WOLFSBURG, Germany -- Simon Thomas, Volkswagen Group's global marketing chief, says VW must continue to improve the way it launches vehicles and uses the massive amount of data it gathers.
The 53-year-old British executive moved into his job in September after two years as VW Group's U.K. boss and 20 years with Nissan.
He spoke to Automotive News Europe Managing Editor Douglas A. Bolduc at VW headquarters here.
Q: Does an automaker know within a few months of a launch whether a car will be a success?
A: You get a good idea from the customer order intake and the Internet reaction. So yes, you probably have a reasonably good feel within a few months, but I wouldn't say it's foolproof. Cars do continue to surprise and overperform.
The main challenge is to ensure that the business KPIs [key performance indicators] are set early enough for the planning and activation to be linked to those KPIs. We have a robust approach to setting those KPIs and it's important for us to create them early and also cross-functionally.
What are some ways to make a launch more successful?
You know before the launch what competitors are being launched and when, so you never adopt the mind-set that you are launching into a vacuum. You're always looking to your product and thinking, "OK, after six months competitor A is launching their new model or their face-lift," so you try and plan your events and communications and your media timing to take some of that limelight away.
The other key point is that a launch is not just an event at the start of sales. A successful launch should address the whole life cycle and be extended beyond the full introduction. Later, product technologies or range introductions can be used to support the long-term positioning of the product. The opposite of this we call "launch and leave" and that's exactly what we try to avoid.
You and your team recently spent time at the Massachusetts Institute of Technology. What did you learn?
We were looking at big data and how we can use new methods and new technology to create a deeper insight across our whole business. The modern car with its on-board diagnostic device produces an enormous amount of data: telematics, engine, performance, etc. We were exploring ways to harness that data to enhance the consumer experience. Some of the ideas we worked with were completely outside the traditional thinking.
As a natural output from our business, we also have an enormous amount of data from customers and prospective customers. Some of this data can be so big and so dense that it is too difficult to analyze. It can be like trying to take a sip of water from a raging fire hydrant.
It was also really interesting to see what industries other than automotive are doing in this area.
How have you used big data so far?
We ran a number of big data-related marketing programs last year. It was just a start, but we have already proved some methods to be highly effective and we will expand and reuse these. One of these used data analytics to find behavioral clues from the way a user of our Web site is navigating. The messages delivered to that specific Web user depend upon how he or she hesitates. We have proved that model works and that we can address a message to a prospect based upon their degree of decision hesitancy. That's about as targeted as you can get, I think.
We've also looked at the whole adult population of one particular country. We came up with some algorithms to determine likely buying considerations and therefore what messages would be most effective.
How else can big data help an automaker?
There are many areas of the whole automotive value chain beyond that of targeting customers.
One is looking at order banks vs. future production needs to support lean inventory objectives, or creating a more detailed market trend and sales forecast to facilitate a more agile market response.
There is so much data in the business of an OEM, that the boundaries of how to use it are only limited by one's imagination. The key point about big data is to think of what questions you want to have answered, and not think about what big data you have.