Harnessing Big Data to Power Real Time Insights in Automotive Retail

By: Ronojoy Bhaumik, Executive Manager – Product and Technology, CDK Global (India)

Terms like Big Data have been creating quite a buzz for the past few years, just like Moore`s law for processing power. What was once considered big is now small.

With the internet and 5G networks on the horizon, everything is mostly about sending and receiving data in galactic proportions – much like autonomous vehicles receiving traffic data to compute routes seamlessly.

What is Big Data?

“Every day, we create 2.5 quintillion bytes of data — so much that 90% of the data in the world today has been created in the last two years alone. This data comes from everywhere - sensors used to gather climate information, posts to social media sites, digital pictures, videos, purchase transaction records, and cell phone GPS signals to name a few. This data is big data.” (Source: IBM)

We at CDK Global, are in the business of automotive retail – we help consumers buy and dealers sell cars. Our software is responsible for selling one in two cars in North America and that’s a claim very few in our industry can make.

The answer lies in the amount of data we can leverage and power businesses from across the auto retail industry. We start in the digital world (websites and click stream data, digital advertising platforms, social media engineering) and move on to the actual retail cycle of – sales, finance and insurance, parts and inventory, and service and after sales.

As consumers go through this journey, we collect data and curate it, which then gets populated on our global data lake.

To give you an idea, this is what a typical monthly volume table looks like.

While these are impressive numbers, it is dwarfed by examples of social media sites like Twitter or Facebook which generate this amount of web traffic volumes within a day or even hours. What we can gather from here is – it’s not how big your data is, but what you can do with it. The question that needs to be addressed here is whether we can generate insights that can power more sales or help the customer make better decisions. From a technical perspective, latest technologies such as Hadoop, Apache Spark, Lambda Architecture, etc. are used within our BI teams. But having said that – what needs to be reinforced is that –

• Technology is a means to an end, not the end in itself
• Customers do not care much about the technologies that are being used
• Customers stay or go depending on how well their business problems have been handled

Now, let me run you through some scenarios of real-time insights that power businesses.

Scenario 1:

• Every vehicle that is sold has a Vehicle Identification Number (VIN) and every consumer interested in buying the car will most probably start his/her journey on the internet - go to a website to search, track, and locate the car of his/her choice.
• The ad that directed the consumer (tracked using cookie IDs on browsers), the page on the website on which the consumer spends more time on, and the interest in purchasing a particular car model, whether an SUV or a sedan. As the consumer spends more time on the website, narrowing his preferences becomes easier, and the propensity to buy the car within two days from the dealer located at a particular location is noted.
• This is information that the auto dealer can use if there is a Sales/Finance person who is generating a quote for customer. The dealer can bring up the VIN and the BI will highlight how popular the car is based on the web activity. When negotiating, dealers can identify vehicles that are most likely to be sold within the next 30 days.

This is the Vehicle Interest Score and this is how it is computed.

Scenario 2:

Personalization is the name of the game. Ever noticed how retail sites like Amazon or Flipkart display items that you have previously looked for on the web? It’s analytics at work and this is called as targeted or retargeted advertising.

Similarly, consumer behavior can be tracked using cookie IDs and user attributes from across devices. Real time is used for constant data processing and machine learning with which the content of digital advertisements or WebPages on the fly are made to match with what the user needs. This kind of personalization makes the difference of converting a webpage visit from a regular browser to actual sales in dollars and cents.

The most difficult part of today’s data world is not technology or computational power but governance. The questions range from whether privacy laws are being violated or if data can be monetized. Now, companies are investing in data stewardship and governance.

Nowadays, telematics data and old service records from vehicles are used to give information to both the auto dealer and the consumer if a car is due for service. This makes both the consumer and the seller aware, help protect their vehicle, and increase business.

To conclude, what makes this so exciting is the possibilities that exist to make life even better for customers in the auto retail world, with technology and data science that can power predictive and prescriptive insights with greater accuracy.