We’re working with a customer at Elevate, and with pure Data Science, we were able to predict what task codes to apply by reading the narrative, unstructured outside council task descriptions within invoices, and using AI and automation to apply the right task code. With pure data science, we were able to get 80% accuracy, and this was a while ago, we were just testing it. When we worked with subject matter experts, we were able to get to about 90% accuracy. And again, this is a capability that’s just using test data before it went to market.
We’ve developed a scalable process to embed subject matter expertise, truly people and technology, not just lip service because the people know best and wrapped the technology around that. So, what surprised us was how clear the advantage was when you applied business subject matter expertise to processes and data to, basically, when something could be interpreted one of two ways, and a subject matter expert knows best, you embed that understanding in your tool, and it performs better.
Nicole: Is that the calibration we hear about? It sounds like it’s larger than that.
Brian: It’s more than that. One of the most useful things about design thinking, for example, companies that use it, generates over 50% more revenue, according to McKinsey. By design thinking, I mean designing processes, designing software, designing services, designing products. One of the things we always want to avoid is digitizing the cow path. We don’t want to digitize bad processes. So, the ability to work with customers to understand what a day in the life looks like today, understand everything that goes into it from the end-user’s perspective.
Again, the subject matter expert who knows better than we do, and we can bring supporting subject matter expertise, but you, the customer, know you best; you’re the use case. And, we’re embedding that process and making it scalable, getting expertise out of someone’s head and into a tool is actually a fundamental problem of AI across industries.
And if that can be accomplished through understanding your processes and understanding your data and mapping them back to clear business results, I can tell you what I think you need. But we’re getting to the point where technology is so configurable that you can tell me what you actually need, and I can adjust accordingly.
Nicole: Where do you see either the most commonplace that you could implement digital design thinking, or what is the one that has the biggest return?
Brian: So, if you look at $1.2 trillion that was spent on digital transformation last year, $1.2 trillion, we wonder if this is coming to the legal industry. It’s a third of all IT spending globally, Yet Digital Transformation is broadly construed, and it’s not just using technology. It is, again, not digitizing the cow path. You have to reinvent the processes that you want to digitize. The value of design thinking and where to apply to it, you know you need to invest in technology if you’re a law firm because your customers can buy the same tools you can and serve themselves with them increasingly.
Forty-eight percent or more of all legal spend is moved in-house at this point. If you’re a law department, you bought, paid for, own, and store work product that you created in the past. But you spend seven hours a week, approximately 20% of your time per lawyer per year, looking for information. Why not find information that is out there, and repurpose it in the context of your current need? Well, what does that mean? What information should you start with?
If you could automate something or reuse your data, those are the two big picture use cases for Digital Transformation across industries; it’s process automation and data reuse. And, you know, 78% of all data in the legal industry that law firms have, that law firms possess, go unused, and we can argue about how useful that data is. But usually, if it’s not directly useful, it’s useful as a data point to tell a story or support something because there’s more data. Well, you know you have to do something, or you want to do something, but it can be very difficult to identify a starting place. Even if you have a specific pain point, it can be difficult to identify the size and shape of that starting place.
Design Thinking is a structured process for identifying a starting place in the same way that lawyers use a structured process for preparing for trial, for M&A transactions, for drafting contracts, etc. This is a structured process for ideation, one that allows for the identification and prioritization of use cases that are technically feasible, not Star Wars, that are economically viable, what can you afford to do, what do you want to spend money on doing, and that people will want to use. Which, in a roundabout way, brings me back to probably the greatest driving change behind everything that we call Digital Transformation.
So, yeah, okay, technology’s becoming more configurable and less expensive. Well, people are using it to get closer to the end-user. In fact, 70% of Fortune 500 CEOs, the reason they invested in AI, broadly speaking, is to get closer to the end-user. Because again, technology prices fall, it’s about how can I show my end-user, my client, my customer, “know-me” insights in air quotes. Well, identifying where to start involves a lot. If you’re dealing with a toolkit full of Legos, how do you configure those Legos so that you achieve your business case, and it’s not a random walk or experiment at the edge of the enterprise?
Design Thinking allows you to identify a practical starting place with digital technologies, whether they’re products or whether they’re capabilities that can be assembled in a custom way around you to create your own custom solutions. Put little AI and analytics and robotic process automation in, stir and see what happens. So, identifying a practical, testable starting place before you make an investment in technology, reducing the risks and costs associated with buying or building technology, improving adoption through better business and end-user fit, and then building a road map for transformation, So maybe it is one project, and that’s all you want to do. But maybe it also involves taking the so-called data exhaust from your ELM platform, seeing what you can do with it to create that new insight.