View profile

🚗 Car Talk - Issue #25 (Bundles & Over-The-Air Connectivity)

🚗 Car Talk - Issue #25 (Bundles & Over-The-Air Connectivity)
By Sachin Seth • Issue #25 • View online
Hi friends
This issue of CarTalk is unlike others.
Ever since I wrote the CarTalk issue on subscriptions, I realized there is much more to explore there. This (very!) long form issue is my attempt at doing just that.
If you don’t have the time to read the entire thing (or can only afford chunks of time), the summary of my thesis is that Over-The-Air Connectivity (OTAC) has the potential to unlock tremendous untapped value and will change not only how we own and operate vehicles, but will end up changing the mobility ecosystem overall. Any player (OEM or startup) working on this technology is worthy of serious attention. Anyone who ignores OTAC will be written off, or be a niche player at best.
Will it truly happen? I don’t know. I’m bullish, however.
This mega-newsletter is divided into 5 sections:
  1. The Economics of Car Ownership
  2. (Optional): The Revenue Benefits of Bundling
  3. The Marginal Costs of Bundling
  4. Over-The-Air Connectivity: The Magic Bullet
  5. OTAC Unlocks New Business Models
All these opinions are my own.
Just.. buckle up, won’t you?

The Economics Of Car Ownership
Car ownership is a burdensome, time-consuming affair. The moment you drive the car off the dealer’s lot, you have signed yourself up for a list of responsibilities that just didn’t exist in your pre-automotive life. Throughout the ownership lifecycle, a car owner will have to pay for many services to ensure the driving experience is legal, safe and enjoyable. I summarize this in the figure below:
Figure 1: Current State - A car owner interacts with many entities to be able to drive a car. With every vendor, the owner incurs a product cost (pc) and a transaction cost (tc).
Figure 1: Current State - A car owner interacts with many entities to be able to drive a car. With every vendor, the owner incurs a product cost (pc) and a transaction cost (tc).
There are various categories of services/features that a driver has to pay for after purchasing a car. Other than obtaining a car loan, all of these transactions are recurring events:
· A driver has to procure financial services to own and operate the vehicle – a car loan with a low interest rate is preferred, an insurance policy with a low premium is preferred e.g. a driver could be interacting directly with Chase Inc. for a car loan, but Geico for insurance.
· To operate the vehicle, a driver must continuously pay for fuel or charging (in case of electric vehicles) and service the car periodically. An example would be paying for gas at Costco and for servicing at Firestone.
· Lastly, a car owners might want to pay for tech/luxury features such as:
  • Paying for a GPS device or high-definition maps (TomTom, Garmin, a phone app) that displays real-time traffic data
  • Enabling heated seats
  • Uncorked acceleration
  • Unlocking hands-free driving software
Each of these transactions comes with costs to the customer – there is a product/service cost (pc), as well as a transaction cost (tc) to pay. Product/service cost is straight-forward – consider the cost of gasoline pumped, cost of insurance, and so on. Transaction costs are harder to quantify but exist nonetheless in the form of time value of money, or customer experience. You will have to waste time on the phone with Geico, you may have to visit many lenders to get the best bank loan, you will spend time calling around many service centers to get the cheapest bumper replacement costs, and so on. A high transaction cost is why some might choose to continue driving around with their “Check Engine Light” lit up – they can’t muster the effort to call around shops to fix (I’ve certainly done it).
Almost all of these transactions are mandatory - there is no getting around them. Given that drivers have to spend all this money to buy these products/services anyway, wouldn’t it be better if they just had one vendor (who could bundle and sell these services) to deal with?
Figure 2: Future State - By interacting with just one entity that can bundle all these services together, the driver can minimize transaction costs (tc). The bundler can also act as one or multiple vendors on the right.
Figure 2: Future State - By interacting with just one entity that can bundle all these services together, the driver can minimize transaction costs (tc). The bundler can also act as one or multiple vendors on the right.
