The pace of innovation in the Automotive Industry has never been faster. The widespread adoption of electric vehicles (EVs), self-driving cars, connected cars, and shared mobility services – in both advanced markets and emerging economies – has pushed the industry to what is unarguably its highest inflection point, many magnitudes of order higher than any point in its 100+ year history.

Automotive original equipment manufacturers (OEMs) and their Tier 1 Suppliers are pouring in capital investment to produce high energy density cells at low cost, to cash-in on what seems to be a quasi-infinite demand for lithium-ion batteries that power EVs.

However, an even more telling trend than the growth of EVs is the growth of “connected” cars. Equipped with hundreds of intelligent, expandable, mobile electronic devices and on-board sensors, connected cars stream millions of data points on the health of cars for proactive monitoring and pattern detection every second. Routine visits to the dealership or local garage mechanic will become less frequent as hungry Machine Learning algorithms feed on a myriad of sensor data, processed in the Cloud (or on the Edge), generated from across the entire vehicle fleet.

Perhaps the real game-changer aspect of connected cars is their ability to add new features, upgrades, and updates entirely over-the-air (OTA). Most automakers are already designing vehicle hardware to support software updates. This enables manufacturers to shift to a revenue model that is based on services—rather than a one-time sale of a car or truck. It should be noted that Tesla began monetizing OTA upgrades in 2019 when it offered Model 3 owners an acceleration boost—from 4.6s to 4.1s—for $3,000. 

Opportunities and Challenges in the Connected Future

Ironically, connected cars will appreciate (not depreciate) over their lifetime. Seems counter-intuitive, doesn’t it? To put it in perspective, during Tesla’s most recent earnings call, Elon Musk likened the upside potential of Teslas on the road today, with the ability to turn on full self-driving (and eventually autonomous driving), to the “largest asset price increase of anything in history!”

Piece it all together and one can begin to see the downstream impacts across the value chain – car dealerships, insurance providers and a confused used car market, all in a state of abeyance and wanting to partake in a share of the increases in vehicle asset prices.

Several leading OEMs are in the process of re-imagining existing sales, marketing, and service processes into a holistic and unified approach, as these departments become more symbiotic. In parallel, OEMs must be quick to embrace their new role as the nerve centre of a rapidly evolving value chain ecosystem comprising hardware suppliers, software vendors, Cloud hyper-scalers, systems integrators, specialist data partners, Tier 1s and secondary (used car) customers… no easy feat!

Capitalising on the Customer Connection with a Direct-to-Consumer Model

In the world of connected cars, the OEM too has a seat inside the car (I say this metaphorically, of course); from customers’ driving behaviours, aesthetics and taste, to their lifestyle preferences, in-car feature purchases, communication preferences, and more, OEMs have an unparalleled opportunity to truly learn about each customer in ways they (and their dealers) couldn’t have fathomed before. The key question is: how can OEMs “hyper-personalise” customer experiences, and offer new features and value-added services that resonate with target customer segments?

News flash: Of all the global major OEMs, Tesla is the only one that operates with a D2C model. No surprise that its operating margin is an enviable 17% compared with the high single digit profit goals of all other major OEMs. How much of these D2C efficiencies went into subsidising recent Model 3 and Model Y price cuts and really turning the squeeze on its competitors?

Here’s an illustration of how Stellantis, a leading global Automotive company, is executing on its vision for software-defined vehicles. It is worth noting that Stellantis is aiming to generate €20 Bn from incremental software-enabled services revenues by 2030 across its 14 iconic brands – including Fiat, Maserati, Jeep, Peugeot, Opel, Citroen, Ram, Dodge, Vauxhall and many more.

So, whether it’s an acceleration boost for $3,000, a golden haptic cube that surfaces scenic destinations at 50 cents a pop, a $1,000 bolt-on feature to summon your car to your GPS location, or an exclusive member service (pay-per-use or priced in) for on-tap access to chauffeured cars, the digital revenue streams for the OEM (within a communications framework that prioritises customer choice and privacy) are limitless. The trick lies in algorithmically identifying customer segments from the D2C data and offering hyper-personalised features and services that align to the brand and market strategy.

