Yale Case Book: Its Club Practiced Three Rivals' Cases

Four recordings, and the case books their titles credit

The case prep guide published by Kellogg's Career Management Center library lists them under the heading "Yale Consulting Case Videos", each tagged by the library as "This resource is freely available online". We pulled the metadata for all four on 30 September 2026.

Read the right column. A club at one school recorded itself working prompts from three other schools' case books, dated 2006, 2010 and 2011, and put the provenance in the title of each file. Then a fourth school's library catalogued the recordings and told its own students they were free to watch.

That single table settles a question every page on this subject dances around. School case books circulate between schools, students know exactly which document a prompt came from, and nobody in the chain treats the boundary between schools as meaningful. The Yale club was not hiding it. It labeled the files.

What we will not claim about those files

The four recordings sit on a YouTube channel carrying a person's name rather than an institutional channel. Their upload dates run from 16 to 25 July 2016. So the safe statement is narrow and we will hold to it.

What is verified: the titles, the durations in seconds, the upload dates, the fact that the channel is personal, and the fact that Kellogg's library lists them as freely available. What is not verified, and we will not assert: that Yale SOM published them, that its club authorized them, or that the three schools whose prompts appear consented to any of it.

We also link none of the four. Nothing in the chain establishes a right to distribute, and a link is distribution. This page names what exists and points at the library record that catalogues it, which is a page on a school domain that anybody can check.

One more limit. We could not obtain a transcript, because the transcript tool is absent from this session and no caption track surfaced in the page HTML. So nothing spoken inside those recordings is quoted here. It is the best first-hand source in this whole batch and it stays half closed, which is written into our research file with the attempt.

Twenty nine minutes, against the only duration a firm publishes

The four durations add to 6,992 seconds. The arithmetic runs 2,094 plus 1,428 is 3,522; plus 1,637 is 5,159; plus 1,833 is 6,992. Divided by four that is 1,748 seconds, or 29.1 minutes per recording.

Set that against the one duration any consulting firm publishes. Accenture's own case workbook states that "case interviews typically last from 45 to 60 minutes, with a focus on the case itself lasting 30 to 45 minutes", checked 29 September 2026.

So a student-run mock of a case book prompt ran, on average, slightly under the bottom of the firm's own case-focus band. That is a useful calibration for anybody scheduling practice. A prompt worked properly end to end is a half hour of work, and the recordings were made by people with no incentive to pad.

Two caveats. A recording includes setup and discussion that an interview does not, and it excludes the personal experience portion that a real round contains. And four recordings are four data points from one club in one year, which is a measurement rather than a norm. Treat 29 minutes as the floor of a realistic practice block rather than as a standard.

Where a Yale case book prompt legitimately comes from

We found no public page on a Yale domain that serves a case book to a visitor. We probed the two student group platforms that host school clubs, groups.som.yale.edu and yalesom.campusgroups.com, across six plausible club identifiers. Every one returned HTTP 200 with an empty application shell of about 10,465 bytes, which is the platform's response for an address with no tenant behind it.

The search results for this query were not readable either. A plain HTTP fetch of one results page returned an anti-bot block rather than results, so unlike the other schools in this batch we did not even observe where copies are hosted. This page says that rather than assuming the pattern holds.

What we do know about a Yale-attributed copy comes from a different finding, already recorded on 29 September 2026. A career services site at an unrelated university republishes six case books in open download, attributed to Duke, Darden, HBS, McGill, Wharton and Yale, with no mention anywhere of permission from the clubs that wrote them. Institutions redistribute these documents too, alongside the aggregators, and the institutions document their right to do it no better.

So the legitimate routes for a Yale student are the ones that do not involve a file: ask the club, use the career office, and work the cases the firms publish themselves.

What the firms publish, which a student write-up cannot carry

Across seven firms we count 65 separate published instructions and 43 practice cases readable without an account. Kellogg's library points at nine firms' own material on the same page that lists the Yale recordings, which is a school career service telling students to read the firms directly.

Kearney publishes something a case book write-up structurally cannot give you, because a write-up has no interviewer in it: "If things aren't going well, don't try to tough it out. Acknowledge that you are stuck, seek more information, or pursue a different logical path." That is written permission, from a firm, to say out loud that you are stuck.