At the end of the day, savvy consumers will pick the vendors that allow them to minimize the sum of pc and tc. If the car company (OEM) itself, or a startup steps up to consolidate all these services and offers them to the new driver as a bundle, the driver saves immensely on tc. Now, pc may/may not reduce (esp. when compared against the summation of all pc’s in Figure 1), but I have reason to believe it will. Lastly, the bundler can also act as one or multiple vendors on the right in Figure 2.
Optional: The Revenue Benefits of Bundling
(Feel free to skip this section if you received a B+ or better in any Economics class!)
Imagine you’re launching a new product in the market and would like to appropriately price it to maximize your revenues. If you price it too high, you wouldn’t expect to sell many units. If you price it too low, it will sell really well but strain you operationally. As you plot the various (Q,P) combinations on a 2-D graph, you get the demand curve.
Figure 3: Demand curves, aka CarTalk butchering Econ 101.
Figure 3: Demand curves, aka CarTalk butchering Econ 101.
If you want to maximize revenue, you’ll pick a point on the demand curve such that the area of the green rectangle is maximized: 
Figure 4: Goal is to maximize the area of the green rectangle by picking the right spot on the demand curve.
Figure 4: Goal is to maximize the area of the green rectangle by picking the right spot on the demand curve.
As you traverse up and down the demand curve, you find that the optimal price of this product is P_optimal and the quantities to produce & sell are Q_optimal.
Total revenue (area of the green rectangle) = P_optimal * Q_optimal.
Easy peasy.
The issue with this “Total revenue” is all the white space under the demand curve that you could do nothing about. By picking P_optimal and Q_optimal, you miss out on:
  1. The fat-cats who were willing to pay a price > P_optimal for your product (thus letting you earn more profit per every sale). This is known as “consumer surplus” and is represented by the blue triangle in the figure below
  2. The bargain shoppers who would’ve bought your product if only you priced it lower than P_optimal, represented by the orange triangle titled “deadweight loss”.
Figure 5: In an ideal world, a seller would want to capture all three colors under the demand curve.
Figure 5: In an ideal world, a seller would want to capture all three colors under the demand curve.
If you were financially savvy, you’ll be devising ways to claw back the blue rectangle and the orange rectangle somehow.
Enter bundling of products.
This is an example I’ll lift straight from Prof. Kent Smetters’ Managerial Economics class notes. Kent and you are both in the market to buy a car. Kent would love a radio in his car but doesn’t care much for AC, you’re just the opposite.
Willingness to pay for features for two different customers.
Willingness to pay for features for two different customers.
If the car company wants to maximize total revenue, they could price both the radio and AC at $4,000 each and earn $24,000 from Kent (Car + Radio, he passed on the AC) and $24,000 from you (Car + AC, you pass on the radio). Total revenue is $48,000. Price the radio/AC any higher and you lose a sale entirely. Price them any lower and while you can make the sale, you lose consumer surplus as well as reduce revenues.
Now, if the car company is crafty, they can come up with a plan! For $5,000 they’ll sell a bundle of AC and Radio. Kent is all over it, so are you. You both take the deal and the car company can earn $50,000 total. They bundled the goods and increased total revenue by $2,000.
Mathematically, what happened has been described already by economists in this amazing research paper titled “Bundling Information Goods: Pricing, Profits and Efficiency”:
Figure 6: As a seller bundles more commodities together, the demand curve of the bundle flattens.
Figure 6: As a seller bundles more commodities together, the demand curve of the bundle flattens.
As you bundle more goods, you’re flattening the demand curve for the overall bundle. Throw more items into the car buying example (ACs, Radios, Heated seats, free tune-ups, you get the idea….) and the demand curve flattens.
Figure 7: New demand curve, who dis?
Figure 7: New demand curve, who dis?
A flatter demand curve allows a seller to increase the area of the green rectangle in a way that a single demand curve wouldn’t have.
You see where I’m going with this, right? Bundling of goods is not only valuable to the consumer (both Kent and you now drive a car with a radio AND air-conditioning), it is also valuable to the supplier.
The Marginal Costs of Bundling
If you skipped the previous section, I concluded that bundles maximize revenue. The discerning reader, however, would’ve already noted that it is profits we want to maximize (“Alex, What is Capitalism 101?”).
