[U24] Update: innovation at the margin
Gamifying safety plus vertical integration to disrupt legacy insurance
Dear Reader,
This update in Product | Strategy | Innovation explores innovation at the margin. The term margin can describe something at an edge, but also represents gross or operating profit as a percentage of revenue. Both meanings are relevant to this update, but are also important within the context of vertical integration. New ventures like AirBnB in their early years might exploit core competencies like customer acquisition and capital efficiency to disrupt incumbent operators. They are lean and hustle. And they are focused on growing a core business. These ventures can also use capital markets to offset losses while scaling operations and revenue. With adequate scale, margins improve over time with operational efficiencies. Vertical integration outside of core competencies is mostly out of scope.
But growth-focused tech companies with sustainable and growing cash flow can take an expense under operations like information technology, vertically integrate the supply chain to advance these systems in-house as a core competency to solve their own problems and then develop a plan to convert the expense into revenue by selling these new capabilities to customers. That is the Amazon playbook used to create its 3 pillars: Marketplace (fulfillment sold to the Amazon ecosystem), Amazon Prime (customer acquisition and loyalty to increase the long-term value of a customer for the Amazon ecosystem) and Amazon Web Services (cloud & edge computing services sold to the Amazon ecosystem by subscription).
This requires significant investment in fixed costs like skilled labor, research into emerging technologies and development for long-term advantage. But first you should determine if there is a real problem that needs to be solved and whose problem is it.
Automobile Insurance
An estimated global fleet of 2 billion vehicles transverse city streets and highway systems on a daily basis around the world. Human error, weather and other factors also lead to automobile accidents on a daily basis. Many of these accidents result in minor damage paid for by most drivers out of pocket, but major damage or a total loss can put significant hardship on many individuals when one or more vehicles are involved. Automobile insurance premiums paid by a large cohort of drivers helps to subsidize claims made by a subset of drivers in that same cohort who are responsible for the accidents.
Automobile insurance is priced by insurance companies using some form of a price calculator with a number of factors to help price the risk profile for a driver. The Hartford uses the following factors to help price auto insurance.
Telematics enabled by specialized hardware and provided by an insurance company to a driver monitor actual miles driven for pay-by-the-mile policies, safe driving for rewards based on activating specific telematics, and driving behavior to provide personalized feedback. But this technology was developed mostly outside of the insurance industry and is available across multiple insurance companies for limited competitive advantage.
Insurify and other technology-focused companies have added a layer services on top of insurance to source bids from multiple insurance companies based on a driver and vehicle profile to help drivers optimize auto insurance while also lowering premiums. Lemonade will use AI to help estimate the risk profile for a driver and process claims more efficiently. And legacy insurance companies like State Farm, Allstate, Liberty Mutual and others bundle auto, pet, home and life insurance to reduce premiums with greater efficiency acquiring and servicing customers across multiple lines of insurance products.
The real problem with auto insurance is safer drivers subsidize claims from drivers with the greatest risk.
But the primary problem for automobile insurance even with all of the incremental advancements within the industry is premiums for safer drivers still subsidize claims from drivers with the greatest risk. This problem negatively impacts the safest drivers the most. And the problem also impacts outliers like safe drivers with a more expensive electric vehicle ahead of mass adoption versus unsafe drivers with a less expensive internal combustion engine vehicle that is widely adopted. Telematics are a step in the right direction when discounts are provided based on actual driving data with the insured vehicle.
When Tesla entered the luxury sedan market in 2012 with the Model S, early adopters in the United States buying these vehicles often paid much higher auto insurance premiums due to the low number of Tesla vehicles on the market, higher prices for parts and higher prices for repairs. Tesla announced in 2019 it would start offering auto insurance in California where Tesla vehicle adoption was very high. Other auto manufacturers have offered auto insurance from time to time with limited success. But Tesla saw accelerating the transition to sustainable energy might require taking on insurance to help manage a key input to the total cost of ownership for electric vehicles.
First principles and the best talent wins!
Tesla Insurance has been a niche product to date outside of the core business, but as innovation within Tesla accelerates as vehicle production starts to scale exponentially, auto insurance becomes more strategic to advance the company’s mission. However, Tesla is not well positioned to out compete with other insurance companies based on insurance industry tactics alone. Tesla has to leverage its entire technology stack and engineering talent to out compete auto insurance companies who are narrowly focused only on insurance. First principles and the best talent wins!
