management/Ebilling, IP, Finance, Contracts and HR software systems, to name a few. Regardless of where your organization is on the BI continuum, the first step to developing a useful BI platform is understanding that integrating data is actually integrating processes. If the work-flow processes that generate the data are inconsistent, the data will be inconsistent, and consequently, the analytics become unreliable.

Between 70% to 80% of corporate business intelligence projects fail, according to research firm Gardner

If we consider the scope of operational processes in a legal organization we can see how the various Software as a Service systems (SaaS) utilized to support each of the organization’s functions, combined with the personnel involved in the data input and administration will often, unintentionally, result in “de-normalized” or inconsistent and redundant data.

Very often, each system has a unique version of shared data, fields, and entities. Consistency is lacking from system to system. This is commonly referred to as a “single version of truth” (SVOT) problem. Couple this with the fact that many or all of the systems are “siloed” and you have a data integration nightmare. Some technology can help resolve these issues, but improving the data processes at the SaaS level will do more to ensure an efficient BI platform than anything else.

Below: Example of Operational Processes in Legal Organizations

Below: Intelligence Infrastructure

Process Improvement is an effort to minimize the SVOT problem, as much as possible, before implementing a BI solution. It should identify the sources of data that everyone agrees is the trusted input for their various systems then operationalizing it across all systems and workflows. System administrator and user training are essential to data consistency. And a data governance policy will ensure the initial BI implementation, and future BI data integrations are reliably orchestrated.

To obtain sustainable process improvement we know the following steps:


  1. Secure Executive support and establish solid IT and vendor partnerships.

  2. Develop a clear and detailed understanding of the BI insights the legal organization is seeking.

  3. Ensure all stakeholders can communicate with a common language by identifying and closing any knowledge and data literacy gaps with cross-discipline training between Legal, IT and Finance personnel.


  1. 4. Map the key data requirements for the Bi Platform.

  2. Develop and implement an enterprisewide plan to insure data veracity and consistency across all SaaS systems.

  3. Establish admin and user guidelines for each SaaS system utilized, implement data governance and audit/control policies to ensure a smooth initial implementation and future integrations.


  1. Identify the data sources that are relevant to the metrics and analytics the Legal Organization is seeking.

  2. Determine whether your organization has the resources to build a BI platform or if buying one is more appropriate for your organization.

  3. Implement a platform that will foster a "discovery environment" for its users. It should go beyond mere reporting and data visualization and deliver 4 specific categories of analytics.

Business Intelligence (BI)

Business intelligence is frequently thought of as just a visualization tool. Ideally, all data sources from the legal systems should be integrated and automated at similar time intervals, not static reporting that requires human data retrieval. Also, a Legal BI platform should foster a "discovery environment" for its users. It should go beyond mere reporting and data visualization and deliver four specific categories of analytics.

The first question to resolve is whether to “build or buy” your BI platform. BI analytics is a multidisciplinary project. Does your organization have the budget? Access to the right talent? And maybe, more importantly, the time to build the software and infrastructure? The demand for data scientist has skyrocketed. Data analysts and engineers, Python/R programmers, cluster computing and Linux admins are going to be critical for an in-house build scenario.

If your legal organization has both the budget, time and talent to build and continuously support a BI solution in-house then building BI from the ground up is an approach that may achieve a better ROI in comparison. If your organization has some or none of the above resources then buying a BI platform solution is more likely a feasible alternative.

BI Platform Functions

Whether you decide to buy or build, a stable BI implementation should look something like

this and have most or all of the features below.

BI Platform Functions Defined

Data Ingestion

The data ingestion should allow for orchestrated extractions with connectors to accommodate any data source with data-at-rest and data-in-transit encryption. It should also allow for a data push or pull strategy to comply with your security protocols.

Data Warehousing

The data warehouse should offer immutable data storage and recovery features for both structured and semi-structured data for the next step of data transformation and insight delivery.

