The charts are drawn wide: turn your phone sideways for a better view.

Carmaker Productivity2021 → 2025

All analyses Automotive Value Map LinkedIn

Automotive industry | October 2026

How many cars each employee builds tells you a carmaker’s class; how much each employee earns tells you whether that class pays

I looked at 41 carmakers over five years, 2021 to 2025, to see whether each class has a normal level of output and profit per employee, who sits off it, and what that says about the job cuts now announced across the industry.

How to read this

  • This is a diagnostic model built from public reports. It shows where a company stands against the norm of its class, not how many people it should employ.
  • Output per employee depends on how much a company makes itself and how much it buys from suppliers or contractors, so a gap says where to look, not which functions or processes cause it.
  • The distance to the median can close through sales as well as through headcount. Deciding which requires the process-level view that annual reports do not show.

Executive summary

Situation

Between 2021 and 2025 the industry went through a whole economic cycle: post-Covid shortages, record profits in 2023 and a sharp fall in 2025 (exhibit 1). Any single year would give a distorted picture, so every number here is a five-year view.

Complication

Volkswagen, JLR, Bentley, Aston Martin and many others are cutting jobs. Cuts get discussed as if they cure everything, but without knowing what is normal for a class you cannot tell whether a company has too many people or simply sells its cars too cheaply.

Key question

Is there a normal level of output and profit per employee for each class, and what does being above or below it say about where a company goes next?

Exhibit 1

Over five years the industry went through a full cycle: from shortages and record prices to a downturn under tariffs and China

Key events of 2020–2025 and their effect on carmakers’ output and profit.

2020–2021

Covid and chip shortage

Lockdowns stopped plants, then the semiconductor shortage cut output by millions of cars. Demand came back faster than supply.

Fewer cars, higher prices, no discounts

2022

Inflation and shocks

Sanctions and carmakers’ exit from the Russian market, a spike in energy and materials prices, the first interest rate rises. The chip shortage still kept prices high.

Revenue grows on price

2023

Peak

Chip supply recovered and pent-up demand met high prices. The industry built more cars before prices came down.

Record profits almost everywhere

2024

Normalisation

Inventories refilled and discounts returned. Expensive credit cooled demand, China went into a price war and EV demand slowed.

Profits decline

2025

Downturn

US tariffs on imported cars and parts, write-downs on EV programmes, foreign brands losing share in China.

Profits fall, volumes hold

Source: summary of company reports and industry events 2020–2025.

  1. Output per employee has a norm; profit per employee does not. A volume maker builds about 24.6 cars per employee a year, a premium maker about 16.0 and a luxury maker about 2.8, and these numbers barely move between companies or from year to year. Profit per employee behaves very differently: it varies several times over within a class and follows the cycle, rising when cars are scarce and falling when discounts come back.
  2. Premium gets paid less and less for building fewer cars. The median premium maker earns $67k per employee, about one and a half times the volume median ($44k), but that figure has moved by −36% since 2023. In 2025 Mercedes-Benz and Audi earned less per employee than GM. Only luxury still holds a clear lead over volume ($139k).
  3. The industry got more productive and poorer at the same time. In volume, median output per employee changed by +22% over 2021–2025, while profit per employee changed by −14% since 2023; in the mid-market profit per employee changed by −62%. The plants keep getting better, and the pressure comes through prices and costs.
  4. Comparing each company with its own class gives four archetypes, and each needs a different agenda (exhibit 4). Cutting headcount is the right main lever only for “structural excess”: Volkswagen, Honda’s car business, JLR, Aston Martin, Great Wall and, borderline, Renault. For the “price trap” (Nissan, Mazda, Volvo, Bentley) the levers are product, price and markets.
  5. Five carmakers build well below the median output per employee of their class, and only Volkswagen has announced cuts that would bring it close. Volkswagen, Great Wall, JLR, Honda’s car business and Aston Martin build between 16% and 55% fewer cars per employee than the median of their class, and they also earn below it. With the cuts they have announced, Volkswagen would still be about 7% short and Aston Martin 13% short; JLR would go from 55% to 51% short, and Great Wall and Honda have announced nothing (exhibit 26). Renault is in structural excess on its five-year average but was within 10% of the median in 2025. The distance to the median is a diagnosis, not a target for cuts: it closes through sales as well as through headcount.
  6. The biggest risk over the next two years sits with companies caught between classes. Porsche, Mercedes-Benz, Audi and JLR carry the costs of a higher class without earning its profit. On five-year averages Porsche and Audi still look like class benchmarks, but in 2025 their profit per employee fell further than their peers’.
  7. New EV makers went two different ways. XPeng and Leapmotor reached the mid-market output norm (about 21 cars per employee) in four years, and Leapmotor already breaks even. Rivian, Polestar and Lucid still ran operating margins of −66%, −66% and −259% in 2025, a long way from the economics of their class.
  8. Whether a company ends up in structural excess is decided when it designs its organisation, long before anyone announces layoffs. Pick the class on purpose, measure the processes that drive the norms, staff for an average year instead of the peak, put AI on processes that are already clear, share what customers cannot see and earn the price with what they can (see “What to do”).
The industryClass norms

The higher the class, the fewer cars each employee builds, and the more profit the company has to make per employee to pay for itThe class norm is the median of 2021–2025 averages across mature companies in the class. New EV entrants and companies with non-comparable headcount are excluded.