A key pillar of innovation at Tesla illustrated above is Artificial Intelligence. One major use case is computer vision with machine learning using billions of miles of Tesla vehicle data to train a neural network. Edge computing with multiple cameras and custom semiconductor chips in Tesla vehicles sense the environment and control the vehicle for autonomous navigation. Tesla Full-Self Driving (FSD) alpha was used by Tesla employees to verify key functionality for the early system, but Tesla expanded FSD beta to about 2,000 testers in 2020 to accelerate the number of miles driven with FSD. This also accelerated training Tesla’s neural network. But a real objective has been to make FSD beta available to more drivers who purchased FSD for their Tesla vehicle. FSD beta v10.2 was determined to be ready for wider use, but Tesla wanted to roll out FSD beyond the 1st 2,000 testers very selectively to its safest drivers to help manage risk while accelerating the cumulative miles driven with a larger cohort of FSD beta drivers. This will become even more important when training the neural network is accelerated as well with the introduction of Tesla’s special-purpose super-computing platform called Dojo to automate this training.
Tesla Gamifies Safety
Tesla wanted to rank order Tesla drivers who opted in to its FSD beta evaluation program based on what it called the Tesla Safety Score. The safety score uses the edge computing already installed in Tesla vehicles to monitor Tesla drivers on specific parameters only if 1.) the driver bought FSD for the vehicle and 2.) the driver opts in to the FSD beta evaluation program. Tesla drivers who opt in are monitored and rated on a scale of 0 to 100 — the higher the score, the safer the driver, and the faster the driver will be given access to the broader suite of FSD features through FSD beta. So Tesla drivers in the FSD beta evaluation program with a Tesla Safety Score of 100 would be prioritized to receive FSD beta first, then drivers with a score of 99 next and so on.
According to Tesla, five safety factors determine a driver’s safety score:
Forward collision warnings per 1,000 miles
Hard braking
Aggressive turning
Unsafe following of other vehicles
Forced Autopilot disengagements
Tesla recently released FSD beta v10.2 to 1,000 additional drivers who purchased FSD with perfect scores of 100 on the safety score. Tesla also announced on its recent October 19th Q3 earnings call there are 150,000 Tesla vehicles using its safety score. But Tesla is also looking at another use for the Tesla Safety Score for even more of its customers. Tesla benefits if their vehicles are the safest vehicles on the road. And that improves when the Tesla drivers are the safest drivers they can be until FSD is widely available and used for the majority of the driving experience.
Tesla drivers who currently use the Tesla Safety Score obtain an updated score daily on the Tesla app associated with their vehicle. Tesla has stated the probability of an accident for drivers using the Tesla Safety Score has gone down 30% compared to drivers who are not using the safety score based on 100 million miles of driving data. This is likely due to the motivation to get a higher score to qualify as a FSD beta tester, but what if incentives and discounts through the Tesla Safety Score were offered to an even broader base of Tesla drivers who simply opt in to telematics in states where Tesla offers an auto insurance product.
Tesla recently launched an Auto Insurance product in Texas where pricing is based on telematics and the Tesla Safety Score and premiums are adjusted month-to-month based on recent safety score results. Tesla is fully transparent with how the score is calculated so drivers know how to improve the score. But claim history, credit history and other standard features to price auto insurance do not play a role in pricing this Tesla Insurance product. As FSD evolves to reduce vehicle accidents and fatalities and becomes more widely adopted, Tesla vehicles will become an even lower auto insurance risk.
A key strategy for Tesla to accelerate the transition to sustainable energy is a relentless focus on continuously lowering the total cost of ownership for Tesla electric vehicles. This was fundamental to the original master plan where each model release would realize a much larger production volume to decrease unit production costs. Designing different models at different price points is also important to offer affordable models in the Tesla portfolio for as many people as possible. Model configurations help balance features and cost. But these are somewhat linear on the product roadmap as part of product portfolio strategy and execution.
Innovate at the margin to continuously lower total cost of ownership.
Innovation at the margin provides unexpected opportunities to simplify process, reduce parts, gain efficiency and uncover other advantages to further lower costs. Higher quality and an enhanced customer experience are also possible. Tesla is already an industry leader for robotics and automation, but as previously mentioned in another update [U10], the Tesla Model Y is benefitting from die-casting underbody segments as single parts versus spot welding 70 parts to create the same component. The Tesla “giga press” used today was sourced from an Italian company called IDRA.
