The new Hertz knows bold moves. On the heels of Monday’s big news on the partnership with Tesla to supply 100,000 Model 3s to Hertz’s rental fleet by 2022, we now know that close to half of those vehicles will be rented to Uber drivers.

(Hertz’s news release does not explicitly tie the 100k Tesla order with the Uber supply. And again, there are caveats: “up to” 50k units and both figures are dependent on “factors outside of its control, such as semiconductor chip shortages or other constraints.”)

Beyond the headline, how will this work operationally?

For guidance, let’s take a step back. About 10 years ago, it became clear to Uber and Lyft that growth was being hindered by potential drivers’ lack of access to four wheels. Hence the creation of Uber’s internal Xchange leasing program and third-party supplier programs.

We know what famously happened to Xchange — Uber shut down the program after it lost an average of $9,000 per vehicle. Renting or leasing to TNC drivers is a tricky business. It’s based on heavy use, higher maintenance costs, and more frequent insurance claims. (Fair, who bought the Xchange business in late 2017, eventually shuttered its own TNC rental program about two years later, citing a rise in insurance premiums.)

But then a workable formula emerged: Rent used midsize sedans with the heavy depreciation already taken out of them — and the holding costs made sense. Hertz started figuring this out in 2016, cascading cars from their rental fleet to weekly TNC rentals. Hertz earned nearly $300 million in TNC revenue in 2018 on 42,000 cars dedicated to that market.

However, reinvigorating the TNC rental market with new vehicles — and with an “electric, entry-level luxury sedan” Model 3 at that — requires a new magic formula.

In this program, participating drivers receive a preferred weekly rate for Hertz rentals, which includes insurance, basic maintenance, and unlimited miles. What is the bundled price point that makes sense for both driver and Hertz? At least for Uber, hurting for drivers once again, the Model 3 will be a draw. (Uber is also incentivizing the experience by offering drivers a “zero-emissions incentive” of $1 a ride for using EVs and 50 cents for every rider.)

The stakes are higher as the Model 3 starts at a steeper initial cost. Hertz will self-insure these Teslas, though insurance in general is substantially higher for a Model 3 than, say, a TNC workhorse such as a Toyota Camry. Yes, the standard maintenance will be less for the Teslas, but the bodywork will be more expensive.

Range anxiety won’t be an issue. But the majority of drivers — the ones not living in single-family dwellings — will rely on public charging. (Hertz is partnering with EVgo on this.) The inability to charge overnight will present some logistical challenges for drivers.

Let’s be clear: Using EVs in high-mileage applications makes sense. European taxi fleets run Model 3s, and that was part of the business model of Tesloop. But those models don’t include a middleman trying to make money on top of his payments.

A prevailing concept on holding costs for EVs in fleets is to run them longer to flatten the depreciation curve. I rode in Tesloop’s Model S, dubbed “eHawk,” which had more than 400,000 miles at the time. It was still a premium experience — and the depreciation was taken out of that EV long before. In the Hertz/Uber model, we’ll be able to see how this plays out on a much larger scale.

Our Autonomous Future

The long play to this partnership has implications for autonomous transportation.

This also was part of Tesloop’s end goal. During my Tesloop trip from Los Angeles to San Diego, the driver used Full Self-Driving Mode (FSD) for the entirety of the freeway portion of the trip (which was about 90% of it). Perhaps Hertz won’t fleet Model 3s with FSD to start, but when they have confidence to do that, the data Hertz and Uber will collect will be invaluable.

This type of partnership creates a pattern for a supplier and transportation provider relationship in an autonomous environment — ultimately, just take away the driver. Uber is unlikely to ever own the metal, that’s Hertz’s job. And fleet management, Hertz’s core competency, is becoming known to the wider world as the backbone to an autonomous service. Yet Uber is still the consumer-facing service in this equation, and Hertz is the back-of-house supplier and fleet manager.

Sixt’s partnership with Mobileye on a robo-taxi service shows that car rental hasn’t given up on being the consumer-facing service provider of the future.

There are many questions to be answered in the short and long term as the model settles into the right parameters for all parties. Will those parameters work well enough for Hertz and Uber to fleet 50,000 units to TNC drivers? There will be keen interest to see how a new EV rental model for TNC drivers works out.