Exhibit 2

Moving up a class cuts output per employee by 1.5–6 times; profit per employee is 1.5 times higher in premium and 3 times higher in luxury

Median of mature companies’ 2021–2025 averages; under each figure the range within the class.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 3

Classes form distinct clusters on output per employee, while profit within a cluster differs several-fold

Averages for 2021–2025, mature companies and the three with non-comparable headcount. Horizontal: vehicles per employee per year, log scale. Vertical: operating profit per employee, $k. New EV entrants are shown in their class sections instead (see Methodology).

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 4

Position against its own class norm splits carmakers into four archetypes with different agendas

Class norm = median of 2021–2025 averages across mature companies in the class, set to 100. Each dot: the company’s average as a percentage of its class norm. Archetype threshold 90%. New EV entrants and Rolls-Royce not shown.

Class benchmark

Output and profit per employee at or above 90% of the class norm.

Pricing power

Output below 90% of the norm, profit at or above: brand and price pay for low output.

Price trap

Output at or above 90% of the norm, profit below: price and mix do not cover costs.

Structural excess

Both output and profit per employee below 90% of the class norm.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Class 1 of 4Volume

Volume makers build a fifth more cars per employee than in 2021, yet earn less per employee than at the 2023 peak, with Chinese price pressure, tariffs and write-downs taking the difference11 companies: Toyota, Volkswagen, Stellantis, GM, Ford, Nissan, Renault, Suzuki, Hyundai, Honda (car business only), Chery. Hyundai and Chery have data for 2024–2025.

Exhibit 5

Almost every volume maker raised output per employee; Volkswagen is still at two-thirds of the group level

Vehicles per employee per year, mature companies, 2021–2025. Shared scale across panels. Dashed: group median.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 6

Profit per employee differs several times over and has fallen for most since 2023

Operating profit per employee, $k, mature companies, 2021–2025. Dashed: group median.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 7

Period averages and trend direction

Average over the years with data; trend from the first available year to the last. For new EV entrants the profit columns show operating margin.

CompanyYears in averageVehicles / employee, avgTrendProfit, $k, avgTrend

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Key observations

  • Plants got more productive while profit went the other way. Median output rose from 21.3 cars per employee in 2021 to 26.0 in 2025. Median profit per employee peaked at $47k in 2023 and fell to $41k in 2025, as price and mix stopped covering rising costs.
  • GM, Toyota and Hyundai set the bar. GM kept profit between $75k and $92k per employee in every one of the five years. Toyota’s car business went from 24.9 to 27.9 cars per employee. Hyundai, with data only for 2024–2025, shows 33.2 cars and $74k per employee, one of the best combinations in the group.
  • Stellantis and Nissan have a price problem rather than a people problem. Stellantis built roughly the same 21–24 cars per employee every year, yet its profit per employee went from $102k in 2023 to $−4k in 2025. Nissan builds at the norm (about 25) and barely earns anything ($3k in 2025). Cutting staff here treats the symptom, while the cause sits in product, pricing and markets.
  • Volkswagen does have a people problem. To compare like with like, I took only its volume brands: no trucks, no financial services, no Audi and no Porsche. Even on that basis Volkswagen builds about 16.1 cars per employee, two-thirds of the group norm, and earns $9k per employee, about a fifth of it. Neither number improved in five years. This is structural excess, and it explains the German cuts.
  • Honda’s car business barely makes money. Taken without motorcycles, it builds 19.3 cars per employee, below the group norm, and earned $−4k per employee on average ($−71k in 2025/26 after EV write-downs). Motorcycles bring in most of the group’s profit.
  • The usual answer to Chinese pressure is to build cheaper, and more and more often with Chinese partners. Chinese makers have reset the price level in China, in emerging markets and increasingly in Europe. Volkswagen works with XPeng, Stellantis with Leapmotor, Toyota with BYD, and Renault developed the new Twingo with a Chinese engineering centre in about two years. In the US the story is different: GM and Ford lost profit to tariffs on parts and to EV write-downs rather than to Chinese competitors.
  • Suzuki and Chery show how far simple cars at scale can go. They build 42.5 and 38.8 cars per employee, the highest in the group, with profit close to the median. Suzuki is the only company in the group where both numbers rose over the five years.
Class 2 of 4Mid-market

The mid-market works at volume-level productivity without volume-level scale, and lost more than half of its profit per employee in 20259 companies: Mazda, Subaru, Mitsubishi, Tesla, Škoda, SEAT/Cupra, Great Wall and new EV entrants XPeng and Leapmotor. Škoda and SEAT headcount found for two years only.

Exhibit 8

Mazda, Subaru, Mitsubishi and Škoda work at volume-level productivity; Tesla and Great Wall well below

Vehicles per employee per year, mature companies, 2021–2025. Shared scale across panels. Dashed: group median.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 9

In 2025 profit per employee fell for everyone except Škoda

Operating profit per employee, $k, mature companies, 2021–2025. Dashed: group median.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 10

XPeng and Leapmotor reached the mid-market output norm within four years

Vehicles per employee per year, new EV entrants.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 11

Operating margins are closing in on break-even, and Leapmotor got there in 2025

Operating margin, % of revenue, new EV entrants. Launch years with fewer than 1,000 deliveries omitted.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 12

Period averages and trend direction

Average over the years with data; trend from the first available year to the last. For new EV entrants the profit columns show operating margin.

