.jpeg)
Sander Meijer runs a mid sized used car operation in Amersfoort, about half an hour east of Utrecht, where he's been handling maybe fifteen to twenty vehicles a week since expanding into EVs in late 2023. He'd been buying Teslas fairly aggressively through the Dutch remarketing channels, mostly Model 3s from lease returns, because the margins were decent and demand from private buyers in the Randstad was steady enough that he rarely sat on one for more than two weeks. In January, he took in a 2021 Model 3 Long Range from a fleet disposal company in Rotterdam, listed at 74000 kilometers with what looked like a clean service file and a reasonable asking price of around 27000 euros. The car drove fine on a twenty minute test loop around the lot, but when Meijer plugged in an OBD diagnostic tool to check the battery’s state of health before listing it, the reading came back at 81%, which, for a car with that kind of mileage and age, is significantly worse than what you’d expect. Tesla’s own 2023 Impact Report showed average degradation of about 15% after 200000 miles for Model 3 and Model Y Long Range variants, so a car at 74000 kilometers showing 19% capacity loss was either a statistical outlier or, more likely, a vehicle that had been hammered with DC fast charging and possibly spent long stretches parked at high states of charge in a warm climate before arriving in the Netherlands.
Meijer said the fleet company couldn’t or wouldn’t provide any charging history, which is the whole problem with EVs right now from a trade perspective. The odometer on a combustion car tells you something imperfect but broadly useful about how much wear the drivetrain has accumulated, and dealers have been reading that signal for decades. On an EV, the odometer tells you almost nothing about the most expensive component in the vehicle. A Tesla that’s done 150000 kilometers of mostly home charging at moderate states of charge on an NCA pack can easily retain 93 or 94% of its original capacity, and I’ve seen data from Recurrent and others confirming this pattern across thousands of vehicles. The same car with the same mileage that spent two years as an airport shuttle getting Supercharged to 100% three times a day could be sitting at 82% or worse, and from the outside, looking at the listing, those two cars are identical. The Swedish used car marketplace Kvdbil ran a study on over 1300 electric and plug in hybrid vehicles and found that eight out of ten retained more than 90% of original battery capacity. The other 20% is where things get messy, because those are the cars that end up repriced or, in some cases, the cars that don’t get tested at all and land on somebody’s lot with a mileage figure that suggests a healthy pack when the reality is very different.
The Netherlands is probably the sharpest test case for this in Europe right now. BEV registrations hit 156000 in 2025, roughly 40% of all new passenger car sales, and the country already has over 209000 charging points, which means the density of used EVs hitting the secondary market is only going to accelerate as lease returns from 2022 and 2023 start flowing through. Arval Trading flagged battery health diagnostics as one of the key emerging metrics in Dutch used EV appraisals earlier this year, alongside charging history and real world range, and that tracks with what I’m hearing from dealers across the Benelux. A vehicle data analyst at carVertical mentioned that the gap between what mileage implies about an EV’s condition and what the battery actually shows is becoming one of the biggest pricing risks in the European used car market, particularly for Teslas and early Volkswagen ID models, where charging behavior varies wildly between owners. With odometer fraud on combustion cars, you can at least pull TÜV or APK inspection records and cross reference them against a vehicle history report to spot discrepancies, but for battery health, there’s nothing like that. No standardized log of how an EV’s pack has been treated, no charging history that follows the car across owners, and the data that does exist is locked up in manufacturer systems that dealers can’t get at when they’re appraising a trade in.
AVILOO, out of Austria, has probably done more than anyone to make battery testing practical for dealers, and their FLASH Test, which is basically an OBD dongle that spits out a state of health number in three minutes, is showing up at more and more dealerships across Europe. They ran a remarketing study in 2024, and the numbers were interesting, frankly better than I would have guessed. Buyers said they’d pay somewhere between 550 and 1100 euros extra for a used EV with an independent battery certificate, and the certified cars moved off lots about 36% faster, which adds up fast if you’re running any kind of volume. AVILOO has been hammering this point for a while, that kilometers and years on their own don’t tell you what you need to know about a pack’s actual condition, and frankly, the recent large scale studies keep proving them right. One study from earlier this year looked at thousands of EVs and came back with an average state of health of about 95%. The spread within the sample was enormous, though, and got wider the older the cars were. Cars in the four to five year old bracket showed the bottom quarter averaging around 91.6% and the top at 96.5%, already a spread that matters when you’re pricing them, but once you looked at the eight to twelve year cohort, the worst performers were down around 82% while the best were still sitting at 90%. If you’re a dealer buying blind, that eight percentage point spread at the same age is where the 3000 euro pricing mistake lives, and almost nobody is testing for it systematically.
Meijer listed the Model 3 with the actual battery health data right there in the ad, knocked about 3500 euros off the asking price to reflect the pack’s condition, and it sold within ten days to someone in Eindhoven who didn’t care much about the reduced range because his daily commute was 40 kilometers and he charged at home. The car was worth less than what Meijer paid for it, though, and he told me the loss was somewhere around 2000 euros once you factored in the APK inspection and reconditioning costs. What bothered him, he said, wasn’t really the money, it was that without the OBD tool he would have listed it at full price and some buyer in Amersfoort would have been stuck with a pack that couldn’t deliver what the mileage suggested, and that most dealers he talks to in the Utrecht area still don’t bother testing because they haven’t worked out how to account for battery risk in their pricing or because they figure the odometer reading is close enough. Mileage as a proxy for battery condition is basically useless on its own, and enough data is floating around now to prove that to anyone willing to look at it, but old habits in the car trade die slowly.
Nobody seems sure whether mandatory battery health reporting at resale is actually coming, or if the trade just keeps eating losses on degraded packs for another few years while regulators figure out what they want. Belgium’s Car Pass showed years ago that if you make mileage documentation mandatory, odometer fraud basically disappears, and you could make a similar case for battery health at the point of EV resale. Polestar has been doing battery health certificates on used Polestar 2s since 2025, Hyundai rolled AVILOO into its German dealer network, and there are maybe half a dozen independent testing outfits now competing for the same space across Europe. There’s no EU directive requiring any of this at the point of sale, and given how long Directive 2014/45 took to produce any real cross border mileage sharing, a standardized battery passport before 2028 feels optimistic. Meijer said he’s budgeting for an AVILOO unit this quarter and plans to test every EV that comes through his lot, which, at current volumes, means maybe eight or ten tests a week. A couple of his competitors in the Utrecht area have apparently started doing something similar, though one of them told Meijer he only tests Teslas because the ID.3 and ID.4 packs haven’t given him trouble yet, which is a small sample size to be making that call from.
