Some people in your dealership might fear artificial intelligence and machine learning tools, but they are already widely in use whether you recognize them or not.

In the last two columns in this space, we took a deep dive into showing how machine learning tools were empowering next generation decision making for the financing teams within dealerships.

We showcased examples of how a credit decisioning platform called Lucy, that was developed by dealers, was being used to get more car deals approved. Lucy is powered by artificial intelligence and pre-qualifies 85 per cent of customers in 16 seconds—regardless of their credit history.

One challenge though, is that many dealership personnel don’t fully understand artificial intelligence and machine learning and this leads to fears and apprehensions about trying out the technology.

In this column, I’d like to take a step back and help demystify and clear up some confusion and bust some myths about concepts like machine learning and artificial intelligence so that dealers can make informed choices about embracing tools that can help fuel productivity and innovation.

Myth: Only large organizations can afford machine learning and artificial intelligence.

It’s true that many big companies use artificial intelligence right now, and have worked on it for many years. But I think people would be really surprised to see how many small companies use AI and machine learning. It’s just the way of the future. Many companies are trying to incorporate AI into their business. It reduces inefficiencies with things that are already being done by human beings. I think everybody’s gonna be doing some kind of AI or machine learning in the future— not just big companies. Mom and Pa shops can, and will use it.

Rather than replace people, a lot of companies use these tools in combination with their people. How can the technology be used to make this person’s job a lot more effective, more accurate and a lot easier to do?

Myth: AI and Machine Learning will replace my people.

That’s a huge misconception. I think there will always be some people that will feel that way. But I think the ones who are up to date with technology will see this as a huge benefit for them.

Rather than replace people, a lot of companies use these tools in combination with their people. How can the technology be used to make this person’s job a lot more effective, more accurate and a lot easier to do? Many systems are not fully autonomous and still need some degree of human interaction. If you have a system that can help you be more effective in your job and does, let’s say 80 per cent of the work for you, why would you not embrace that? So I would just say, don’t be afraid of the future. Don’t be afraid of AI. It’s not meant to replace you. It’s meant to support you and help you be more, more effective in your job.

Myth: Artificial Intelligence and Machine Learning are the same thing.

Artificial Intelligence systems are built to solve complex problems, in some cases by thinking like a human. Machine learning is a subset of that system. Machine learning allows machines to learn from data without being programmed. As data is entered, it starts to learn those algorithms, and gets better, and more accurate in its analysis and predictions as it gets more data to assess.

Myth: No machine can replace the “gut” instinct of my experienced F&I Manager

When it comes to Lucy, our Credit Decisioning Platform for car dealerships, she has the “gut instinct” of hundreds of F&I Managers, who have done this for 20 years! Even your most experienced F&I Manager won’t have seen as many clients as Lucy has. It’s not that it’s a competition, it’s just the volume of data that provides insights into which deals get approved and that’s why Lucy can be a very sharp ally within your dealership. Lucy can work with your experienced—and less experienced finance teams—to help them be more efficient, close more deals and better understand the complex world of credit approvals, particularly in the non-prime market.

Myth: Your AI system will reach the same conclusion as my Finance team so why bother.

When the Founder of DecisioningIT wanted to build this product for his dealerships, he took five of his F&I managers and asked them to work on a credit application file. All five came back with a completely different analysis and recommendation. A system using machine learning can reduce human error because it focuses on actual stats, information, lending criteria, all in real time, and makes decisions based on a massive amount of information in the system.

Myth: If I just keep doing things as I’ve always done, I’ll be fine.

A lot of dealerships might not recognize how much Artificial Intelligence is already baked into their own personal lives and their dealerships too. For example, companies that mine CRM data and trigger alerts to touch base with customers after a certain time or service interval, or make predictions based on their expected mileage and vehicle usage. Or the voice activated assistants streaming their music or powering their search. The reality is that many dealerships are embracing a host of advanced technologies that help them run their businesses, and help provide a better customer experience. Machine learning tools that help make faster, more accurate decisions are just part of that.

The post Busting myths about artificial intelligence appeared first on Canadian Auto Dealer.

A look behind the scenes helps shed light on why lenders reject many good clients, and what dealers can do about it.

Some finance managers at dealerships are mystified when seemingly good financing applications are rejected, but in the hustle and bustle of trying to get the next deal closed, they often just move on and don’t give it much thought. 

 But there are lots of hidden opportunities for dealers in working rejected applications, and perhaps more importantly understanding why certain deals get rejected in the first place. 

 When dealers get an auto decline from a lender, they often get an “answer” that comes back in less than a minute that says: “Consumer does not fit the lender criteria.”

 With so little information to go on, it’s understandable why finance managers just move on. They often just send the deal to another lender, and if rejected, then to another lender, and eventually they just close the file.

 Behind the scenes, lenders have their own internal scoring system. They have a view of the client’s past history and their own historical data not available to anyone else. So, in some cases, no one from outside their four walls has an idea of the actual score. For example, many banks are “allergic” to lending money to any former client who missed a payment with them as a lender. 

 But the vast majority of the reasons why lenders would decline a file are actually in the hands of dealers—they have the data. Lenders base 90 per cent of their decisions on the data provided by the dealership and on credit bureaus. Dealers are just not reading it to figure out if they’re gonna be declined or approved prior to submitting.

There are lots of hidden opportunities for dealers in working rejected applications, and perhaps more importantly understanding why certain deals get rejected in the first place.

