Lessons from 10,000 Electric Cars and 100 Million Miles

This time last year, the EV world was very different. Ford had not yet shipped a single F-150 Lightning. Hertz had not yet ordered 100,000 Teslas or 175,000 EVs from General Motors. Gas prices in the US had not yet soared above $5. Build Back Better had not yet become the Inflation Reduction Act with tax credits for both new and used electric vehicles. 

And twelve months ago, Recurrent was wrapping its first year as a company, with 5,000 community members, a lot of early learnings about batteries and range, and some mixed signals on EV adoption in the US. 

A lot can change in one year. 

Recurrent has just passed our second anniversary. We’ve more than doubled our community, with over 11,000 EV drivers sharing data and receiving insights on their battery and range with Recurrent’s monthly reports. We’re now drawing analytical insights from over 100 million miles across 50+ different makes and models of battery electric and plug-in hybrid vehicles. So, what have we learned this year? 

1. EV odometers still don’t matter. 

Odometers are the rule of thumb that has always been used to determine the value of a used car - but it simply doesn’t make sense with electric cars. We see “low mileage” used EVs with batteries that are substantially worn down compared to their original state. We’re seeing “high mileage” cars that have been carefully charged, or had their batteries replaced along the way, with real-world range similar to what they had initially. 

We published research in April about how manufacture date is a better indicator (but still far from accurate) of remaining battery capacity than model year. We’ve watched as battery replacements have rolled out across nearly all Chevy Bolts over the last year. The average maximum range for a replaced-battery Bolt has jumped by 13%, based on over 1600 Chevy Bolts in our connected community. And the same is true for a Bolt with 50,000. miles as it is for one with 5,000.

2. Used EV values are getting smarter. 

Used car buyers have begun to figure out that the value of a used EV should be determined by its battery health, not its odometer reading. This is visible at all stages of the car buying process. 

One of the largest wholesale used auto auctions in North America, ADESA, noted a $4000 - $7000 increase in sale price and a 7% to 9% increase in vehicle sales rate when Recurrent provided information about the EV and its battery. 

That wholesale premium matches up with what Recurrent has observed in our retail shopping research. Potential used EV buyers are willing to pay a $3000 (or higher) premium for vehicles with batteries in great condition, even if they have high mileage. Buyers were even willing to pay a premium for a used EV with a battery in “fair” condition when compared to having no information about that battery.

Our partnership with Black Book, one of the leading valuation services, shows an early shift in how the auto industry is rethinking used EV prices. Owners of EVs who are considering selling can now get a vehicle-specific valuation from Black Book along with a recommended battery premium from Recurrent based on their battery’s condition. And Recurrent will help them find a buyer, too.

3. EPA range estimates are confusing. 

We’re not sure how the EPA got the job to do range testing for EVs – it seems to have been a natural scope creep from MPG ratings to MPGe to EPA range. The problem is that EPA range accuracy leaves something to be desired. Recent Consumer Reports testing on a variety of new EVs showed that EPA range estimates are 10-70 miles off from real-world driving tests in mild and warm weather, with even bigger differences (over 100 miles, in some cases) in cold weather driving. 

Our own studies from last winter, involving 1000’s of vehicles from our subscribers in real-world driving, showed up to 40% range loss in cold temperatures, a figure that varies substantially from model to model. 

As a result of the inaccuracies in original EPA range estimates, one of the major efforts that our data science team tackled this past year was coming up with a predictive model to determine the “Range When New” for every EV make, model, year and trim level, derived from data from subscribing vehicles. We use that number internally, rather than the EPA range. For instance, it is the denominator for the Recurrent Range Score, so that we’re showing the range of every used EV vs. how far it would have gone when it was new.

But in a nod towards the ubiquity of the inaccurate “EPA Range” figure, we still show the EPA number on all our reports as a point of reference (see example here).

4. New EV models are accelerating adoption. 

It’s not just Tesla anymore. We started Recurrent in 2020, the year that, in retrospect, may have represented Tesla’s high-water mark in terms of market share. In 2020, 63% of all new EVs were Tesla, before this number fell to 53% in 2021 and 48% in the first half of 2022. (Note: EVs in this context include battery electric and plug-in hybrid vehicles, but not mild hybrids.)

Dozens of new EV models have launched in 2021 and 2022 with just as many planned for 2023. This year we've added a thousand connected drivers just within some of the newest EV models on the road: Hyundai Ioniq 5, Rivian R1T, Ford Mach-E and F-150 Lightning. A wider selection of new vehicles and price points is bringing in new first-time EV drivers, as is a maturing used market.

The EV market analysis team at Boston Consulting Group factors in model selection as one of the major inputs to their adoption forecast. The accelerating growth that they are forecasting is at least partially reflective of all of these new models coming on the scene. 

5. High gas prices drive people to electric vehicles.

In March, when US gas prices reached all-time highs, I looked to the past to predict the future. At the time, I said that EV and hybrid sales would surge as a result of historic gas prices. That’s exactly what happened:

The one thing I missed predicting at the time was an even bigger milestone: plug-in model sales surpassing conventional hybrid sales. With the brief exception of a few months in 2018 when the initial Tesla Model 3s were delivered, EV (fully battery + plug-in hybrid) sales have never exceeded regular hybrid sales. We quietly slipped past that incredibly historical milestone, likely for good, this summer.

That shift was another notable change in BCG’s EV forecast from 2021 to 2022 – last year, they predicted a 2:1 ratio of hybrid to EV sales in the US in the year 2025. In this year's forecast, the overall % adoption of both had grown, but the ratio of hybrid to EV sales was 1:1 for the year 2025… in other words, what we’re already seeing today. So it's not too hard to make another bold prediction: that next year’s BCG 2025 forecast will have that ratio skewing even more in favor of EVs.

OK, that’s a wrap for our second full year in business. I’m so thankful for every member of our team, all of whom have chosen to take a bet with their careers to be here at the ground floor of the used car industry’s transition to electric vehicles. I’m grateful to our investors, who believe in us and in this transition enough to make meaningful bets on the future. And thank you also to every EV owner who has joined us on this mission to electrify transportation.

‍