CompanyYears in averageVehicles / employee, avgTrendProfit, $k, avgTrend

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Key observations

  • These makers have volume productivity but not volume scale. Mazda, Subaru, Mitsubishi and Škoda build 23–31 cars per employee, as many as the volume giants. Yet each sells a fraction of their volume and gets no premium price, so profit holds up only in good years.
  • US tariffs took the Japanese trio’s profit. In 2025 Mazda earned $7k per employee, Subaru $7k and Mitsubishi $18k. All three ship cars from Japan to the US.
  • Škoda is the exception on profit, with $83k per employee in 2025 at the same productivity as the Japanese. It runs on Volkswagen Group platforms, purchasing and engineering, so it gets group scale without carrying the cost of building its own.
  • Great Wall builds only 14.3 cars per employee and earns about $16k. Part of the gap comes from making many components in-house, part from the price war in China.
  • Tesla is drifting away from premium economics. Output per employee rose (9.4 → 12.1), while profit per employee dropped from $107k in 2022 to $32k in 2025.
  • Chinese EV newcomers reached the norm faster than anyone. XPeng went from 7.0 to 21.6 cars per employee and Leapmotor from 13.3 to 20.7, which is the class norm. Leapmotor broke even at operating level in 2025, and XPeng’s operating margin improved from −36% to −4%.
Class 3 of 4Premium

Premium kept its productivity but lost a third of its profit per employee in two years, and is now closer to volume makers than ever10 companies: BMW, Mercedes-Benz, Audi, Volvo Cars, JLR and new EV entrants Polestar, NIO, Li Auto, Zeekr, Rivian.

Exhibit 13

Mature premium productivity is stable; JLR builds half as many cars per employee as its peers

Vehicles per employee per year, mature companies, 2021–2025. Shared scale across panels. Dashed: group median.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 14

Since the 2023 peak, profit per employee has fallen at every premium maker

Operating profit per employee, $k, mature companies, 2021–2025. Dashed: group median.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 15

Output per employee at premium EV entrants depends on how much they build themselves

Vehicles per employee per year, new EV entrants.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 16

Only Li Auto reached a positive operating margin, while Polestar and Rivian still lose two-thirds of their revenue

Operating margin, % of revenue, new EV entrants. Launch years with fewer than 1,000 deliveries omitted.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 17

Period averages and trend direction

Average over the years with data; trend from the first available year to the last. For new EV entrants the profit columns show operating margin.

CompanyYears in averageVehicles / employee, avgTrendProfit, $k, avgTrend

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Key observations

  • Premium productivity held steady while profit shrank. Mature premium brands built 14–16 cars per employee in every one of the five years. Over the same period median profit per employee fell from $78k in 2023 to $50k in 2025.
  • The German three are losing their lead over volume makers. Between 2023 and 2025 Mercedes-Benz went from $130k to $57k per employee, BMW from $129k to $74k and Audi from $78k to $43k. In 2025 Mercedes and Audi earned less per employee than GM. Most of this comes from China, where German premium brands are losing share and pricing.
  • Volvo charges premium prices without earning a premium margin. It builds about 16.4 cars per employee, like BMW, but its profit of about $49k per employee is at volume level.
  • On productivity JLR has already left premium. It builds 8.3 cars per employee, half of what its peers do. In a good year ($71k in 2023/24) Range Rover covers the gap, but any shock, like the cyber attack and tariffs in 2025/26, takes profit down to $5k.
  • Of five premium EV newcomers, only Li Auto has reached the economics of the class. Its operating margin was +6% in 2023. In 2025 NIO ran at −16%, Rivian at −66% and Polestar at −66%. Polestar’s high output per employee flatters it: it has no plants of its own, and Geely and Volvo build its cars.
Class 4 of 4Luxury

In luxury the brand sets the profit and productivity hardly matters: by 2025 the leader earned $500k more per employee than the laggard8 companies: Porsche, Ferrari, Lamborghini, Bentley, Aston Martin, Rolls-Royce and new EV entrants Lucid and Lotus. Rolls-Royce does not disclose profit. McLaren and Morgan are excluded because their accounts for these years are only partly available.

Exhibit 18

Pure luxury builds 2–4 cars per employee; Porsche sits in between

Vehicles per employee per year, mature companies, 2021–2025. Shared scale across panels. Dashed: group median.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 19

Ferrari keeps raising profit per employee, most others are losing ground

Operating profit per employee, $k, mature companies, 2021–2025. Dashed: group median.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 20

Lucid’s output per employee is already close to the luxury norm

Vehicles per employee per year, new EV entrants.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 21

Lucid and Lotus still spend far more than they sell, with operating margins deep below zero

Operating margin, % of revenue, new EV entrants. Launch years with fewer than 1,000 deliveries omitted.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 22

Period averages and trend direction

Average over the years with data; trend from the first available year to the last. For new EV entrants the profit columns show operating margin.

CompanyYears in averageVehicles / employee, avgTrendProfit, $k, avgTrend

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Key observations

  • In luxury, productivity barely moves and explains almost nothing. Pure luxury makers build 2–4 cars per employee a year. Ferrari (2.5) and Aston Martin (2.2) are almost identical on output, yet in 2025 Ferrari earned $416k per employee and Aston Martin $−89k, a loss.
  • Ferrari is the only one that grows year after year, from $276k to $416k per employee on the same ~2.5 cars. All of that growth comes from price, personalisation and keeping cars scarce.
  • Porsche is stuck between premium and luxury. Its output per employee fell (8.2 → 6.7) because sales dropped while headcount kept rising. Profit per employee collapsed from $187k in 2023 to $11k in 2025. Porsche now carries luxury costs on premium revenue.
  • Lamborghini and Bentley are moving apart. Lamborghini holds $245–335k per employee, while Bentley fell from $115k to $61k and is cutting staff in 2026.
  • The luxury EV newcomers are nowhere near the economics of the class. Lucid already builds 1.8 cars per employee, close to the luxury norm, but its operating margin in 2025 was −259%: it spent about $3.6 for every dollar it took in. Lotus Technology ran at −82%.
Outside the normsNot comparable

Three makers need their own lens, because a different headcount perimeter hides a different business modelKia, BYD, Geely. Their headcount is reported on a different perimeter, so they are left out of the norms. Trends within each company still say something.

