Three Easy Steps to Get Ready for AI

(This piece by Scott Nelson originally appeared in Commercial Break Volume 31 - Issue 3, a quarterly publication of the AACFB)

Are you getting tired of hearing about AI yet? Do you find headlines like “Why I can’t live without ChatGPT” or “AI can’t do my job” (I paraphrased both of those titles) esoteric, intimidating, or just plain boring at this point?

Believe it or not, I know how you feel. That may be surprising coming from someone who is developing AI-based applications and frequently writes about how AI will change the way equipment finance operates. The fact is every new technology trend starts with two things: hype and fear. The hype elevates the impact, innovation, and indispensability of the technology to compel clicks, views, and, hopefully, adoption. The fear follows the hype naturally because if you don’t adopt this new, indispensable technology, others who do will put you out of business. Actual Harvard Business Review (HBR) title: “AI Won’t Replace Humans — But Humans With AI Will Replace Humans Without AI.” Fortunately, legions of MBA-toting consultants who are now “experts” in the new technology are available to help. If desperation is the mother of invention, then fear is the mother of business consulting.

So, what is a finance brokerage or equipment finance company leader supposed to do? Should they learn the ins and outs of LLMs (Large Language Models) and how to manage terabytes of unstructured data in Snowflake? Fortunately, the answer is no. The reality of this technology trend is that getting started is pretty easy and the tools are developing so fast that the thoughtful and pragmatic person can avoid early adopter mistakes while still benefiting from the returns.

You have data, start using it.

The first step to getting ready for AI is actually the easiest because all you have to do is start using your data. Finance brokerages are transactional and transactional businesses generate tremendous amounts of data. Every application, every customer conversation, every lender question, and every lender decision generates data. Data that is rich in information about customer preferences, product strengths, vendor strengths, lender preferences, and timelines for changes in the marketplace. Your data is good enough, but you have to start capturing it if you are going to use it.

The right partner is a collaborator.

This leads to the second step which is often intimidating for business leaders because they perceive that it involves understanding the technology at hand – in this case AI. But picking the right technology partner for data and AI is not about how deeply that vendor understands ChatGPT, LLPs, or neural networks, nor is it about how widely they have applied it to everything they do. Rather, the use of data and the application of AI is about how well a service provider collaborates in the digital ecosystem. Unless you are talking to Apple, Microsoft, Amazon, or Google, none of the technology partners in equipment finance have the ability to vertically integrate all of the digital technologies and applications required to be a one-stop-shop. So, first look for partners who know and understand equipment finance and finance broker workflows. AI can be applied to most business practices, but hasty AI training can lead to bias amplification and poor outcomes. So, you want someone who knows the business. By the way, someone who knows the finance broker business will know that it is a business built on the management of operating expense, not ROI from capital expense. Their pricing should reflect this insight.

Finally, look for vendors who know and clearly state that that they are not best-in-class for each step in the workflow or software stack. They will have and be able to describe partnerships for key parts of the workflow and why they chose those partners. They should have designed their business for collaboration so that they can collaborate with both you and your other preferred providers to enhance the uniqueness of your business and your strengths to make you more competitive.

If you hear a vendor say, “We can solve all your AI problems,” they can’t, and you can move on.

Self-Assess: How can we do things better?

After you find your vendor partner, the last step is easier for some than others. It’s easiest for the curious and those who constantly ask, “How can we do things better?” Data, no matter how much you have of it, only has value to those who have questions. What do you wish you knew about your customers? What do you need to know to win consistently in a new market expansion? Did you pick the right vendor to help you capture share in that geography? Your collaborative and equally curious vendor will know how to best capture the data streams needed to answer these questions, whether internally generated, generated by a funding partner, or externally by market sources like FICO and PayNet. This data, when properly analyzed and presented either directly or with AI prediction will show you the trendlines, ratios, relationships, and documented outcomes you need to perform faster and better.

Eric Reis introduced build-measure-learn as an innovation and technology adoption methodology in his book “Lean Startup.”

Expanding your “digital field of view” in this way is the key to capturing and aggregating the right data. That will, in turn, help build the business intelligence and AI predictors that enable the exploration and automation of outcomes. Learning from your data is the objective and AI, when properly implemented, is very good at learning because it is fast and stubborn. A good learning system will practice Build-Measure-Learn until better results are consistent.

AI is coming and it will change how finance workflows – from start to finish – are implemented. But this should not be intimidating or cause a loss of sleep. Three easy steps will get you started and deliver immediate returns. Get started now and be ready to add AI when the returns are clear so that you do not become a validation data point for an HBR article.