Nobody practicing alone off a PDF ever exercises that. The recordings above are closer, because a partner is present and can be asked. This is the argument for the pair protocol below and against silent reading, and it comes from a firm rather than from us.

No firm publishes a weight. Across forty-four firm research files we hold, zero contain a published scoring scheme. Any page that shows you a firm's rubric with numbers on it is showing you something invented.

What Yale publishes about consulting, and the word doing two jobs

Yale splits consulting into two lines in its own salary tables, and the gap between them is the finding. From the Class of 2024 employment report, checked 30 September 2026.

The median difference between the two function lines is $60,000. Both carry the word consulting. Yale is one of very few schools that price them separately, so a number taken from a school that merges them is describing a blended figure.

The school also warns about its own category in plain words: "More SOM students accepted their first role in consulting than in any other industry, but that category contains striking variety. Students accepted consulting roles that focus on the nonprofit industry, healthcare, and strategy among many other fields."

The percentage shares for that class are published only inside pie chart images whose text has been converted to vector paths, so they cannot be read from the page. The prior year's PDF does print them: for the Class of 2023, consulting services took 44.2 percent by industry and the consulting function took 54.2 percent, split 43.5 external and 10.8 internal. Our Yale SOM recruiting page works both years and the rounding inside them.

Working one prompt when a partner is in the room

The recordings above are pair work, and that is what a case book prompt is for. Everything after the prompt is one club's reconstruction from memory.

1. The partner reads the situation and the question once. Then silence. No exhibit, no hint, no additional sentence, whatever the pause costs.
2. Ninety seconds on paper, called out loud. The partner announces start and stop. Four minutes is a different exercise.
3. Delivered aloud, uninterrupted. Name each branch. The partner notes which branch you could not finish in one breath.
4. Use the permission. If you are lost, say so and ask for a different path, exactly as Kearney's published advice instructs. The partner records whether you said it early or late.
5. One sentence of recommendation, answer first. Then the supporting number, then the risk that would reverse it.

Then swap. Two people working one prompt each way runs close to the 29 minutes each that the Yale recordings measured, so budget an hour for a full exchange. The partner's job is to withhold, to time and to record. Volunteering a figure or nodding at a thin branch deletes the difficulty deliberately.

Naming the dimension a Yale miss belongs to

A miss is worth recording when it has a name that can recur. A missing cost branch is a coverage problem and it comes back next week in another industry. One forgotten line item in one prompt is noise.

These weights are ours, read from src/lib/grading/rubrics.ts. No consulting firm publishes a rubric, so weights attributed to a firm are fabricated.

Coverage takes 45 of 100 because an answer that misses a whole branch cannot be rescued by precision inside the branches it covered. The mapping onto the protocol is direct. A branch the partner marked absent is Coverage. Two branches counting the same dollar twice is Separation. A structure delivered on time with no specific line under any branch is Depth.

Every rubric and its reasoning sits on how an answer is assessed. The free drills score these dimensions on prompts we wrote ourselves, with no account, and the structuring drill returns the first correction rather than a grade. What they do not do: no live interviewer, no business content, and none of our six assessment simulators is live yet, which that page states itself.

What changed on this page, and the pointer we deleted

30 September 2026. Rewritten after the four club recordings were located through a peer school's library catalogue and their metadata read. The earlier version ran 269 words and described what a case book is good and bad for without ever establishing who publishes the Yale one or whether anybody does.

The earlier version closed on a pointer to our own free resources page, telling readers that case books from several schools were collected there. Nothing of the kind is on that page, no such file has ever existed here, and our rights position means none will. The pointer is deleted. We first recorded this error in the shared case book research file on 29 September 2026, and it was printed on eleven of our pages.

Three controls stay red on purpose. We have no transcript of the four recordings, so nothing said inside them is quoted. We could not read this query's results at all, so we make no claim about where Yale copies are hosted. And we cannot say who holds the rights in those recordings or in the file that circulates, because neither the school nor the club claims either on any public page. The attempts and their dates are written into the research file.