IAVC framework

This item was originally published by Forbes, where Raj Sethuraman is a member of the Forbes Technology Council.

The buzz around data and the business potential it holds is, at this point, not necessarily new. But it continues to grow louder, perhaps because unlocking that potential has proven to be a mystifying process for businesses across industries — especially as emerging technologies like artificial intelligence and machine learning are increasingly becoming part of the mix.

During my years working with data, though, I’ve constructed a framework that’s been extremely helpful for ensuring the information being collected actually breeds better business outcomes. In the legal space, for example, I've seen that data has tremendous potential to optimize workflows and manage spending — but not if it’s collected without a plan. I refer to my three-step framework as “IAVC” — Inject, Analyze, and Value Creation — and I’ve seen firsthand how it can help reduce many of the biggest data mistakes that teams and companies make. Here's how to implement it to improve your data analytics strategy.

I: Inject Data

The first component of this framework is to “inject data” — in other words, to receive and store data from multiple channels. All companies have data from numerous sources — partners, third-party vendors, customers, other business units and so on. Particularly in the wake of near-constant experimentation with artificial intelligence (AI) and machine learning, it’s extremely important to collect a substantial amount of data, as data is what feeds such technologies, makes them smarter, and helps automate processes. Plus, sometimes information you don’t realize is important holds significant value. During this part of the framework, you may unveil information that will prove vital to your business efforts.

As data is being collected, though, you can already be doing early-stage analysis and cleansing. Companies often over-collect data — and collect a lot of “garbage data” — because they fail to understand patterns of data as they relate to business applications well in advance. It’s important to collect only relevant data that is accurate and complete. To do so, work backward. Know what goal you are trying to accomplish. Is it internal, like improving operational efficiency by automatically generating reports, or is it external, like generating usage insights for your customers? Once you pinpoint the goal, you can connect the dots backward and focus on injecting data that supports it.

A: Analyze Data

Once the information is clean and collected, it naturally needs to be mined. This builds on the early-stage analysis I mentioned but goes deeper and encompasses a full understanding of the data’s characteristics and whether or not it’s valuable. Indeed, I believe the crux of data analysis is segregating valuable and non-valuable data based on business drivers. This requires an acknowledgment of the relationship between data points.

For example, consider a data point like average spend by legal practice area and business unit (PABU). The spend spread related to various legal matters, based on their type, jurisdiction and other factors, and the relationship between the number of matters opened, closed or resolved in a calendar period are both important parts of the analysis process. Such relationships help provide a full, 360-degree view of the corporate legal department in question.

As alluded to in the last step, this framework is often a push-and-pull between staying focused on specific business outcomes and staying open to possibilities. Mining and monetizing data usually requires experimentation. Though I’ve made the mistake of assuming otherwise numerous times, data will not give you an answer on day one. There have been times that I’ve tried to generate a pattern; for instance, I was collecting data about customer usage of products without accepting that experimentation with customer segments is part of the process. But experimenting with data is important because it offers the chance to create a hypothesis and validate assumptions.

Experimenting with causal relationships among different data sets is a great place to start. These experiments can be run quickly and can generate new hypotheses to test more deeply while building on existing business knowledge and market understanding. In corporate legal departments, for example, you can uncover the relationship between network fees and invoice reviews — and then refine those insights to implement new billing norms.

Understanding the data’s life cycle is the cherry on top. Some data, like legal matters and invoices for a law firm, are transactional and tied to a specific point in time. On the other hand, aggregated historical data — like the number of matters created or closed in a quarter or the average number of invoices from a given law firm — can be used to make predictions about the future. These two broad categories are themselves tied to different business goals.

VC: Value Creation

Once the data is well understood, creating value — whether it’s internal or external — will inevitably follow. Value is created by understanding how to improve efficiency, productivity and time to market. Let’s say a product is not working effectively and is consuming a lot of resources internally. Data that you've liberated through a dashboard or visualization can illuminate how to fix the workflow, perhaps by enabling the automation of key processes.

To identify the potential value in liberated data, look at processes in which team members execute steps repeatedly or use rule-based decision making. The data you generate from such processes can enable experimentation and hold significant hidden value.

In the legal industry especially, you can optimize spend management — which includes budgeting, selecting and negotiating with outside counsel and processing invoices — with proper data analysis and complemented by artificial intelligence. Improving spend management and other efficiencies has an external benefit, too, as you can pass cost savings along to customers. Data can also help customers make better decisions about the products and services they need. For example, they can use data to decide which law firm is best-suited for a particular matter.

Using data and metrics effectively often seems simple in hindsight, but it can be daunting and difficult without a plan. No matter how much experience I have working with data, I don’t stray far from this framework because it ensures information gathering is tied to business outcomes.

There’s no point in collecting data without any purpose, data that’s not clean or data that just sits somewhere and is never used. The IAVC framework can help you avoid such mistakes — which is important because, done correctly, the art and science of data collection can be extraordinarily powerful for any business.

Visit the website to read more from the Forbes Technology Council.

About The Author

Raj Sethuraman

Raj Sethuraman is focused on delivering enterprise platforms that leverage technology, such as artificial intelligence, machine learning, data analytics and more, to improve productivity and business growth for global legal and claims teams. As CTO, Raj oversees all product engineering and software development activities of the 225+ global technology team at Wolters Kluwer's ELM Solutions.

Raj has substantial leadership experience in managing software development teams with global companies, including Intuit Inc., United Health Care Group, Brillo Inc., and Agilent Technologies. Raj is a graduate of Harvard Business School’s Advanced Management Program. He earned his MBA from the University of Southern California, a Master of Science from the SJCE School of Engineering in Mysore, India, and a Bachelor of Science in Electronics and Instrumentation from Annamalai University in Chidambaram, India.