Haley Toyota Certified Sales Center
What IBM Watson Can Teach US About Big Data
You may remember the scene in the picture from 2011. IBM developed a super computer named Watson who appeared on the game show Jeopardy. He was in a match against two returning Jeopardy champions, Ken Jennings and Brad Rutter. Watson was uploaded with the entire contents of Wikipedia, dictionaries, encylopedias and other data stored on a 4 terrabyte hard drive. During the game when the host, Alex Trebek, provided the clue, Watson parsed this massive amount of data to come up with the correct answer. All of this takes place in a matter of seconds. So, how did Watson the computer do against the two returning champions? He smoked them. Why? Because he was making his decisions based on a data set, not a feeling set.
I believe that we can learn something from Watson's decision making process. Remember, Watson is a computer and does not have the "gut" feelings that we humans do. In my world of being a used car dealer I have to make decisions each and every day on what types of vehicles to purchase for my store. I am a vAuto dealer and use their Provision tool to help me source my inventory. With all of the data that Provision feeds me, I sometimes let my "old car guy" thinking get in the way of trusting the data. Sometimes when shopping for inventory I see a vehicle where my old car thinking disagrees with the data that the tool provides. For instance it may be a high market days supply vehicle with limited sales history in my market area. The data tells me not to buy that car. My old car guy thinking then shifts into high gear and says " I know they are telling me this is not a good car for my store, but if I can buy it cheap enough I can get it sold".
Watson would not have made that decion. Whenever I have an aged unit in stock, I check the merchandsing, pictures, descriptions, and price to figure out why it has not sold. In most cases I find that these units have 25+ fabulous pictures, have descriptions that Ernest Hemingway would be proud to have written, and seem to be priced right. The reason they are not sold is they are units I never would have purchased in the first place if I trusted the data more than my gut. They are high market days supply vehicles that were not a good fit for my store. My "old car guy" thinking thought that I could outsmart the computer. That will never happen.
I think that we can all learn something from Watson's success. We are in the age of big data in our industry and we have some very powerful technologies available to help us do our jobs more efficiently. I used the example of an inventory tool but the same can be said about the massive amount of data that is in our CRM's, and DMS. Are there reports that we could be paying closer attention to that could help find more deals? More service customers that have defected to the independants? You get the idea. If we are truly in the age of big data and how we use it can determine our success, then lets be more like Watson. And less like the old car guy. We have to learn to trust the data. I sure did. 
5 Comments
Eric Miltsch
DealerTeamwork LLC
Bill, great perspective. Curious to know how long it took you to finally let go of that gut feeling and trust the data your tools were providing you? I know you're very active in vAuto - do have to still fight those urges?
Bill Simmons
Haley Toyota Certified Sales Center
Great question Eric. I have to admit that it took awhile. But I am absolutely committed to staying true to the data.Once I learned that the mistakes in my inventory were my fault, and not that of the software, I was converted. .
Anne Fleming
Women-Drivers.com LLC
Bill - really a fabulous and fascinating article! Bravo! I am wondering how we reconcile the "Blink" (personal gut POV) vs. the Watson (cold, hard facts) in these matters? Does knowing that big data makes a difference actually make a difference when it comes to ego, behavior and action? Of course, the question is rhetorical and different for each person. From consumer reviews, our company provides data with KPI metrics to dealers with clear actions to take that will improve sales to women and all customers. Yet, when it comes down to the "right answer or way to do something", frankly, our humanness sometimes just gets in the way.
Dennis Galbraith
Dealer e Process
Great article Bill. I always love hearing from you with your fantastic humbly-experienced perspective. Anne, you bring up a great point. In practice, I think the book Blink was a setback for data driven decisions, but that was not at all the point of the book. We all carry around a large number of core values, predefined if-then decisions that are based on an objective view of the world. We get in trouble when we reject those and start to remake the decision with a bunch of rationalization. Should I buy that lady a drink and flirt with her even though I'm married? I already have that one worked out, but if I think long enough on it I might be able to rationalize that doing so would be perfectly innocent and would certainly not lead to any bad behavior. (Note: Mrs. Galbraith appreciates that I'm logic based and don't do that.) It's the same for Bill buying cars. He knows he is not going to pay 85% price to market for a car with 120 market days supply. It's a ready-made decision for any vehicle in that situation, based on data. However, if he think about it long enough, his emotions have the opportunity of overrunning his logic. When we have the data, we don't need a lot of time to make logical decisions. We can do it in a blink. We get in trouble when we give our emotions enough time to drag us away from the logic.
Bill Simmons
Haley Toyota Certified Sales Center
Anne and Dennis, thank you. I'm glad you enjoyed the article. Learning to trust the data and keep the ego out of the decision is the big learning curve. But once we do that the benefits are wonderful. I used the example in the article of a car that I should not have purchased. The same holds true for vehicles that the data suggests that I buy that I have never purchased before. It is a real leap of faith to purchase those units but when they sell within the first 30 days it confirms for me that the data is more trustworthy than my gut feeling. When I consider how many more vehicles I could have sold over time making better stocking decisions based on market data, it is mind blowing.