By applying Artificial Intelligence (AI) and sophisticated Machine Learning techniques to customer behaviour data captured in the car, compounded by powerful network effects, there’s a clear path OEMs can tread to increase customer lifetime value. Considering that the marginal cost of these digital services is zero or negligible, the impact on bottom line when scaled across the vehicle asset/customer base cannot be ignored.

To ensure customers stay happy, new digital sales and marketing efforts must go hand-in-hand with a preventive service management framework. From automated alerts for failing parts, to highly optimised booking of services or repairs, to sensor-based driver insights, there’s plenty of opportunity for personalisation within the maintenance and repair cycle. It is vital for the AI algorithms to collate a 360° view of both the customer and the car and integrate next-best-action models into customer communications.

An aside: If you’re over 40, there’s a good chance you’ve watched the action crime drama television series, Knight Rider – the 1980s show was the quintessential nostalgic look at how we perceived AI if it was something other than a computer or a robot.

If you have, ask yourself this question – if Michael were forced to pick between KITT’s autonomous driving or interactive voice capability, which of the two would he choose? What are the chances he would opt to retain the ability to talk to KITT even if that meant foregoing the auto-drive?

From the lens of a conjoint analysis, customers derive different utility (path-worth) from different features and therefore exhibit different willingness to pay for those features depending on which customer segment(s) they belong to.

Tip: The journey to a fully autonomous car has undoubtedly been made easier by modern, scalable software development architectures, de-coupling of hardware and software cycles, robust data management and deep-learning capabilities.

At the same time, the massive investment in AI horsepower (specifically the compute and energy costs of GPU clusters to train complex deep learning models at millions of frames per second) over long gestation cycles, and the increasing complexity of code inside the car to keep pace with safety regulations means that only a handful of the major OEMs are true contenders in the race to develop the first autonomous car.

For the vast majority, they may do well to shift their innovation and AI budgets away from the development of an autonomous car to more lucrative personalised features and services… so long as they have a flourishing data ecosystem and a refreshing data-driven mindset, they need not play second fiddle when it comes to monetising software-enabled digital revenues.

Seizing the opportunity

The Automotive Industry is at a tipping point. Driven by significant and lasting changes in consumer demand, electrification and sustainability, and a new age of software-defined vehicles (particularly full self-driving and connected cars), those OEMs who embrace an AI-based market differentiation strategy will thrive.

At a time when Industry profit margins are under pressure, AI and Machine Learning offers a silver lining. By using Machine Learning to answer key questions – which digital features and services to offer, when and to whom, on what channels, and at what price – OEMs can algorithmically identify and target customer segments at zero marginal cost.

Preventive maintenance is equally important, whether to detect anomalies, signal deviations from specs, or optimise repairs and logistics costs. From a Customer Experience (CX) standpoint, even such high service standards may soon be considered table stakes. Moreover, since it is typically desired across all customer segments, proactive service management could be the perfect starting point for OEMs to focus on when building out the customer lifetime value picture.

While the silver lining is real, it will take much more than a unified sales, marketing and customer service approach to increase revenues and profit margins. Replacing classic form factor-based A-B-C segments with algorithmically derived customer segments to guide R&D cycles, optimise factory mix, predict supply/parts lead times, and scout out supply chain efficiencies requires a shared understanding of the benefits of AI and Machine Learning across the enterprise.

While OEMs re-allocate their innovation budgets towards orchestrating the data pipelines and Machine Learning models that enable hyper-personalisation, it would be nice to see financial analysts and the investor community bringing in dimensions of customer lifetime value into their assessment of OEM performance – perhaps a better proxy for OEM brand valuations than the mundane, short-term focus on demand-to-production ratios, and average selling price of cars every quarter.

If you have any questions about getting started with AI and Machine Learning, developing your strategy and roadmap, or addressing specific use cases, feel free to reach out to me at gaurav@inawisdom.com.

———

Inawisdom are leading consultants in AI, Machine Learning and Data Analytics. As an AWS Premier Services Partner, we help data-rich businesses discover valuable insights that enable data-driven decision making. Inawisdom has been part of the Cognizant family since 2020.