Other car manufacturers had the same opportunity, but Tesla made it happen with a custom aluminum alloy to inject molten metal into the mold under pressure to fabricate such a large part. Materials Engineering is a cross-company function between Tesla and SpaceX led by Charles Kuehmann, PhD who holds the position of VP, Materials Engineering at both companies. His team developed the custom aluminum alloy. Elon Musk has reported that the giga press process to die-cast both the front and rear Model Y underbody segments as single parts eliminates 600 robots combined. That saves time and cost and improves quality. That is innovation at the margin.
Linking the Tesla Safety Score, Tesla FSD and Tesla Insurance leads to lower total cost of ownership through Tesla’s vertical integration. Tesla will lead this pursuit with its commitment to advance all of its technologies with the top engineering talent today across all industries. As the proportion of driving becomes more autonomous, Tesla absorbs more of the insurance burden for transportation while concurrently reducing the rate of vehicle accidents and fatalities if Tesla FSD realizes its objective to be 10x better than a human driver. That lowers the cost of insurance to the driver. And Tesla Insurance lowers the premiums paid as the Tesla Safety Score improves driving behavior and safety when the driver is operating the vehicle. Innovating at the margin expands the scope of what is possible to accelerate the transition to sustainable energy by aggregating Tesla FSD and Tesla Insurance into a complementary, yet comprehensive direct-to-customer experience using the Tesla Safety Score as well.
Conclusion
Automobile Insurance is a problem for safer drivers because they subsidize claims paid out for the least safe drivers who do get into accidents. And this problem will be even greater in the future for Tesla drivers who benefit from continuous innovation to reduce automobile accidents in Tesla vehicles. But through vertical integration across an entire supply chain, technology stack, innovation pillars and top engineering talent, Tesla can offer its own auto insurance product to lower the total cost of ownership for its vehicles with customers.
This combined Tesla FSD, Tesla Insurance and Tesla Safety Score offering is highly disruptive to legacy vehicle manufacturers and insurance companies who cannot provide the same degree of innovation within their own industries unless they partner is some creative ways across these industries. Tesla’s vertical integration and mission to accelerate the transition to sustainable energy are key contributors to its competitive advantage. If you outsource everything as a legacy vehicle manufacturer except: 1.) your brand, 2.) manufacturing internal combustion engines and 3.) final vehicle assembly, how do you compete with Tesla’s direct-to-consumer business model over the next decade once the lithium ion battery supply chain evolves to provide as many batteries as Tesla needs to meet consumer demand? But Tesla is not waiting. They are also sourcing their own raw materials to make their own 4680 lithium ion battery cells for Tesla vehicles.
And Tesla’s actual gross margins are improving as well. The most recent Q3 earnings report disclosed vehicle gross margins improved to 30.5% even with supply chain challenges. This is due to multiple factors including lower input costs for vehicles manufactured in Shanghai and lower unit production costs with continuous innovation. And Tesla continues to sell more high margin add-on services to a growing base of customers through its mobile app and website. This improves overall margins and earnings per share. Demand for Tesla vehicles is also accelerating. Tesla has increased the sales price multiple times over the last year due to this demand.
Tesla’s market cap passed $1 trillion this week for the first time. That milestone has only been accomplished to date by a limited number of companies including Alphabet (previously Google), Apple, Amazon, Facebook, Microsoft and now Tesla. However, Tesla’s accomplishment is not due just to its current financial performance. Speculation on earnings far into the future are also playing a role. Hertz also announced this week it ordered 100,000 Tesla Model 3 vehicles to start the transition of its rental fleet to electric vehicles. Another update with the CEOs of Hertz and Uber later in the week stated that order could increase to 200,000 Tesla electric vehicles with 150,000 rented to Uber drivers. That could evolve later into Robotaxis to provide Hertz and Uber a strategy to partner in that transition, too. Elon Musk confirmed no vehicle discounts were provided on this multi-billion dollar vehicle sale.
But Tesla’s 2021 revenue is still a small blip compared to the aggregate revenue of the legacy auto manufacturers. Tesla’s market cap also exceeds the market cap for this same group of the auto manufacturers. I’m not recommending a stock. I’m writing about a company as an innovation leader that continues to innovate at a pace far greater than industry norms. But those are very different topics. With eroding speculation or other reasons, Tesla’s market cap could easily fall 50% or more in a matter of weeks.
For full-disclosure, I follow Tesla with a primary interest in their products, strategies and innovation. As a retail investor, I’m long ABNB, AMZN, GOOG, LMND and TSLA mentioned in this update. Nothing in this post is intended to serve as financial advice. Do your own research.
Best,
Stephen