Automotive Consulting 2026: Twelve Firms, No Ranking

Why this automotive consulting page has no ranking

Three of the results that rank automotive firms order them by a survey of clients and consultants, and none of the three publishes the sample. A fourth ranks named individuals, which is a thing we do not do at all. A ranking built on an undisclosed survey is a number a reader cannot check, and our own rules forbid publishing one.

What a reader can check is a test applied identically to every firm, so here it is, stated before the list rather than after it. A firm appears below if it publishes an automotive practice page on its own domain, we opened that page ourselves on 30 September 2026, and the page names at least one vehicle segment, one client type, or one figure. Sixteen firms were checked. Twelve passed.

The test has a consequence worth stating. A large firm with real automotive work and no readable automotive page does not appear, and its absence says nothing about its presence in the sector. Accenture is exactly that case, and the reason is recorded further down with the request that failed.

The test also means the list mixes kinds of firm on purpose: generalist strategy firms, restructuring and performance firms, an audit-network practice, a firm owned by a carmaker, and two firms that do nothing but automotive. Those are different businesses that sell to the same buyers, and a candidate who treats them as interchangeable will prepare for the wrong interview.

Twelve automotive consulting practices, in their own words

Every quotation below was read on the firm's own page on the date given in the source ledger.

One thing to notice across the rows. BCG is the only firm of the twelve that publishes a practice headcount, and Porsche Consulting is the only one that publishes a firm headcount. Nobody publishes automotive practice revenue.

The German center of gravity, read from where the co-leads sit

The old claim that Germany holds a disproportionate share of automotive consulting is easy to assert and awkward to source. Four readings support it, and each one is a fact rather than an impression.

McKinsey's Automotive & Assembly page, checked 30 September 2026, lists two global co-leaders of the practice. One sits in Stuttgart. Porsche Consulting is headquartered in Germany, is owned by a Stuttgart carmaker, and lists five German offices before any other country. Roland Berger publishes its automotive services in English, German and Chinese, and in that order on the page. And Berylls, the automotive-only firm, published the summary of its 2026 supplier study in German on its own site before the English version.

That is a fact about where the work is bought, and it has two practical consequences for a candidate. First, a German-language application is a real requirement at some of these firms for some of these offices, and the firm's own careers pages are where to check it rather than a list like this one. Second, the primary material on this sector is partly not in English, which means the supplier economics you would use in a case have a reading list that a US-only search will miss.

Houston, the other co-lead's city, is the counterweight and it says something too. Automotive strategy work sits next to energy and industrial work, which is why the same practice at McKinsey covers "automotive and mobility, machinery, aerospace" in the words of its own page.

Two firms do only automotive consulting, one owned by a carmaker

Three rows in that table are structurally different from the other nine, and the difference matters more than any ranking would.

Berylls trades as Berylls by AlixPartners and does nothing but automotive. Its public output is an annual study of the hundred largest suppliers, and its 2026 edition states three things plainly: the industry's interest coverage ratio "is now at its lowest point since the onset of the electrification wave", the investment rate among the largest suppliers "has fallen to its lowest level in at least eight years", and "the supplier industry is losing its financial strength at precisely the moment when it needs to finance the most expensive transformation in its history." The full study sits behind a form, and we do not have the underlying values, so those three sentences are all we can attribute.

Ducker Carlisle is the other automotive-only firm, and its published business reaches past strategy advice. The page names "automotive and aftermarket data collection, benchmarking, and insights" and claims "over six decades" in the sector. A firm that sells benchmark data sells a different product to the same purchasing department.

Porsche Consulting is the case nobody on the ranked lists mentions. Porsche AG's own newsroom states it: "Headquartered in Bietigheim-Bissingen, Porsche Consulting GmbH is a subsidiary of the Stuttgart-based sports car manufacturer Dr. Ing. h.c. F. Porsche AG. Founded in 1994, the company currently employs 800 people". A consultancy owned by a manufacturer in the sector it advises is an unusual structure, and it is a reasonable thing to ask about in an interview there.

Worth flagging that the firm and its parent disagree on details. The newsroom places the headquarters in Bietigheim-Bissingen and names a Belmont office; the firm's own "who we are" page says "In addition to our headquarters in Stuttgart, we have offices in Hamburg, Munich, Berlin, Frankfurt, Milan, Paris, São Paulo, Atlanta, Los Angeles, Palo Alto, Beijing and Shanghai." Both pages were read on 30 September 2026. We publish the divergence rather than pick a side. Our page on what Porsche Consulting is counts all four of its office lists. The group gives two headcounts for the same firm.

