How Artificial Intelligence, Machine Learning and Predictive Analytics Sell More Cars

Imagine this scenario: it’s been a slow month in the dealership. The sales guys are playing Angry Birds or fiddling with their fantasy football (though they are supposed to be combing through cold leads). Nobody’s walked in the door of the dealership for hours. You have a long, long day ahead of you. Sound familiar?

Here’s another scenario. You arrive at the dealership in the morning, and your customer relationship management (CRM) solution informs you of the top 10 activities you can do that are most likely to lead to a sale. How does it know that? It’s fortified (like your breakfast cereal) with machine learning and predictive analytics and built on a platform of artificial intelligence, or AI.

You already use artificial intelligence in your dealership. Don’t think about robots or science-fiction scenarios, think about something a little closer to home, like Siri on your iPhone or Alexa on your Android tablet. The truth is, you do use AI in your dealership operations, probably every day. It’s time to start harnessing the technology to help you sell more cars.

What Is AI?

“Artificial intelligence” is a broad term, and it’s generally defined as a computer system that can perform tasks that would normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages. It’s often used interchangeably with the term “machine learning,” but there are differences.

Machines That Can Learn

“Machine learning” is a type of AI that involves computer systems that can learn and adapt based on their operating experience without any direction from their programmers. It’s translation or dictation software that adapts to the way you speak, or a news aggregator that “discovers” your favorite sports team and actively seeks out information about it for you. It’s when Siri puts search results for Thai food on top of your local restaurant listings because she has “learned” that you have a passion for Pad Krapow Moo Saap.

Predictive Analytics

Can you use what happened in the past to predict the future? There’s a body of science that says you can. Many software solutions today have built-in predictive analytics, a science that uses machine learning to analyze data from previous events to predict what will happen in the future in a way that is far too complex for the human brain to do. Predictive analytics can find patterns in data that human brains simply aren’t fast or smart enough to recognize.

In your dealership, predictive analytics could be used to comb through previous sales information to determine that individuals who live in higher income ZIP codes in family homes are more likely to purchase a new vehicle on a weekend in mid-May in anticipation of family summer road trips, for example. Or, it might discover that people who lease cars for both business and personal use are most likely to seek an upgrade in the early fall when trade show season begins, and new products and marketing campaigns are unleashed.

Essentially, software solutions with predictive analytics built in can examine customers’ demographics and past behavior — past purchases, product views, or information such as location, age and gender – to predict what they’re likely to buy in the future, when, and determine what the ideal selling approach to succeed with them is.

As Usual, It’s All About the CRM

Chances are good your customer relationship management (CRM) contains more customer information than any other solution you use. Since AI and predictive analytics need good information to work, it’s vital that 1) Your CRM is error-free and regularly updated; and 2) That your CRM is properly integrated with other software functions such as BDC and marketing. The more information your AI has, the better it can identify the “low-hanging fruit” that will lead to a sale by the weekend.

“CRM analytics software provides a single, unified platform through which all authorized team members can access specific information on individual customers, in order to provide them with the personalized service,” according to Salesforce. “And, given that 30 percent of marketers say having disparate data sources is the main reason they cannot glean useful insights from a customer, a system that collects all available data into a single location is perhaps the most important step towards creating a working customer relationship.”

Predictive Analytics Stems Customer Turnover

Why do you lose customers? Why did that couple who seemed so hot on buying a car last week suddenly drop off your radar? Unless customers tell you directly, it can be hard to know. Many things can contribute to customer attrition, including not following up properly on leads, not approaching the customer in the right way, not offering relevant product suggestions, or gaps in your customer support channels, according to Ray Mendoza, Principal Engineer at SoftClouds.

“Using predictive analytics, we can determine what causes customers to make the decision to leave your brand, your product space, and ultimately cast their economic vote elsewhere,” he wrote. “Some factors are weak in determining attrition, while some are strong indicators. Predictive analytics allows a model to be built to understand what causes this problem so that businesses can improve customer retention and provide better service to their customers.”

Understand, Motivate and Reward Your Sales Staff Better

On the cutting edge of CRM, there are even apps that can measure your sales team’s “mood” and use it as a factor in setting recommendations for activities most likely to lead to a sale. Other hard-to-gauge factors that affect selling include salespeople’s communications style and even work ethic. It’s a way to recognize that not all the factors in a sale (or lack of a sale) are based on hard numbers.

Using a CRM app that measures these “softer” factors, you can better understand how to motivate individual sales staff members (since their motivations may be different) and measure their performance with better tools. (For example, is a salesperson who sold 10 cars with very hot leads working at the same level as a salesperson who managed to sell 10 cars with very cold leads? Probably not. So, are you really measuring their performance accurately and compensating them fairly?)

While it’s true that intuition and experience play roles in successful selling, it’s clear that humans simply don’t have the mental processing power to evaluate all relevant factors and fill in all the gaps. By allowing “smart” solutions to do some thinking for you, you can open new venues and create more opportunities to sell vehicles.