Exhibit 23

Absolute values are not comparable with the other groups because of the headcount perimeter

Vehicles per employee per year, 2021–2025. Shared scale across panels.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 24

Trends within each company are still worth reading

Operating profit per employee, $k, 2021–2025.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Exhibit 25

Period averages and trend direction

Average over the years with data; trend from the first available year to the last. For new EV entrants the profit columns show operating margin.

CompanyYears in averageVehicles / employee, avgTrendProfit, $k, avgTrend

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Key observations

  • Kia publishes only its Korean headcount, so the absolute values are overstated. The trend is still telling: profit per Korean employee rose from $125k to a peak of $260k in 2024, the best trajectory among volume brands.
  • BYD runs a different business model, with 4.7 cars per employee and about $6k of profit. Its headcount includes batteries, chips and contract electronics. To compare BYD’s car business fairly, subtract the headcount of the separately listed BYD Electronic and use automotive segment revenue (see Methodology).
  • Geely Auto leaves out some parent-group plants that build its cars, so its output per employee is overstated (35.8).
Who cuts and whyJob cuts

Almost everyone is cutting jobs, but only some of them are fixing the actual causeThe archetypes separate two cases: too many people, where cuts work, and a price or product problem, where cuts only buy time.

Exhibit 26

Five carmakers build between 16% and 55% fewer cars per employee than the median of their class; only Volkswagen’s announced cuts would bring it close

Companies in structural excess. How far output per employee in the latest year falls short of the class median. Filled circle: today. Open circle: the same output divided by headcount after the cuts each company has announced. Volkswagen on its volume-brand perimeter, with cuts counted without those at Audi and Porsche; about half of its cuts have no deadline. Renault, in structural excess on its five-year average, was within 10% of the median in 2025 and is not shown.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Key observations

  • Only Volkswagen’s announced cuts match the size of its distance to the median. Its volume brands build 29% fewer cars per employee than the volume median; if the announced cuts land in these brands, the shortfall narrows to about 7%, although about half of them have no deadline[1][2][3]. Aston Martin’s cut of up to a fifth of its staff narrows its shortfall against the luxury median from 31% to about 13%.
  • JLR is furthest from its class. It builds 55% fewer cars per employee than the premium median, and its first round of 4,000 roles narrows that to 51%[4][5]. Great Wall is 42% short of the mid-market median, partly because it makes many of its own components, the same effect that keeps BYD out of the comparison. Honda’s car business is 16% short; Renault, borderline on its five-year average, was only 6% short in 2025.
  • Selling more cars closes the distance too. To reach the premium median with its post-cut workforce, JLR would have to sell about twice as many cars as today; Great Wall would need about 1.7 times its volume. In a flat or shrinking market that is hard, which is why these companies feel the pressure on headcount first.
  • These numbers describe a position, not a plan. They do not say which functions or processes create the distance, and companies in the price trap (Nissan, Mazda, Volvo, Bentley) are not shown at all: their output per employee is already at the norm, and their problem sits in price and product.

Exhibit 27

21 of 25 major carmakers have announced cuts, but the reasons behind them differ

Announcements from January 2024 to October 2026. Indices: the company’s 2021–2025 average as a percentage of its class norm.

CompanyArchetypeOutput, % of normProfit, % of normAnnounced cutsDiagnosis

Source: company releases, Reuters, Bloomberg, AP and trade press (listed under Data); author’s analysis.

Key observations

  • For companies in structural excess, cuts go after the real cause. That means Volkswagen, Honda’s car business, JLR, Aston Martin, Great Wall and, borderline, Renault. Volkswagen’s volume brands build 29% fewer cars per employee than the class median. The group has committed to about 50,000 cuts in Germany by 2030[1][6] and in September 2026 approved about 50,000 more positions group-wide[2], part of them at Audi and Porsche, which would take its volume brands close to the median. JLR, 55% short of its median, has announced 4,000 roles[4]; Great Wall (42% short) and Honda (16% short) have announced no cuts[8].
  • For companies in the price trap, cuts treat the symptom. Nissan (about 20,000 roles[14]), Volvo Cars[15], Bentley[16] and Mazda[17] build at the norm (Nissan at 107% of it), but their profit is well below it (Nissan 33%). Without new products, better pricing and new markets the margin will not come back, however many people leave.
  • Benchmarks and pricing-power brands cut as a precaution. BMW[26], Audi[23], Porsche, Ford[27], GM[30] and Mercedes-Benz[18] trim mostly indirect staff to protect the margin through the downturn. Porsche is a special case: on output it looks like a luxury benchmark, yet its 2025 profit per employee dropped below volume level, and this is already its second round of cuts[24][25].
  • Toyota, Subaru and Mitsubishi have not announced cuts, and all three are class benchmarks. If the framework works, this is exactly what we should see.
  • Can a company get out of structural excess without layoffs? There are three ways, and none of them is quick. It can grow into its headcount, as GM did, going from 18 to 25 cars per employee on recovering volume with flat headcount. It can rely on attrition and buyouts, as Volkswagen does with job security for its German staff until the end of 2030[39], but then the plan has to run for years, and Volkswagen has already left production at four German plants from 2031 open[2]. Or it can move up a class, as Jaguar is trying to, so that low output becomes normal, provided the price really goes up. In a shrinking market with flat prices cuts cannot be avoided, and the only real choice is how fast.
  • Who may be next. Stellantis went from $102k profit per employee in 2023 to below zero in 2025[40][41], and its €6bn cost-reduction plan names no role count[21]. Honda’s car business builds 16% fewer cars per employee than the volume median and has so far only promised “a fixed-cost structure appropriate for the scale”, with no number of roles[8]. Renault sits on the edge of structural excess and was reported in October 2025 to be planning about 3,000 voluntary departures in support functions, which it has not confirmed[10]. Mercedes-Benz has halved its profit per employee and so far relies on a voluntary severance programme, which about 5,500 people took by March 2026 according to Handelsblatt[19].
Sources41Show
  1. electrive, 10 Mar 2026 electrive.com
  2. Volkswagen Group, 3 Sep 2026 volkswagen-group.com
  3. Reuters via AOL, 3 Sep 2026 aol.com
  4. ITV News, 7 Sep 2026 itv.com
  5. BusinessToday, 7 Sep 2026 businesstoday.in
  6. Volkswagen Newsroom, 20 Dec 2024 volkswagen-newsroom.com
  7. CnEVPost, 30 May 2024 cnevpost.com
  8. Honda, 12 Mar 2026 global.honda
  9. Japan Times, 15 May 2024 japantimes.co.jp
  10. Reuters via Investing.com, 4 Oct 2025 investing.com
  11. Fortune, 26 Feb 2025 fortune.com
  12. RTÉ, 25 Feb 2026 rte.ie
  13. PA via Express & Star, 25 Feb 2026 expressandstar.com
  14. WardsAuto, 13 May 2025 wardsauto.com
  15. Volvo Cars, 26 May 2025 volvocars.com
  16. ITV Granada, 17 Mar 2026 itv.com
  17. Mazda, 22 Apr 2025 newsroom.mazda.com
  18. all-about-industries, 17 Mar 2025 all-about-industries.com
  19. Handelsblatt, 2 Apr 2026 handelsblatt.com
  20. Reuters via Investing.com, 21 Mar 2025 investing.com
  21. Stellantis, 21 May 2026 media.stellantis.com
  22. Electrek, 15 Apr 2024 electrek.co
  23. Audi, 17 Mar 2025 audi.com
  24. Euronews, 14 Feb 2025 euronews.com
  25. electrive, 28 Jul 2026 electrive.com
  26. Reuters via BusinessDay, 29 Jul 2026 businessday.co.za
  27. Euronews, 20 Nov 2024 euronews.com
  28. electrive, 16 Sep 2025 electrive.com
  29. NBC News, 22 Nov 2024 nbcnews.com
  30. Automotive News via Yahoo Finance, 29 Oct 2025 finance.yahoo.com
  31. TechCrunch, 11 May 2026 techcrunch.com
  32. electrive, 29 Jan 2024 electrive.com
  33. Business Insider via Yahoo Finance, 30 Oct 2025 uk.finance.yahoo.com
  34. electrive, 24 Oct 2025 electrive.com
  35. TechCrunch, 16 Jun 2026 techcrunch.com
  36. TechCrunch, 20 Feb 2026 techcrunch.com
  37. Reuters via Yahoo Finance, 22 Jun 2026 finance.yahoo.com
  38. EV (eletric-vehicles.com), citing NIO annual report, 10 Apr 2026 eletric-vehicles.com
  39. electrive, 1 Oct 2026 electrive.com
  40. Stellantis, 15 Feb 2024 stellantis.com
  41. Stellantis, 26 Feb 2026 stellantis.com
Case studiesFour cases

One company from each class: only Volkswagen’s announced cuts would bring it close to its norm, Aston Martin’s get close but leave it loss-making, JLR’s barely move it and Great Wall has announced noneFour companies in structural excess, one per class. For each: output per employee today, after the cuts announced so far, against the class norm, and what else stands between the company and its class.

VolumeVolkswagen

Volkswagen’s cuts could close most of the gap, but half of them have no date and the bigger problem is profit

Exhibit 28

If the cuts land in the volume brands, output per employee gets within a tenth of the class norm

Volume brands: VW Group less trucks, financial services, Audi and Porsche. Cuts counted without Audi and Porsche.

Vehicles per employee now

17.5

After announced cuts

22.9

Class norm

24.6

Cars a year needed to reach the median

7,264,000

Source: company reports and announcements (see Sources below); author’s analysis.

Key observations

  • On paper the cuts nearly close the gap. Volkswagen’s volume brands built 17.5 cars per employee in 2025 against a class norm of 24.6. If about 90,600 of the announced cuts land in these brands, output rises to about 22.9, close to the norm. Only the German programme has a deadline, 2030[1][6]. The further 50,000 approved in September 2026 has neither a timeframe nor a split by brand or country, and Handelsblatt expects about half of it in Germany[2][42].
  • Headcount is the smaller half of the problem. The Volkswagen brand earned a 3.0% operating margin in 2025 and has moved its 6.5% target from 2026 to 2029[43][44]. The December 2024 agreement takes about €1.5bn a year out of labour costs[6], while the group has to get from 2.8% in 2025 to 9% by 2030[1][45]. Most of that distance is price, mix and product cost.
  • China is where the volume went. Group deliveries in China fell from 4.23 million in 2019 to 2.69 million in 2025[46][47], and the joint ventures’ share of retail sales slipped from 12.2% to 10.9% in a single year[48]. Volkswagen now develops cars for China with XPeng, and the first joint model went from start to series production in 24 months[49]. About 60,000 of the group’s employees work in the Chinese joint ventures, whose profit does not show in the operating result[50], so Volkswagen’s profit per employee looks worse than the business behind it.
  • The path is volume, and the risk is that it does not come back. The plan assumes nine million vehicles a year and cheaper electric cars: the ID. Polo from under €25,000 and the ID. EVERY1 at about €20,000 from 2027[45][51][52]. European capacity still exceeds demand by more than 500,000 units, and the decision on four German plants is postponed to the end of June 2027[45]. If volume stays flat, the next round will be about plants rather than offices.
Sources11Show
  1. Handelsblatt, 3 Sep 2026 handelsblatt.com
  2. Volkswagen Group, 12 Mar 2026 volkswagen-group.com
  3. dpa via Yahoo Finance, 13 Mar 2025 finance.yahoo.com
  4. Volkswagen Group, Sep 2026 volkswagen-group.com
  5. Volkswagen Group, 14 Jan 2020 volkswagen-group.com
  6. Volkswagen Group, Jan 2026 volkswagen-group.com
  7. Reuters via MarketScreener, 12 Jan 2026 marketscreener.com
  8. Volkswagen Group, 13 Mar 2026 volkswagen-group.com
  9. Volkswagen Annual Report 2025, Mar 2026 annualreport2025.volkswagen-group.com
  10. Volkswagen Newsroom, 29 Apr 2026 volkswagen-newsroom.com
  11. Volkswagen Newsroom, 5 Mar 2025 volkswagen-newsroom.com
Mid-marketGreat Wall