We know that Profits = RevenueCost (….hold that Nobel prize, please).
In the paper I referenced earlier, the authors say on page 6: “In the basic model, we assume that marginal costs are zero. While very large bundles will typically continue to be profitable even in the presence of non-zero (but small) marginal costs, bundling becomes unprofitable for goods with substantial marginal costs”.
Well, then - let’s look at marginal costs of our proposed automotive bundle, shall we?
1. Insurance: The business of insurance (automotive, Obamacare, etc.) is predicated on getting a lot of individuals to sign up so the healthy population can subsidize the ones that are infirm. Thus, the more insurance policies a bundler can directly sell, the better rates it can command from its underwriters and pass on the savings to its customers.
Since it is a legal requirement for every driver on the road to have insurance, insurance can be the crown jewel of any bundler’s offering. The marginal cost of adding another driver is low (but can be made to go even lower, as we’ll see shortly).
Today, companies like Porsche, Mercedes, Tesla already offer their own insurance products.
2. Car Loan: Ford has a finance company inside the corporate entity. Toyota offers in-house financial services too. So does Honda and Nissan and VW and BMW and Mercedes-Benz and GM and.. I’ll stop. These companies are called captive finance companies and their sole purpose in life is to help their parent companies move metal.
I’m not an expert on loan/banking, but selling loans bundled with other automotive offerings ought to keep marginal costs very low. If a bundler can increase the number of car loans, there are second-order benefits to reap noted at the end of the write-up.
3. Fuel/Charging: Every car needs gasoline / charge to move. Focusing on electric vehicles for now, a number of companies are setting up charging infrastructure nationwide. GM already has partnered with EVgo, so if you’re the lucky owner of an electric Hummer, you can charge your car on EVgo’s network. Similarly, if you’re the lucky owner of a VW ID.4, you can charge your car free for 3 years using Electrify America’s networks.
What is the marginal cost of adding a driver on a pre-existing charging network? Non-zero, certainly. At the same time, there have been research studies showing that consumers tend to charge their electric vehicles more at home rather than at public charging stations (ease of use, better user experience). So even though marginal cost of car charging isn’t zero, there is reason to believe that consumers will pay more relative to what they’d consume piece-meal.
4. Servicing: Adding service packages to bundles can get costly. More bundles leads to more service packages being sold, which leads to requiring more servicing locations + manpower needed for a good customer experience. Marginal costs of including servicing in a bundle are non-zero, but there are ways to minimize.
5. Tech/Luxury features such as:
  • HD Maps
  • Elevating in-cabin entertainment: Buy media (audio, video)
  • Elevate driving experience: Heated steering wheels, heated seats, uncorked performance
  • Hands-free driving or self-driving features
Most cars today come pre-fitted with hardware to enable these features, all that’s needed are bits of code to activate them. Marginal cost of these features? Marginal cost of software, which is a basically zero. This category of offerings will become very lucrative since every sale made is pure profit for the bundler.
To summarize:
Why a bundler would consider bundling, and why a consumer will buy the offering.
Why a bundler would consider bundling, and why a consumer will buy the offering.
Bundling will be attractive both for:
  • Customers, because it reduces their transaction costs immensely (one-stop shop for all auto needs) and product cost too, in cases like insurance.
  • OEM and startups because bundles allow them to maximize total profits by maximizing revenue (previous section) while keeping marginal costs low (the marginal cost of a lot of the bundled offerings is zero, or near zero).
Don’t believe me? The authors of the amazing paper say so too: “With independent linear demands for the individual goods, bundling any number of goods with zero marginal cost increases the seller’s profits
Capitalism 101.
Over-The-Air Connectivity - The Magic Bullet
In the previous section, I glossed over the marginal costs of Insurance as well as Servicing. Let’s address that now. Resurrecting an older figure, with a twist this time:
Figure 8: Over-The-Air Connectivity (OTAC) allows the car to communicate with services 24x7, unlocking newer ways of doing business.
Figure 8: Over-The-Air Connectivity (OTAC) allows the car to communicate with services 24x7, unlocking newer ways of doing business.
