The old adage– “How do you eat an elephant? One bite at a time” might seem out of context today - clearly the hunter-gatherers of the Middle Ages were actually trying to eat the elephant. But the modern-day version, attributed to Bishop Desmond Tutu, conveyed a philosophy about how to change culture and socio-economic systems.

I introduce the adage here – with a digital “byte” twist - because the scope and scale of digital transformation in equipment finance can often feel elephant-like. Data migrations, workflow translations, and employee training requirements present the possibility of months or even years of effort in the eyes of CIOs and CEOs - many of whom have wounds from previously failed attempts to “deploy better systems” and experienced the challenges of such large-scale change initiatives.

Two recent developments in equipment finance present the ‘eating digital elephant’ experience:

  1. First, embracing the opportunity and capability of technology today is existential. Competitors who have already adopted cloud architectures with enhanced business intelligence and are able to effectively deploy machine learning and AI to improve the efficacy, efficiency, and the speed of their businesses. They are learning faster and using the plethora of data they have to win in the marketplace.
  2. Second, the sharpness and ease-of-use of today’s data tool set removes excuses. Never has it been so easy to leverage data in new workflow applications and use it to make strategic business decisions.

If your digital transformation looks like an elephant, how would you start? Let’s consider what a five-course digital elephant meal might look like.

The First Byte: Your data is good enough.

The First Byte is about people and attitude, and the organization’s culture. The challenge here is that many business leaders don’t have the confidence and fortitude to even start the meal or take the first bite. They worry about not having enough data to solve their problems, but the reality is that leveraging the data they do have will provide enough insights to get started and, more importantly, will immediately guide the organization in how to generate more and better data for improving operations going forward.

Every equipment finance company we’ve worked with has immediately experienced a cultural change in regard to using data. This is because the teams see leadership’s investment in data and see how data can reveal insights into challenges that previously may have been felt, but not verified. Salespeople, for example, are naturally competitive and results or “scoreboard” driven. Once sales teams see what is happening at their level, without exception they have asked for access to more business intelligence dashboards to help themselves improve their performance. They are often the heaviest users of the business intelligence dashboards and always the leading group looking for ways to do their jobs better.

Data-based business intelligence is naturally addictive and can help teams evaluate and improve their performance. Your data is good enough, so get started.

The Appetizer: Outcomes that matter.

The next step in getting started is getting oriented. Any transformation of a business is often motivated by the goal of improving business performance. AI can deliver predictions of outcomes and are usually driven by the plethora of data inputs that will eventually influence those outcomes. This step is focused on orienting the project by sitting down with your team and discussing what matters most to the business and where the business needs to become more efficient. Identifying the outcomes that define success, how you define being effective within each business function as well as what success looks like at the mission level of the business.

There are four areas where outcomes can be measured: origination, collections, syndication and funding and portfolio management.

Origination
Look-to-Book

Is this deal worth our time? How do we help brokers help us?

Delinquency and Tier (Rate)

Does this deal fit our strategy? Automate underwriting.

Delinquency

Is this deal worth the risk?

Collections
Pay Anyway

Do we need to call this delinquent account?

Delinquency with Updates

Is this deal weakening? Should we act now?

Cash Generation

Is the business performing to make payments?

Syndication and Funding
Syndication

Do we hold or sell this paper?

Lender

Which funding partners will want this deal?

Lender and Tier(Rate)

Can we add a premium to this deal, or must we discount?

Portfolio Management
Term with End State

Is this deal likely to renew or return?

Syndication with Lender-Match Updates

Has this deal aged properly for sale?

Residual Monitoring

Are residual values following projections?

Predicting outcomes helps solve operational problems faster.

Some outcomes are numeric like interest rates, and others are defined by a set of outcomes. For example, delinquency can be 30-60-90-Loss, syndication is Hold-the-Paper or Sell, and Lender Partner can be list of fifteen banks. Productivity for an organization is defined by the measures of its outcomes so identifying those measures is first. Then your next step is identifying every parameter that you believe will be important to those outcomes. Think about the data you have and the data you don’t have – if you don’t have it, you can usually partner or buy the data you need. Again, your team can help identify the sources for the input data relevant to your business outcomes and you should be sure to leverage the experience and knowledge of each function in the organization: origination, collections, portfolio management, and funding. Map your processes, identify the data that will get generated and identify the missing data you might need.

Bread and Butter: All your data in one place.

Now that you know the data that matters, you need to get all of it – historical and present -- into one place so that you can use it and improve your business by using it. Equipment finance workflows involve a multitude of data sources generated in a variety of ways and at a variety of rates. The business receives customer applications, credit bureau reports, vender pricing and availability, bank lending rates, etc. -- and you must find a way to get that data into a common data structure that enables you to use and reuse it. Business intelligence applications use data to provide visibility and analysis of where you have been and how you are performing against plan.

As you begin to use data analysis, you will validate the data against outcomes and identify new needs or data that needs more diligence where there are gaps. If you don’t have an IT team that works with modern data tools like Azure, Power BI, and Open ML, find a partner who does. Once you have the data organized and data streams connected for ongoing capture and use, you are ready to move on to the main course of the digital elephant – learning.

The Main Entree: Build-Measure-Learn

Eric Reis introduced build-measure-learn as an innovation and technology adoption methodology in his book “Lean Startup.” This sequence is fundamentally a learning process and, as such, is also a methodology for continuous improvement in a digital enterprise. Machine learning models are built on the historical data of past performance in order to predict operational outcomes for future engagements. AI automates learning from variations between the predictions and actual results while integrating those learnings along the way.

Be on the lookout for new measures or new data that will answer hard questions or can guide action. Pay close attention to the rate at which data is captured and the precision with which it is measured. Do you need to measure payment delinquency “Yes or No?” or do you want to know “What are the max days of delinquency?” The latter allows for many more customer engagements without creating additional losses.

Pay particular attention to “outliers.” Outliers often identify inputs either not anticipated or outside normal policies, e.g., “knows the President of the bank” or “I think restaurants are going to come storming back from the lockdowns.” Sometimes outliers catch unexpected behavior and other times they can identify new strategies.

Dessert: More Effective and Efficient

At this point, the hard work is behind you and the organization is naturally gathering and using data. You have an operating system of business intelligence built on the continuous analysis of data from enterprise workflow and results. When you add in machine learning and AI predictions you will enhance the productivity of your team by automating simple decisions and improving the hard ones with more focus from your team thereby increasing the pace of your business without increasing the risk.

AI technology serves the dessert of this 5-course meal and a special treat you will never want to pass up. It helps an organization be more effective in its mission through better application of human resources while improving efficiency with technology automation. More effective and efficient means more productivity. And increased productivity drives growth.

When you sit down at the table, the elephant (digital transformation) might seem way too much to eat in one sitting. Trust yourself and your data – your eyes aren’t too big for your stomach. If you engage in this five-course meal following the steps above, you will find that eating the elephant of digital transformation is much less effort than you first thought.

To learn more about how AI can make your organization more efficient and effective, visit Tamarack.ai.

 
Written by

Scott Nelson

President & Chief Technology Officer, Board Member

Scott Nelson is the president and chief technology officer of Tamarack Technology. He has more than 30 years of strategic technology development, deployment and design thinking experience working with both entrepreneurs and Fortune 500 companies. Nelson is a sought-after speaker and contributor on topics related to IoT and digital health. His involvement in technology in the local and national technology community reflects an ongoing and outstanding commitment to technology development and innovation.

  Connect

« Back
Get Started
  • Should be Empty: