The art, science and unicorns of data analytics
It’s been an exhausting but exciting week of nerdism. After spending last weekend with Tony and Arya Stark (unrelated), I spoke at the Ark Group’s Business Intelligence and Analytics in the Legal Profession conference in Chicago and enjoyed the many presentations and conversations throughout the day and evening. So much so that I am close to nerd-overload, but #nottoday. So here I share some brief thoughts and takeaways from the conference.
Business intelligence is not analytics
I often hear these terms used interchangeably along with “dashboards”, “Tableaus”, “infographics” and “that report with the chart”. This is beyond semantics: data analytics is different than business intelligence (BI) or competitive intelligence (CI) or artificial intelligence (DUH). Especially in legal, where BI typically represents centralized and general financial reporting. This is certainly an essential capability to have, but not all firms have this perfected yet. Particularly with “advanced” business concepts such as profitability.
Simply put, analytics is the capability to analyze, which requires analysts. Go figure. And we can analyze lots of things: business and financial performance, a new business opportunity, and even the practice of law…but analytics always starts with a question to answer, or a problem to solve or a hypothesis to validate. This is how business intelligence fundamentally differs from business analytics. BI is generally good for reporting, sometimes dolled-up with charts and graphs, but without an analyst to interpret it BI is just information at best.
Because the insight-driven analysis required to solve problems or assess opportunities often extends beyond data from a single source, it requires finding and stitching together the right types of qualitative and quantitative analyses to synthesize into an evidence-based point of view. And after speaking with many firms, the capability to do this is quite a challenge to build, buy or both. Presenters described resources with these skills as “data scientists”, “market analysts”, “translators” or “storytellers”. But the most fitting description was “unicorn”.
Data visualization and storytelling are essential but rare skills
One presenter, a Director of Client & Market Intelligence at an AmLaw firm, talked about his team and the work that they do. He shared a set of infographics that summarized trends and opportunities within specific industry sectors and firm practice areas. These were produced by collecting, organizing and analyzing a combination of firm transactional data and third party company and market data from D&B, Yahoo Finance and other sources, then “storyboarding” the desired infographics based on questions being answered and who is asking them, and ultimately handing off to a graphic designer to deliver the final product. The end result was excellent, IMHO.
When asked about the skills required to produce these documents, he described analysts with competitive intelligence, market research, data visualization and storytelling experience coupled with legal industry domain knowledge, familiarity with data tools and resources, and, and…a colleague sitting next to me leaned over and said “I need to get the names of these people!”. Yup…🦄
In my talk, I described BI and analytics as the first mile and the last mile of decision making: the first mile requires tools and resources to get usable data while analytics covers the last mile of interpretation and translation required to generate useful insights and answers.
The fact is, at the end of the day executive committees and other firm leaders and decision makers don’t care about the first mile effort, and don’t really want reports, dashboards or even beautifully designed infographics; they want a briefing, ideally from a knowledgeable unicorn that can translate for them. And the last mile to produce and deliver this briefing is the hardest but most valuable part of the analytical process.
As one of the rarest of these unicorns that makes time to work AND write, I strongly suggest that you read Jae Um’s post “An Rx for Data Dizziness: Getting Started with Data Requires Smarter Strategy” over at Legal Evolution to dig deeper.
Data quality vs. sufficiency
During my panel discussion I suggested using the term “data sufficiency” over “data quality”, as it begs the question “sufficient for what” and forces a conversation about context and purpose. But it also subtlety points out another difference between “traditional” legal BI and analytics. Finance-led BI projects quite naturally focus on the absolute accuracy and completeness of the numbers included. But with most analytics it is acceptable to be directionally correct. An analyst simply doesn’t need GAAP certified financials to describe future trends and models that are, by definition, educated guesses. I’ve personally seen this derail analysis efforts due to collisions between CFOs, CSOs and their teams.
Sufficient data for analysis often goes beyond the data found in firm financial and BI systems, and analysts need to identify, collect and correlate data from a variety of sources. This is where law firm business analytics projects benefit from competitive intelligence and research experts. But alas, I have observed a distinct lack of collaboration or coordination across these functions in many firms, if they exist at all. Data analytics is a team sport and requires multi-disciplinary expertise that span departments and silos, which is not easy in any large organization let alone a law firm.
Another challenge facing analysts is that traditional law firm tools for business intelligence and reporting generally do not offer sufficient flexibility when it comes to data analytics. Most BI solutions today are built on traditional database platforms using dimensional data warehouse designs. This essentially defines the destination for your data to analyze in advance, which is great to ensure consistency and accuracy but very limiting when exploring data for insights. Analysts often need to iterate through different datasets, experimenting and testing different models and schemas in order to zero in on useful results. This includes adding new data, or inferring new attributes from existing data. Most data warehouse solutions and the reports and dashboards in front of them cannot support this flexibility, making the “last mile” work tedious, costly and error prone.
A more modern analytics stack must include more sophisticated data visualization tools such as Power BI or Tableau, and more flexible data capture, storage and processing tools such as schema-less database systems, data lakes and graph databases. While broad platforms such as Microsoft Azure and Amazon Web Services can provide the raw technology for these solutions, I’ve yet to see an end-to-end law firm analytics solution on the market that meets all of these needs. This is a huge opportunity, IMO, for traditional BI, reporting and data management vendors to modernize and expand their value.
Analytics requires new tools, techniques and *talent*
I’ve said it before, and I’ll say it again: the innovation wars to come will not be fought over tech, robots and business models. They will be battles for talent. Business talent, such as unicorn analysts, skilled data experts familiar with a range of modern data tools and techniques, expert researchers and competitive intelligence professionals, and story tellers who can pull it all together to help firm leaders understand options and make better decisions.
These resources are not inexpensive and are increasingly difficult to find and recruit. But firms without them might as well have blinders on when it comes to making better decisions, smarter investments and playing the long game.