Data Transformation

It should support data cleaning, partitioning, and segmentation features that use common SQL language for transformations. Ideally, it should allow for the integration of third-party or public data to create more relevant insights (macroeconomic, demographic, industry-specific, social data, etc.)

Insight Delivery

It should support the ability to embed analytics into the user’s workflow with customizable interfaces. It should offer a variety of visualization types, such as charts, geo charts, dynamic tables, KPIs, alerts, and reports that can be exported in many formats for collaboration and insight sharing. It should allow you to grant permissions from a high to a granular level to drill in and drill across the data. It should enable skilled users to discover new insights with advanced predictive, statistical, mathematical, and text functions.

Data Security

It should deliver multi-layered approach to information security and ideally be Compliant with Service Organization Control (SOC) 2 Report for Security and Availability Principles under AT 101 Management, the TRUSTe® Privacy Program, HIPAA compliant. Abiding by the EU Data Directive EC/95/46. On track to comply with EU Data Protection Regulation GDPR before May 25, 2018.

Augmented Intelligence (AI)

What is beyond Business Intelligence? That is a question that requires navigating the extreme opposites of science and magic. Artificial intelligence was invented in 1956 at Dartmouth College in the U.S. by engineers interested in how computers could be used to model the human brain. There were, and still are, many different approaches to AI, so the term “artificial intelligence” is quite vague. We are purposefully using the term “Augmented Intelligence” because most applications in this space require human input and are therefore not truly artificial. Artificial Intelligence, as it exists today, is best described as a suite of technologies. Machine/Deep learning, Natural Language processing/generation, and Neural Networks are a few AI technologies making progress in the Legal space.

If you only pay attention to the news headlines touting AI’s advances it would appear AI is progressing at a breakneck rate. While it’s true AI research is intense and becoming more so the technologies are not being commercialized at the same pace. For now. It seems unlikely that a “killer AI app” will materialize anytime soon to solve all your organization’s Business Intelligence needs. However, these technologies are advancing, and in-spite of the marketing hype, they are technologies your organization cannot afford to ignore.

We feel the rational approach to capitalizing on these technologies is by having your organization become expert with existing BI technologies and implement a BI platform that can adapt to these emerging technologies. This approach will enable your organization to be better judges and consumers of AI technologies and implement them into your platform as they become useful.

There is another compelling reason for implementing a robust BI platform. A commonly overlooked or misunderstood issue for AI technologies, especially for machine learning and neural networks, is the need for relatively large and reliable data sets. The algorithms used require training, validation and testing data sets to function. It also needs sufficiently large enough ongoing operational data to allow for continuous learning. Having a well structured BI platform should not only deliver the data sets required for AI but also help inform your organizations' planning and implementation of AI technologies.

Buyer Beware

Software and vendor evaluation is essential.

Some of the tools in the legal space where AI is being marketed fit squarely in the fake or semiautonomous categories. We are not aware of any fully autonomous technologies available yet for the legal industry.


The path to a successful Business Intelligence initiative starts by considering how data originates across the organization’s day to day operations and then establish protocols to enable data consistency across the various systems. Empowering users by embedding the analytics in their workflow will maximize the discovery environment that Business Intelligence offers.

Developing a BI platform that goes beyond just visualization tools and delivers predictive, prescriptive, diagnostic and descriptive analytics that will drive the value proposition to an entirely new level.

Future proofing your organization with forward-thinking BI designs will allow your organization to leverage the inevitable deluge of AI technologies that can give your organization the efficiencies and competitive advantage it needs to thrive in an ever-evolving business environment.


About the Author

Jennifer Vandersmissen is a Co-Founder of The Business Intelligence and Process Improvement Group, (BiPiG).

BiPiG offers a technology platform that delivers a robust Business Intelligence solution designed specifically for legal organizations. It is scalable to any size organization, agile enough to adapt and utilize Artificial Intelligence technologies and capable of delivering embedded analytics directly into the workflow of its users. BiPiG also offers Process Improvement training courses and consulting.

#JenniferVandersmissen #NeedtoRead #AllUpdates

bottom of page