The demand forecast, and the same numbers worked backward

AlixPartners publishes the only dated, fully readable sector study of the twelve. Its 23rd annual Global Automotive Outlook, checked 30 September 2026, carries a Detroit, London and Shanghai dateline of 25 June 2026. Two figures anchor everything else in it.

"U.S. new-vehicle sales in 2026 forecast to fall 2.5% to 15.8 million, rising to 16.8 million by 2030." And: "China sales are forecast to fall 10% this year to 24.6 million, rising to 26.2 million in 2030."

Work those backward, because the implied base year is the part nobody quotes. If 15.8 million is 2.5% below 2025, then 2025 was 15.8 divided by 0.975, which is 16.205 million, so about 16.2 million US vehicles. Check it forward: 16.205 times 0.975 gives 15.80. For China, 24.6 divided by 0.90 is 27.33 million, so the 2025 base was roughly 27.3 million units.

Now the growth. The US goes from 15.8 million in 2026 to 16.8 million in 2030, which is 1.0 million more units across four years, and 1.0 divided by 15.8 is 6.3% in total. China goes from 24.6 to 26.2, which is 1.6 million more units, and 1.6 divided by 24.6 is 6.5%. Two very different markets, forecast to grow by almost exactly the same percentage over the same four years, from bases that differ by a factor of about 1.7.

One more from the same release: "Chinese automakers will export nearly 10 million vehicles in 2026, up from 7.1 million in 2025." That is 2.9 million additional units, and 2.9 divided by 7.1 is 40.8% growth in one year, against home-market volumes the same release forecasts falling 10%.

Where two thousand dollars a vehicle would land

The release includes a cost figure that is exactly the kind of number a case interviewer would hand you: "The hidden compliance costs for 'USMCA 2.0' could add up to $2,000 per vehicle in annual costs."

Multiply it out, carefully. Two thousand dollars per vehicle against the 15.8 million US new-vehicle sales the same release forecasts for 2026 gives 15.8 million times 2,000 dollars, which is 31.6 billion dollars a year. Check the units: vehicles times dollars per vehicle gives dollars, and the per-vehicle figure is described as annual, so the product is dollars per year.

Then say what it is worth. AlixPartners does not publish the base the 2,000 dollars applies to. USMCA compliance covers vehicles traded under that agreement, which is not the same set as every vehicle sold in the United States, and the release never names the volume. So 31.6 billion dollars a year is a ceiling rather than an estimate, and the honest way to present it in an interview is to state the assumption out loud before the multiplication, then ask for the compliant share.

That distinction is the whole skill. The arithmetic takes ten seconds. Knowing that the answer is a ceiling, and saying so, is what separates a candidate who can be trusted with a client number from one who cannot.

Ninety-one percent fewer parts, and what that does not let you compute

Kearney, checked 30 September 2026, publishes the single most quoted structural number in the sector: "the move to e-mobility is a move from engines with roughly 4,500 parts to those with 400." That is a reduction of 4,100 parts, and 4,100 divided by 4,500 is 91.1%.

Oliver Wyman, checked 30 September 2026, publishes the other half of the picture: "Today, parts, modules, and systems from suppliers are responsible for about 60 percent of the typical added value of a car."

The tempting move is to combine them and conclude something about supplier value at risk. Do not. Neither firm publishes a mapping from a part count to a share of value, and a powertrain part and a seat frame are not worth the same. Combining the two would produce a confident number with nothing underneath it, which is the error that our own arithmetic rules exist to catch. What the two figures jointly support is narrower and still useful: the supplier base holds most of the added value, and the part count of the component it is most specialized in is collapsing.

Berylls, reading the same industry from the accounts rather than the bill of materials, reports that supplier investment rates are at an eight-year low while the transformation is the most expensive in the industry's history. Those two statements point the same way without needing to be multiplied together.

What automotive consulting changes in a case, and what it does not

No interviewer expects you to know a bill of materials. Ask for any industry fact and you will be given it. What the sector changes is which structure fits, and three published facts decide that.

The fixed-cost base does not flex. Roland Berger, checked 30 September 2026, states the constraint from the incumbent side: "unlike the new start-ups they have to deliver expected profits every quarter, while future proofing and transforming today's core business". A profitability case in this sector therefore lives on volume through the plant before it lives on price, because price moves without moving cost and utilization moves both.

