Data scientists represent a fairly new professional discipline, but they are already in the front rank of technologists being recruited to analyse the large volumes of data that automotive retailers will want to scrutinise as emerging omnichannel sales models become the norm, and digitally-informed marketing strategies take over.
IBM says that the data scientist represents an evolution from the traditional business or data analyst role; but these new hires differ from the IT crowd in that their role requires them to serve as the direct interfaces between a wide range of data sets and the showroom sales agents. They need to help turn streams of data into actionable insight and intelligence, sometimes in real-time.
Data-delving specialists are already widely among us, with vacancies for data scientists advertised in business and technology magazines and on jobsites. Data analytics techniques – the data scientist's stock-in-trade – are the subject of multiple conferences and seminars this year. Right now most of the vacancies are in well-healed sectors such as financial services and luxury retail, but the automotive market is set to join these early adopters.
According to Nick Gill, a senior VP at Capgemini and Chairman of its Automotive Council, dealerships should be getting ready for their own version of the 'data deluge' that's been customary in other retail sectors for some time now. There is a growing expectation that massive ‘Big Data’ sets may prove key to developing reliable and consistent new and used car sales opportunities in the 2020s, and beyond. Such opportunities will be achieved by scrutinising the data to profile car buyers, and predictively reveal characteristics that will inform courses of action that dealers can leverage to win loyalty, boost customer satisfaction and, most importantly, close more sales.
The discipline of data analytics is a development of earlier techniques often referred to as Business Intelligence or BI. BI software suites helped order and analyse enterprise data in order to glean insights into how companies could serve their customers better. They might uncover routes to efficiencies in-store, through the supply chain or at the back in the warehouse. There are various definitions of data analytics available. One that will probably be of most use to car dealerships explains it as 'the process of examining large data sets containing a variety of data types to uncover hidden patterns, unknown correlations, market trends, customer preferences, and other advantaging business information'.
Nick Gill explains;
Data scientists start with a hypothesis – such as people living in a certain area, and of a particular demographic, are more likely to buy an electric car. They will then 'mine' data from a wide range of sources to substantiate the theory, and if it stands up, they identify the people to target with that particular message. Delving into their data would enable dealers to answer key questions like ‘which of your customers and prospects who will be buying a car in the next three-to-six months?’; ‘what types of vehicle are they considering?’ and ‘what price do they want to pay? Having an accurate picture of [which customers are] likely to remain loyal and which will churn, is valuable,” Gill adds. “If you know one of your customers is switching from one brand to another, as a group with both brands in your portfolio, you can….introduce them to the sister dealership – and [thereby] keep their custom within the group.
The ability to gain reliable foresight of car buying patterns would naturally enable dealers to focus their sales and marketing strategies on demographic categories likely to yield greatest customer engagement. For the commitment to return dividends, however, dealers need to change their view of data, Gill says, and recognise it as “a genuinely scientific approach to communicating with car buyers”.
Gill also predicts that, initially, dealer groups will commission data science services ‘from specialist providers’, but he thinks that within five years dealers will be employing their own in-house data science teams. Embedding such specialists into a dealership sales structure will, however, bring challenges.
As indicated, data scientists are supposed to differ from typical information technology professionals in that they are not necessarily 'techies', nor do they necessarily have computing expertise (although many have); however they are likely to come from statistics or mathematical modelling backgrounds.
If the automotive sector may be welcoming these data scientists a little later than other verticals, it has the advantage of being able to learn from the experiences of its predecessors. As with any 'new' manifestation of IT, managing expectations is critical to delivering long-term value.
Whatever approach to embracing data analysis dealer groups adopt, automotive data scientists will have no dearth of data to work with. CRM solutions, omnichannel platforms, and online customer enquiry systems are already generating petabytes of the stuff. Whether a dealership decides to in-source or outsource this new resource, the process is not as straightforward as simply taking data sets from the CRM system, say, and handing them over to a data scientist to work on, as Derek Franks, a data scientist with IBM's Centre for Applied Insights has noted:
Models are only as good as the data you feed them, so it's important to think about how you organise and manage the data. Data scientists are often jokingly referred to as 'data janitors', because we spend 80 per cent of our time 'cleaning-up' data.
Looking ahead, it is the connected vehicle revolution that many market-watchers see as releasing masses of commercially-actionable car usage-based intelligence that, ultimately, will trigger the ‘big data’ revolution in this market. Data sent from customers' cars will include information about vehicle usage and servicing that should be captured and used to inform CRM and omnichannel strategies.
Dealerships need to ensure they are qualified and have the resources to exploit this rich seam of on-road intelligence. They need to do so to ensure that they remain part of the connected vehicle value chain. The wider hope within some quarters of automotive retailing is that the arrival of connected cars in large numbers will be a driver for the technology industry investment that will deliver the increased processing power and the more efficient algorithms needed, to sift the big data sets quickly enough, to put the new intelligence to work before it goes stale.