Great Wall took on a quarter more people for 3% more cars, and has announced no cuts

Exhibit 29

With no cuts announced, Great Wall builds about 40% fewer cars per employee than the mid-market norm

Listed group, year-end headcount from the annual reports.

Vehicles per employee now

13.6

After announced cuts

13.6

Class norm

23.4

Cars a year needed to reach the median

2,284,000

Source: company reports and announcements (see Sources below); author’s analysis.

Key observations

  • Headcount ran ahead of sales. Between 2021 and 2025 Great Wall’s headcount grew from 77,934 to 97,600, by a quarter, while sales rose 3%, from 1.28 to 1.32 million[53][54]. Output per employee fell from 16.4 to 13.6 against a mid-market norm of 23.4. In 2025 alone headcount grew 15% and sales 7%.
  • Part of the gap is the business model. The listed group owns its powertrain, interior and chassis makers outright[55], so its headcount includes work that other carmakers buy from suppliers. Batteries are the exception: SVOLT, which supplied more than half of Great Wall’s battery packs in 2024, belongs to the chairman’s holding company rather than to the listed group[56]. Even with that caveat, the direction is clear: each year the same output takes more people.
  • Much of the new hiring seems to have gone into selling cars rather than building them. Sales expenses rose 44% in 2025 as Great Wall built its own direct-sales network[57], and net profit attributable to shareholders fell 22% to CNY 9.9bn on record revenue of CNY 223bn[54]. Revenue per car reached a record CNY 168,300[58]: Great Wall stayed out of the price cuts in a market its chairman described as having an Evergrande that “has not collapsed” yet[59].
  • The path is overseas. Foreign sales grew from 25% of the total in 2023 to 38% in 2025[54][60]. Great Wall builds complete cars in Thailand and Brazil, opened its first Brazilian plant in August 2025 and announced a second one for 200,000 cars a year in 2026[61][62][63]. The target is more than a million overseas sales by 2030[64]. If it gets there with today’s headcount, the excess disappears through growth. If it does not, Great Wall will face the cuts it has so far avoided.
Sources12Show
  1. Great Wall Motor Annual Report 2022 (HKEX), 30 Mar 2023 www1.hkexnews.hk
  2. Great Wall Motor Annual Report 2025 (A-share), 28 Mar 2026 stockmc.xueqiu.com
  3. OFweek, 25 Jun 2019 nev.ofweek.com
  4. Great Wall Motor circular (HKEX), 11 Dec 2024 www1.hkexnews.hk
  5. news18a, 30 Mar 2026 english.news18a.com
  6. 10jqka, 1 Feb 2026 stock.10jqka.com.cn
  7. CarsGuide, 28 May 2025 carsguide.com.au
  8. Great Wall Motor Annual Report 2023, 28 Mar 2024 res.gwm.com.cn
  9. Great Wall Motor, Jun 2021 gwm-global.com
  10. CnEVPost, 16 Aug 2025 cnevpost.com
  11. CarNewsChina, 27 Feb 2026 carnewschina.com
  12. Great Wall Motor Annual Report 2024 (HKEX), 28 Mar 2025 www1.hkexnews.hk
PremiumJLR

Cutting 4,000 roles is meant to save about £1.7bn, but it does not bring JLR back to the premium productivity normFiscal year April 2025 – March 2026.

Exhibit 30

Even after the cut JLR builds half as many cars per employee as the premium norm

Headcount before the cut per press reports (September 2026)[65].

Vehicles per employee now

7.2

After the cut

7.9

Premium norm

16.0

Cars a year needed at the premium median

624,000

Source: JLR interim reports, Tata Motors presentations, Business Standard/PTI (September 2026), Auto Express (October 2026); author’s analysis.