If the vehicle has the ability to interact with the bundler at all times (a bi-directional communication where the bundler can “check-in” on the car in the field), it unlocks the potential to play some very strategic games. Specifically, OTA connectivity (OTAC) allows the bundler to lower its marginal costs even further for all the offerings touched by the brown line. How so? Let’s take a look:
1. Insurance – If the bundler can somehow know these things about you:
  • where you are geographically situated at a point in time (e.g. a safe suburb, or downtown in a big metropolis?)
  • your driving style (are you a lead-foot, or do you drive like my grandma?)
  • your rates of disengagement of hands-free driving
  • how many times you change lanes
  • …hundreds of other variables
it could, in theory, change your insurance rates in real time! In practice, the bundler certainly could re-calibrate your rates on a monthly basis (i.e. faster than the usual six-month insurance refresh cycle). The beauty of the idea, though, is that it cuts both ways - it could also reward you by lowering rates if your driving habits are good and you park in a known safe neighborhood, etc.).
In the context of a bundle, the OTAC allows the car company to raise or lower rates at will, thereby protecting its profits even more so that it is not exposed to your bad driving. Without OTAC, without bi-directional data transfer, this isn’t possible.
To some extent, companies like Metromile already offer this product. However, the only two data points they have about a car are its:
  1. Geographical location
  2. Speed/acceleration
Metromile cannot use the tens of other sensors a modern day vehicle is outfitted with (radar, cameras, driver monitoring system), putting them at an information disadvantage and discouraging fair pricing.
2.  Service: If a bundler could use bi-directional data transfer to gauge the health of the car and perform pre-emptive measures to ensure the longevity of the car, it could save tremendously on operating costs. “A stitch in time saves nine” as the saying goes. With “predictive maintenance” made possible by OTAC, the bundler can lower its marginal costs even further and make the bundle more profitable for itself.
3. HD Maps & Software Services: OTAC really unlocks the pay-per-use business model for these use-cases. Suppose you have a long road trip planned and you would like to use the autonomous driving feature, get Apple music on your car, and get heated seats. With the click of a button, you can enable these features in the car . When you’re done, turn them off.
Remember, almost all cars come with the hardware pre-fitted. BMW is doing this, as we’ve discussed in prior CarTalk issues. You merely need bits of code to disable/enable the hardware - easily accomplished via OTAC. Every sale is profit. Do it as many times as you’d like, the transaction costs (tc) is literally zero, enabled by OTAC.
So far, I’ve made the case that bundling products will allow bundlers (OEMs/startups) to increase their total revenue. I’ve also made the case that OTA connectivity will allow the same bundlers to lower the cost of their offerings, making the bundles much more profitable than if they were to sell offerings piece-meal.
It did not come as a surprise to me when I read a headline that has managed to fly under the radar so far: “BMW Group rolls out huge OTA upgrade”. More than 750k vehicles will receive a new OS..
BMW, I see what you’re up to.
OTAC Unlocks New Business Models
As you’ve caught on by now, the bundle in Figure 8 comprises of merely digital bits. There is no tangible good that has changed hands. Therefore distribution cost of these bundles is basically $0 (okay, you’ll have to pay T-Mobile or AT&T to push/pull data to the vehicles but you get the idea). As a result, OEMs and startups can start thinking about new ways of doing business. Some of the off-shoots of what OTAC + bundles unlocks is listed below:
1. OTAC allows price discrimination
$0 distribution cost means automotive bundles can be turned on and off at the driver’s whims. The driver could toggle it every month. Every day. Every week. Every hour. The physical limit (the “quantum” of usage) is the round-trip time of the ping from the car to the bundler’s servers and back.
This technology allows the bundler to capture a ton of consumer surplus. Reviving my old demand curve:
Figure 9: Due to no distribution costs, the bundler can capture the entire area under the curve.
Figure 9: Due to no distribution costs, the bundler can capture the entire area under the curve.
If you were a car company and had a working OTAC the likes of BMW above, wouldn’t you want to traverse up and down the demand curve and monetize every paying customer, as in Figure 9?
