How chatbots are revolutionizing the auto industry

Virtual assistants wheel us through the process

Unless you’ve been under a rock this past year, you’ve probably conversed with a chatbot, i.e. algorithms that replicate human conversations. Estimates show that one out of four consumers and nearly half of all millennials interact with a chatbot daily. They’re changing the way we get information and purchase products.

For car shoppers, chatbots are becoming increasingly integral to the experience.

With the help of a bot, customers don’t have to disclose their email at the outset to be guided through an order process; check their credit scores; schedule service appointments, and get their car diagnosed. Chatbots are particularly useful for late nighters or early risers-- those who prefer off-business hours.

These artificial intelligent robots have already increased sales and improved the customer experience by proving more accurate, and actually, yes, more trustworthy than their human counterparts. True they’re not yet Mechas, the advanced humanoids with thoughts and emotions from the movie A.I. Artificial Intelligence. But they’re getting smarter and we’d do well to get used to them in the auto industry as they’re going to be driving the entire cycle of buying, owning, servicing and selling.

AI and ML ensure higher accuracy

While artificial intelligence (AI) and machine learning (ML) techniques have been under development for at least 50 years now, the current technology, though still in its infancy, is enabling higher-than-human accuracy in chatbots.    

“For the most part, artificial intelligence (AI)/ machine learning (ML) techniques are not new; neural networks have been around since the sixties. However, natural language processing is an AI capability that is getting more accurate,” said Greg Shaffer, Executive Director of Product Innovation at Edmunds. “Chatbots are more accurate than human interaction in directing customers to areas that can help them.” But the key, he added, is to develop the right tools to synthesize the data. “In the past, we couldn’t come close to problem solving. Now with deeper processing power, we have the tools.”

In other words, thanks to 50 years of Moore’s Law, computing power is so robust it can support an always-on, always-learning, and always-fine-tuning-for-accuracy, AI-powered bot that improves with more usage and more data inputs.  

“Where we think we have an advantage is that we’re focused on one vertical. We can leverage literally 20 years of text and chat data and we’re able to train our bots pretty efficiently so we can easily narrow down the path that users take from the top of the funnel discovery down to the specific make, model and trim,” said David Domm, Senior Director of Product at Cars.com, an automotive digital marketplace. “The challenge for any company is nailing down the NLP (natural language processing component) so we focus quite a bit on the fail rates. That’s the toughest nut to crack,” he added.  

More trust, More efficiency

To trust is to be able to predict what someone may do. Even if someone is a thief, you can “trust” them to steal. Often trust comes with time-- enough of it to determine a predictable pattern. So why is it not odd that humans “trust” machines they’ve never interacted with more than a human? How come studies show that people more easily confide in, obey and take orders from a machine vs. a human?

Some say it’s because humans view machines as objective, after all, they’re logic-deducing algorithms-- like a math equation.

And if ever there was an industry-- besides politics, of course--that needed help in the trust department, it is the auto business.  Decades of car salespersons being pushy “sellers” vs experts looking out for the best interest of a buyer, has created a stereotype. Car salesmen jokes abound. In fact, one Gallup poll found that car salespeople rank right up there with politicians as your least trustworthy humans. Little wonder, then, that it is for chatbots to provide that much needed assurance of trust and transparency.

“There are a lot of shoppers that don’t like to talk to car salespeople, especially if most of this time is spent at the dealership,” said Andy Lapin, Chief Architect at Kelley Blue Book, a division of Cox Automotive.  “People like to do things efficiently. Would consumers rather talk to a chatbot through their car shopping experience rather than to salespeople, who to some buyers don’t come with the best reputation? We want to improve and even remove this negative perception to make the purchasing process easier, faster and more enjoyable for the consumer.”

Another reason chatbots may be preferred is that consumers, particularly millennials, want intelligent engagement and they’re skeptical of sharing information on a form for fear of being targeted by spammers.

“Consumers don’t want to drop their name and personal information into a black hole not knowing what’s going to come back,” said Cars.com’s Domm. “They don’t want to be spammed.”

Importantly, chatbots are able to collect, archive, organize and even share data. “AI can do a lot of the grunt work by collating information on forms rather than forcing consumers to fill out the same information over and over again,” said KBB’s Lapin.

As Cars.com’s Domm observed: Bots can gather valuable consumer data (vehicle attributes,  purchase budget and timeline, even creditworthiness) and arm salespeople with pre-filled data that reduces human involvement, and saves time going back and forth with consumers. Bots are “facilitating the progression from ‘traditional lead gen’ to more qualified ‘curated connections,’” said Domm.

Moreover, chatbots reduce wait times. “A number one consumer complaint  is that when they finally arrive on a dealer lot, the average wait time is four hours,” disclosed Domm. “If we can automate that process where at the end of the day we’ve helped them to make a better choice on their vehicle; helped them choose the right dealership, and when they get to the dealership, ensure that everything is done for them, that’s a lot of friction we’ve removed, convenience we’ve added and that’s a huge win for us.”

Greater opportunity beyond the sale

If bots can help create transparency and efficiency during the buying and selling process, imagine the efficiencies they can create through the entire ownership.

“I think there’s a huge opportunity to cater to the dealership,” said Cars.com’s Domm. “On the fixed operations side, when it comes time for service and repair, dealers are inundated with phone calls. Service bays are overloaded with repair orders, the phone just rings and rings and they can’t efficiently cater to and effectively serve their end user, which is the consumer,” he said. “We think the bot can provide a full-concierge service by automating scheduling, diagnosing and deciphering what’s wrong and connecting the consumer with the right dealer.”

Domm acknowledges that selling a vehicle is only an entry point and that dealerships actually make most of their money from financing, insuring and servicing. Consumers spend roughly $1200 a year on car maintenance over an average of six years.

“Building a relationship that can bear fruit in terms of service and repairs and other items during the ownership cycle is an area that has been overlooked,” he said.

Kelley Blue Book’s Lapin agrees that when a consumer schedules a service appointment, “… they want to know how long it will take and how much it will cost.”  These are simple tasks that can be handled by a Bot, creating for the franchise dealer a much more valuable consumer experience.

Both Lapin and Edmunds’ Shaffer concur that chatbots can increase customer engagement and dealer retention. They believe the automotive industry is poised to develop a new model that touches customers along the delivery chain—consumers, manufacturers, dealers, salespeople, and service technicians. The bot, knowledgeable about our preferences, walks or better yet, talks us through the entire process.

Be where the ‘conversation is happening’

According to Kelley Blue Book’s Lapin, it’s still too early to see an increase or decrease in revenue from chatbot integration. Nonetheless, every product group in his organization is moving toward a conversational interface, with a significant focus on adaptive messaging to cater to individual consumer responses.

“This is more than just putting a toe in the water,” he said. “I think really every product group we have is talking to us about using UIs that make their product easier to use with a conversational versus a web interface.”

Similarly for Shaffer, it’s premature to be expecting monetization benefits despite Edmunds already having 8,000 dealers using their messaging platform. “Right now people are feeling the growing pains of chatbot technology. But we need to be wherever the conversation is happening and to evaluate how many car shoppers are having conversations on new channels. We are currently working with platform providers like Facebook. Apple hasn’t opened up their platform, but those conversations are beginning to happen.”

Editor's note: Steve Loeb and Bambi Francisco contributed to this piece.

Image source: xcubelabs