Key observations

  • The cut hardly moves the metric, and JLR itself calls it a first round. Losing 4,000 roles over two years[4][5], about 9.3% of staff, takes output from 7.2 to 7.9 cars per employee. To reach the premium median with 39,000 employees, it would have to sell about 624,000 cars a year, twice today’s 307,915.
  • On productivity JLR sits between premium and luxury, next to Porsche, with 8.3 cars per employee on average against 7.5 at Porsche. Yet even in its best year ($71k per employee) it earned at the premium median, far from luxury.
  • Jaguar Type 01 is a bet on the price path, for one brand. The electric GT was revealed in New York on 6 October 2026: from £130,000, above Porsche and just below Bentley, with first deliveries expected in the second half of 2027[66], built in Solihull[67]. Jaguar prices have roughly doubled and volumes will be smaller than before. In our metrics that means fewer cars and more profit on each, which is how luxury works. But Range Rover and Defender earn most of JLR’s profit, so Type 01 will not show in the group figures before 2028, and nobody yet knows how much demand the new Jaguar will find.
  • JLR has two paths, and they are hard to combine. The volume path means getting back to 400,000 cars or more and continuing to cut indirect staff. The price path means accepting luxury-level output and lifting profit per employee through Range Rover and the new Jaguar to the level of Porsche’s best years. The current cut saves money, but on its own it does not change JLR’s class.
Sources3Show
  1. Euronews via Yahoo Finance, 8 Sep 2026 uk.finance.yahoo.com
  2. JLR, 7 Oct 2026 media.jlr.com
  3. Automotive World, Oct 2026 automotiveworld.com
LuxuryAston Martin

Aston Martin can cut a fifth of its people and still lose money, because its problem is volume and price

Exhibit 31

A cut of a fifth brings Aston Martin to within about 15% of the luxury norm

Year-end headcount from the 2025 sustainability data; cut of up to 20% announced in February 2026.

Vehicles per employee now

1.9

After announced cuts

2.4

Class norm

2.8

Cars a year needed to reach the median

6,000

Source: company reports and announcements (see Sources below); author’s analysis.

Key observations

  • The cut brings productivity close to the norm, but not profit. Aston Martin plans to cut up to 20% of a workforce of about 2,800[13][68]. That takes it from 1.9 to about 2.4 cars per employee, against a luxury norm of 2.8. The savings of about £40m a year cover roughly a fifth of the £189m adjusted operating loss of 2025[68].
  • Volume and price fell together. Wholesales dropped from 6,620 cars in 2023 to 5,448 in 2025, revenue fell 21% and the average price 15%, as US tariffs and weak demand in China hit and the Valhalla hypercar arrived late[68][69]. Ferrari, with a similar two to three cars per employee, made an operating margin of 29.5% on 13,640 cars in the same year[70]. In luxury the brand sets the profit, and headcount decides very little.
  • The company runs on outside money. Since Lawrence Stroll’s consortium rescued it in 2020, Aston Martin has raised equity or debt almost every year: from Saudi Arabia’s PIF and Geely, from its own shareholders, from selling its stake in the Formula 1 team and, in July 2026, through a £550m loan[71][72][73]. Net debt still grew to about £1.54bn by mid-2026[73].
  • The path is price, and it depends on one car. The 2026 plan rests on about 500 Valhalla deliveries and a five-year investment plan cut from about £2bn to £1.7bn, with the electric platform pushed back[68][73]. That is the luxury model, fewer cars and more profit on each, but unlike Ferrari, Aston Martin has to prove its price every year on borrowed money.
Sources6Show
  1. Aston Martin FY2025 results (RNS), 25 Feb 2026 lse.co.uk
  2. Aston Martin Q3 2025 trading update (RNS), 6 Oct 2025 lse.co.uk
  3. Ferrari FY2025 results (SEC 6-K), 10 Feb 2026 sec.gov
  4. CNN, 31 Jan 2020 edition.cnn.com
  5. Aston Martin (RNS), 31 Mar 2025 investegate.co.uk
  6. Aston Martin H1 2026 results (RNS), Jul 2026 investegate.co.uk
AdviceWhat to do

Companies drift into structural excess through their own design decisions; six moves help avoid itWhat I would tell a carmaker’s board, in short.

  1. Choose your class on purpose. Then keep price, output per employee and costs within that class’s norm. The companies now stuck between classes (Porsche, Mercedes-Benz, JLR) moved their price and positioning but left their cost structure where it was.
  2. Measure processes, not only results. Output and profit per employee show a problem a year or two after it starts. Earlier signals are the ratio of indirect to direct staff, the number of variants per platform and engineering hours per programme. Excess builds up in indirect functions and complexity rather than on the line, and that is why JLR, Audi and BMW are cutting non-production roles.
  3. Staff for the average year, not the peak. Design processes simple and clear enough that contractors can cover a peak and leave when it ends. Porsche kept adding staff while its deliveries fell (exhibit 32 below).
  4. Put AI where the process is already clear. Indirect work grows with complexity rather than volume: change requests, supplier queries, certification files, warranty claims, reporting. Once such a process is described and measured, most of it runs as plain, predictable automation, and AI takes the small judgement steps in between, with a person signing off what the model cannot check. That is how indirect work can grow with the business without growing the headcount. AI laid over a process nobody can describe only makes the confusion faster.
  5. Share what customers cannot see, differentiate what they can. Platforms, batteries, electronics and software can be shared: Škoda, on Volkswagen Group platforms, earns more per employee than its Japanese peers at the same productivity. Design, interior, driving character and service have to stay your own.
  6. Earn the price. You cannot build premium from cheap materials. Save where the customer does not look, and never on what they see, touch and pay for.

Exhibit 32

Porsche, JLR and the luxury brands hired faster than their sales grew, while the strongest volume makers sold more cars with the same people

How to read it: each row compares two changes over the same years, from the first to the last year with data. The white circle shows how much the company’s sales changed, the dark circle how much its headcount changed. A red line means headcount grew faster than sales, so each employee now builds fewer cars. A green line means the opposite. The number on the right is the gap between the two, in percentage points.

Source: company annual reports, results and press releases 2021–2025; some headcounts from secondary sources (see Data); author’s analysis.