OTAC enables price discrimination (the good kind), allowing sellers to capture all the area under the demand curve. In this business model, you keep the bundle offerings fixed but allow variability in the time of use.
2. OTAC allows subscriptions
With a $0 distribution cost and zero (or near zero) marginal cost of offerings, these bundles are begging to be purchased as subscriptions. In this business model, you keep the time of use fixed (monthly, semi-annually), but allow variability in the offerings that comprise the bundle.
As a customer, let’s assume you really don’t care for a bundled car loan because you want to buy the car outright. You also don’t care for charging privileges in your bundle because you charge at home. You do want all other offerings. Easy enough to package and deliver to you since distribution costs are zero.  
This is exactly what BMW did. Recently, BMW started to sell subscriptions to its car features, a must-read article into how the company is thinking about its place in a digital future.
3. OTAC → subscriptions → recurring revenue → debt instruments
This is a second-order benefit of OTAC and was alluded to under “Car Loans” earlier. Predictable/recurring subscription revenue from customers/drivers can be thought of as a fixed-income instrument, no different from corporate bonds. Recurring revenue can be securitized and offered as a debt instrument as a way for bundlers to raise capital, instead of issuing equities and suffering from dilution.
Figure 10: https://twitter.com/patio11/status/1223104192094461961
Figure 10: https://twitter.com/patio11/status/1223104192094461961
The idea is not new, but is certainly new to the automotive world. The concept of securitizing recurring payments is less “out there” than I make it sound. In the past, banks have routinely offered bond deals made up of auto loans to borrowers with deeply tarnished credits. If they can get away with it, the bundlers could too.
4. OTAC allows bundlers to be efficient Aggregators
Aggregation Theory was coined by Ben Thompson of Stratechery. Let’s take a figure we’re familiar with by now and view it in a different light:
Figure 11: With a critical mass of consumers/drivers on the left side, a bundler can commoditize vendors on the right side and continuously swap them out to lower its own costs.
Figure 11: With a critical mass of consumers/drivers on the left side, a bundler can commoditize vendors on the right side and continuously swap them out to lower its own costs.
If the bundles are selling like gangbusters (and why wouldn’t they, given their value to customers?), bundlers start to enjoy immense market power. In Ben Thompson’s words, the bundlers become efficient aggregators whereby “..suppliers can be commoditized leaving consumers/users as a first order priority. By extension, this means that the most important factor determining success is the user experience: the best distributors/aggregators/market-makers win by providing the best experience, which earns them the most consumers/users, which attracts the most suppliers, which enhances the user experience in a virtuous cycle.
I emphasized part of that quote because the future I envision is one the automotive business is headed towards being a service-oriented business, more so than ever. When every electric car offering out there will have mind-boggling ranges, neck-snapping accelerations, the hardware differentiation will be less and less important. The market-leaders might be someone not with the best hardware and specs, but with the best end-to-end customer ownership experience.
All of this can be enabled by bundling services made possible by OTA Connectivity features.
Conclusion: In this long-form version of CarTalk, I make the case that Over-The-Air Connectivity (OTAC) is going to be very crucial, unlocking tremendous value for those in this business. Unlike autonomy/robotaxis where the timelines may still be years away for most players, OTAC can unlock value today. Any OEM or startup working to enable a connected-car future, or starting to formulate business models underpinned by OTAC is positioned (both technically and economically) to pull ahead of its peers and usher in new mechanisms of car ownership and, frankly, mobility.
Bonus Meme Conclusion:
Step 1: Make Bundles
Step 2: Sell Subscriptions
Step 3: ???? Use OTAC to lower costs
Step 4: Profit!!
Did you enjoy this issue?
Sachin Seth

This weekly newsletter on new mobility is curated by me as a passion project. Yes, the name is an homage to the NPR show of the same name! If you like it, please forward it to whoever is interested in this space. Cheers!

I have worked for many years on automotive products and currently work @ Tesla. All opinions are my own.

In order to unsubscribe, click here.
If you were forwarded this newsletter and you like it, you can subscribe here.
Powered by Revue