Margin arrives after the sale. EY writes it directly: "Long‑term profitability in engineered products is shaped after the sale, where recurring services drive stability, margins and valuation." Oliver Wyman quantifies the customer side of it at "around EUR 300,000 on automotive mobility during his lifetime", a number published without a reference year or a geography, which is how we cite it.

The customer relationship is loosening. Deloitte's 2026 study of "more than 28,500 consumers across 27 key automotive markets" reports that brand loyalty "continues to shift", and Deloitte's own summary of it on its automotive page says "Just over half of U.S. consumers surveyed plan to switch to another brand for their next vehicle purchase." Half the installed base intending to leave is the reason aftersales and financing get their own workstreams.

Put those together and the structure writes itself: units through the plant, contribution per unit, then the post-sale revenue per unit and the retention rate that decides whether it arrives. Our page on automotive cases works that structure, and the interview process covers what varies by firm and region.

Access notes, and the four firms that fell off

Publishing the method is part of publishing the list. Three of the twelve pages refused a direct request and were read another way. BCG returns 403 to any path and was read through a text proxy. Kearney and Bain did the same. McKinsey's practice page returned nothing at all and its text proxy returned 403, so it was read from the 11 June 2026 Internet Archive capture, found through the CDX API, which also showed captures on 15 January and 12 April 2026. Deloitte's automotive page redirected and was read through the proxy.

Four firms were checked and did not make the list, each for a reason we can state.

1. Accenture. Its automotive path redirects and the index path returns 404. No Accenture automotive page opened on 30 September 2026, although its High Tech industry page opens normally. Excluded for want of a readable page, and not for want of presence.
2. Horváth. Direct requests return 403 and the proxy serves site navigation. The only automotive wording we could read was the generic phrase "All the way from A for automotive to U for utilities", which names no segment and no client.
3. L.E.K. Its industries page returns 200 and serves navigation only, with no readable automotive sub-segment. It is on our technology list, where its sector page does open.
4. Simon-Kucher. Its industries page states firm scale, "over 2,200 employees in 30+ countries", and names no automotive industry in the served text.

None of the twelve publishes an automotive recruiting date, a sector-specific application door, or a practice headcount other than BCG. Recruiting runs through each firm's general entry process, which is why this page does not carry a calendar.

Drilling a case whose fixed costs do not move

The structure above is what to practice, and it is practicable without knowing anything about cars. Our drills run one exercise type at a time and score each against published weights, so an estimation problem and an interpretation problem stay separable; market sizing is Assumption chain 30, Scope 25, Arithmetic 25, Sanity check 20, and the reasoning behind those numbers is on how an answer is assessed. What the tool does not do: it holds no automotive case bank, it teaches no sector facts, and our assessments page states that 0 of 6 simulators are live. For pay, the salary pages carry what is sourceable, including Roland Berger, which publishes its automotive services in German as well as English.

One order of work, for a candidate with a first-round date in this sector. Start with the plant, because the utilization question is the one that separates a structure that fits from one that was memorized: units produced, units sold, contribution per unit, fixed cost per period. Then add the post-sale line, because two of the twelve firms publish that margin arrives after the sale and neither of them treats it as an afterthought. Then, last, read one of the published sector documents named on this page end to end, so that when an interviewer hands you a volume forecast you recognize the shape of it and ask for the base year rather than multiplying immediately.

That last habit is the transferable one. Every number quoted on this page came with something missing: a base year, a geography, a compliant share, a sample. Asking which is missing, out loud, before the arithmetic starts is the behavior a sector case is actually testing.

Corrections to the earlier automotive draft

30 September 2026. Rewritten from a 257 word draft. Three of its claims were unsourced and two of them are now sourced differently. It said Germany and Japan hold "a disproportionate share" of the sector's dedicated firms. The German half now rests on four readings named above, and the claim about Japan is gone, because nothing we read supports it. It asserted "the heaviest fixed-cost base of any consumer industry", a superlative across industries that no page we read establishes, and that is gone too; what replaced it is Roland Berger's own sentence about quarterly profit obligations.

It also asserted that "a large share of lifetime margin arrives after the sale" and that "the manufacturer usually does not own the customer relationship". The first now stands on EY's published sentence. The second was replaced by Deloitte's measured finding on brand switching, which is a narrower claim and a checkable one.

Two spellings were corrected as part of the rewrite: "utilisation" and "memorise" were British forms in our own prose, which our style rules do not allow outside a quotation.