Key observations

  • Hiring into the peak shows up clearly in the data. Porsche grew headcount by +13%% while its deliveries changed by −7%%. JLR took on +23%% more people for +5%% more cars, Great Wall +25%% for +3%%. All three are now cutting staff or sit in structural excess.
  • The strongest volume makers did the opposite. GM grew volume by +33%% while its headcount changed by −1%%, and Ford, Honda, Toyota and Subaru also grew output faster than staff. This is how a company works its way out of excess without layoffs.
  • Luxury brands hired too, but part of that is data quality. Early headcounts for Lamborghini and Rolls-Royce are rounded company statements, and Aston Martin’s 2021 figure may cover a narrower perimeter than later press figures. Read the luxury rows as a direction rather than a precise number.
For leadershipImplications

The agenda is set by the archetype, not by company size: headcount cuts help only one of the four

ArchetypeCompaniesDiagnosisPriority agendaWarning sign
Class benchmarkOutput and profit per employee at or above 90% of the class median.Protect pricing and cost flexibility; do not add fixed cost at the peak of the cycle.Profit per employee drops below the norm while output holds, as at Porsche and Audi in 2025.
Pricing powerOutput below 90% of the class median, but the brand pays for it.Keep scarcity and price, do not chase volume; invest in brand and personalisation.Losing the price premium in a key region (Mercedes-Benz in China) turns pricing power into structural excess.
Price trapOutput at the class median, profit below 90% of it.Rework product mix, pricing and geography; seek platform partnerships. Cuts achieve little.Dependence on a single export market, as Mazda and Subaru on the US.
Structural excessBoth output and profit below 90% of the class median.Restructure headcount and capacity and make an explicit class choice at the same time: volume or price.Cuts without a change in volume or price barely move the company towards the norm (the JLR case).

What to watch

  • Output per employee against the class norm. A persistent gap of more than 20–30% signals a restructuring of headcount and capacity.
  • Profit per employee against the norm while output holds. That gap signals a change of product and pricing strategy, not cuts.
  • Output and profit per employee falling together. The sign of being stuck between classes, and of a positioning decision coming soon.

Methodology

How the metrics are built

Vehicles per employee = annual deliveries (or wholesales) ÷ year-end headcount. Profit per employee = operating profit ÷ headcount, converted to dollars at the annual average rate. Where a company headlines an adjusted figure (GM, Ford, Stellantis, Mercedes-Benz, Renault, Aston Martin, JLR), that figure is used so one-offs do not hide the norm. Company average = arithmetic mean over the years with data. Class norm = median of those averages across mature companies in the class. Archetype: index = company average ÷ class norm × 100, threshold 90 on both axes. Positions reflect five-year averages and can hide sharp 2025 moves (Stellantis, Porsche). Hyundai, Chery and Škoda have data for 2024–2025 only and SEAT/Cupra for 2021–2022, so their positions against a five-year norm are indicative. Shortfall against the median = how far output per employee in the latest year falls below the class median, in per cent; the figure after announced cuts divides the same output by headcount minus the roles announced. It describes a position, not a plan: a company can also close the distance by selling more cars.

Classes and new EV entrants

Classes are set by realised revenue per vehicle, not by how brands describe themselves. This is why some brands sit differently from the Automotive Value Map, which groups by positioning: JLR is at the top of premium here (luxury on the map), Tesla is in the mid-market since its price fell (premium on the map), and Porsche, at about €117k per car, is at the bottom of luxury. New EV entrants (XPeng, Leapmotor, Polestar, NIO, Li Auto, Zeekr, Rivian, Lucid, Lotus) are shown in their classes but excluded from the norms and from the overview map. Per-employee profit does not work for them: several buy their manufacturing from other companies (Polestar from Geely and Volvo, Lotus without its UK plant), so a small office headcount carries the whole loss, and launch-phase write-downs inflate it further. Their charts show operating margin (% of revenue) instead, omitting launch years with fewer than 1,000 deliveries.

Separating businesses

Volkswagen is shown on a volume-brand perimeter: the group less trucks and buses (TRATON), Power Engineering, Financial Services, Audi Group and Porsche AG, using the group’s own division data for headcount, deliveries and segment operating result (before group reconciliation). Škoda and SEAT/Cupra remain inside this perimeter and also appear on their own in the mid-market. For Toyota and Honda, headcount and operating profit are for the automotive segment only (financial services excluded, and motorcycles for Honda). For Hyundai, global headcount comes from sustainability reports (2024–2025 only). Kia, BYD and Geely are kept outside the comparison. For BYD, the proposed next step is to subtract the headcount of the separately listed BYD Electronic, use automotive segment revenue and allocate operating profit by segment gross profit.

Coverage and limitations

Volume 11 companies, mid-market 9, premium 10, luxury 8. McLaren and Morgan are not included: their accounts for these years are only partly available. Škoda, SEAT, Chery and Hyundai headcount was found for two years only. Polestar, Zeekr and SEAT have some cars built in other companies’ plants, and Lotus’s UK plant is outside its perimeter, so their output per employee is overstated. Japanese companies and JLR report April–March years, assigned to the starting year. Exchange rates are approximate annual averages (±2–3%). The move from combustion engines to electric cars changes how work is spread: companies that build their own batteries, motors or software carry staff that others buy in from suppliers, so comparisons within a class are rougher for 2024–2025 than for 2021. Class norms rest on small samples: 11 mature companies in volume, 7 in the mid-market, 5 in premium and 6 in luxury, and a company is part of the median it is compared with. Read the norms as reference points rather than statistical constants. The conclusions hold at the level of orders of magnitude and trends, not decimals.

Anton Semianin | Data from public company reporting, author’s analysis. Not investment advice.