Dealers could spot a lot of issues by pulling the client’s credit bureau. For a full picture they should examine both Equifax and TransUnion reports.

They should also remember that when it comes to credit applications 60-65 per cent of all transactions are not even looked at by a human, and that means auto approvals and auto declines can contain mistakes. 

In one example, we had a client who had applied for financing while under bankruptcy. When we pulled the file, we could see a vehicle repossession occurring nine months after the bankruptcy, but in fact it was included in the bankruptcy. This triggered a rejection, but it was a case of bad timing—not another issue on the credit file. 

In this case, the issue was flagged by Lucy, our automated system that uses predictive AI to help dealerships navigate prime and non-prime finance applications. Lucy flagged it and identified there was a Ford Motor repo, right after the bankruptcy that didn’t add up on the timeline. Also, based on a historical timeline, the client used to always pay on time, so something wasn’t adding up. The dealership just needed to get the necessary documents from the bankruptcy trustee, and then resubmit to the lender to get approval. When presented with the information, the lender reversed their decision.

In this case, a skilled F&I manager might have been able to spot the problem, but the reality is that with the volume of transactions and submissions, and the variability in the criteria for each specific lender, prime and non-prime, an automated artificial intelligence platform like Lucy can act as a trusted companion to your business office.

Lucy does a good job at talking to automated financial systems, because, well, it takes one to know one. When she assesses a client file, she will return the lenders that match the lender programs and who will “green light” the deal, the “red lights,” which won’t approve it, and the “yellow lights” that represent lenders in the gray zone. 

Because Lucy’s algorithms consider each lender’s financing requirements, she will provide insights to the F&I manager about which issues are causing the uncertainty. In some cases Lucy will recommend reducing the vehicle value or adding cash into the transaction to minimize the risk so the lender will say yes. 

In effect, Lucy knows her lender friends well, and knows how to navigate yes, no, and maybe. She can also do this in a matter of seconds, and has no emotional attachment to the client’s hopes of acquiring a particular vehicle. 

So, while she will provide the information needed to present a credit application in the best possible light, she won’t lose any sleep if a lender rejects or accepts a submission. She is, after all, a machine who learns!

The post Why do lenders reject so many credit applications? appeared first on Canadian Auto Dealer.

Time to raise the red flag about what’s coming up, and how to manage the inevitable.

As a general principle, the non-prime market did somewhat well over the course of the pandemic, due to government aid.

The help consumers received during the COVID-19 crisis led to increased purchases for vehicles and even power sports. There was an increase in value for many reasons, and the price for used vehicles actually jumped about one per cent month-over-month—for power sports as well.

With government aid, many consumers were able to save more and spend more on things like more expensive vehicles. Some of these consumers are subprime, and specifically subprime consumers buying vehicles that are more expensive today.

The problem with this, that will inevitably present itself before us in the near future, is that consumers that buy into these more expensive vehicles will have a much higher negative equity when they want to resell or repurchase a new vehicle and trade-in that vehicle.

The problem we have today is not related to consumers getting approved at credit, because the credit approval and the default rates have been great. There is no issue with the markets as it is today.

But every lender knows that what is coming soon are rising (general) inflation rates—and higher interest rates mean consumers that are in non-prime will be the first to be impacted due to having purchased a vehicle over market value.

The whole non-prime market will see this potential issue coming fairly soon, where people will want to get out of those very expensive payments. And when they return those vehicles to a store or to try to sell them, they will be losing quite a lot, because their interest rates are higher and the vehicle price is higher. Therefore, their negative equity will be much higher.

This is a problem we are creating right now—that we created as of around last November, December, and into January. These consumers are buying high with high interest rates, and we are creating an impact that—in a few months from now, or years, or maybe a year or two—will have a lot of people buying vehicles with huge negative equities.

So for now, dealers are happy to sell vehicles during what may otherwise be a difficult period, with the vehicle inventory issue and other pandemic-related challenges. The problem is that we are shoveling negative equities, and we are not going to see such a nice scenario in a few months from now.

There is something coming—there is a wave coming, or it is already here. And what I can say is that I’m certain about that.

If you become more selective on prime,  then the non-prime segment will just grow by itself.

There will be more non-prime in the coming months and years because the interest rates will be higher. Therefore, consumers will be stuck with payments and it will be complicated and challenging for dealers that are not decoding non-prime, to cater to that segment. And the reason being is that the files will get more complex.

Dealers in the market will be choosing their consumers much more wisely, because they will not be able to afford issues with clients delaying payments since their interest rate will be high. They will become more narrow in terms of the actual prime consumers they want, which will raise the non-prime segment.

That’s the effect. If you become more selective on prime, then the non-prime segment will just grow by itself. There will be more non-prime consumers in the future, and with higher negative equities.

The challenge with that is that dealers need to secure the right inventory and the right employees to deal with the more complicated files that will result from this situation.

It’s not a simple prime approval—it will require expertise. Some dealers have that expertise, but I expect artificial intelligence technology to become a key or important player in this overall situation.

These types of tools, like predictive AI, are being adopted by some dealers to help with loan-matching and customer credit management, so that more consumers are approved, and dealers don’t leave money on the table.

And this year, dealers must investigate and implement this new technology, as well as prepare their non-prime processes for the coming months. Unprepared dealers will struggle with non-prime, they will pay a high price for experienced non-prime business managers, and their sales volumes will suffer as a result.

The post Decoding the subprime lending market for 2022 appeared first on Canadian Auto Dealer.