NYU Stern Case Book: A 2006 Prompt, Recorded in 2016

One Stern case book prompt, three institutions, ten years

We pulled the metadata on 30 September 2026. The recording runs 1,833 seconds, which is 30.6 minutes, and it was uploaded on 18 July 2016 to a YouTube channel carrying a person's name.

Read the chain. A Stern club wrote a prompt in 2006. A Yale club worked it and recorded itself doing so. Somebody put the recording on a personal channel in 2016. Kellogg's Career Management Center library then listed it for its own students as a free resource, and the library page still does.

That single record answers the question this query keeps asking sideways. A Stern edition from 2006 was still being worked a decade later, by students at another school, and the provenance was written into the filename rather than hidden. It also tells you what these documents actually are: prompts with a long shelf life and no maintained owner.

Three of the four recordings in that library listing credit a different school, and the other two are Kellogg 2011 and Ross 2010. We work all four on our Yale case book page.

What we will not claim about that recording

The channel is personal rather than institutional. Nothing establishes that Stern authorized the use of its prompt, that its club authorized anything, or that Yale's club authorized the upload. So we state what is verified and stop.

Verified: the title, the duration in seconds, the upload date, the personal channel, and the fact that a school library lists the file as freely available. Not verified, and not asserted: any permission anywhere in that chain.

We link none of the four recordings either. A link is distribution, and nobody in the chain has established a right to distribute. What this page links is the library record, which sits on a school domain and anybody can check in one request.

One limit worth naming. We could not obtain a transcript, because the transcript tool is absent from this session and no caption track surfaced in the page HTML. Nothing spoken inside that recording is quoted here. It is the closest thing to a first-hand account of a Stern prompt being worked, and it stays half closed.

Thirty minutes on one case book prompt, and what that buys

Take the duration seriously, because it is a measurement rather than an estimate. One case book prompt, worked end to end by students with no reason to pad, ran 30.6 minutes.

Set that against the one duration a 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 landed at the bottom edge of the firm's own case-focus band.

The planning consequence is concrete. Budget half an hour per prompt worked properly, and an hour for a pair to swap so both people get an attempt. Anybody claiming to have worked twenty cases in a weekend was reading, which is a different activity with a different result.

Accenture also explains where its prompts come from, and the contrast with a 2006 student document is the point: "the cases given tend to be real business situations, often drawn from the interviewer's actual project experience". A firm's prompt comes from a live engagement. A club's prompt comes from somebody's memory of an interview, and in this instance that memory is twenty years old.

Five addresses, and no Stern club page

We could not find a public page for a Stern consulting club. Here is exactly what was tried on 30 September 2026, so that anybody can repeat it.

The two 200 responses are worth explaining, because they are the trap our process warns about. Both returned 25,669 bytes, which is the exact size this platform serves for an address with no tenant behind it. A probe that only checked status codes would have recorded two successes and found nothing.

So this page makes no claim about what a Stern club publishes or withholds. We did not reach it. Elsewhere in this batch we did reach the clubs at Wharton, Booth, Columbia, MIT Sloan and INSEAD, and the results differ sharply: MIT Sloan's club says in writing that members receive its casebook, while Columbia hosts a URL named casebook.html that serves no file at all.

The search results were closed too. A plain fetch of one results page for this query returned an anti-bot block rather than results, so unlike most schools in this batch we did not even observe where copies are hosted. That absence is recorded rather than filled in.

What Stern's own pages say, and where they disagree

Stern publishes a consulting career page with a list of fifteen firms under the heading "Select Consulting Employers of NYU Stern MBAs", and a sentence about access: "internship opportunities with top-tier consulting firms like McKinsey, Boston Consulting Group and Bain".

That list has not been maintained. It carries "A.T. Kearney", a name the firm stopped using in 2020, and it lists "EY" and "Parthenon - EY" as two separate entries. The school's own 2025 employment report, by contrast, writes "Kearney" and "EY-Parthenon". Two pages from one school spell the same firms differently, and only one of them has been revised in six years.

The staleness runs deeper. The school's year over year industry page serves Class of 2018 data, with consulting at 28.4 percent and a $147,000 median. Its parent page promises data for the classes of 2023, 2024 and 2025. We searched the extracted text of the child page for the string "Class of 202" and it does not appear once.

None of that makes the school unreliable. It makes the source you pick matter. Our Stern recruiting page works the 2025 report itself, where consulting reads 32.8 percent by industry and the three year series shows nine points lost.

Working a twenty year old prompt without being taught wrong

A 2006 prompt is still usable, with one restriction. The business situation may have aged, and the shape of the exercise has not. Use the prompt and discard the surrounding apparatus.

1. The partner reads the situation and the question once, then stops. No exhibit, no hint, no additional sentence, whatever the silence costs.
2. Ninety seconds on paper, announced out loud. The partner calls start and stop. A structure that takes four minutes is a different exercise.
3. Deliver it aloud and say why it fits this business. Name every branch. The partner writes down the branch you could not finish in one breath.
4. Ask for data by name, and sparingly. Deloitte publishes the reason: "Ask questions, but don't use them as a crutch. While it's important to ask clarifying questions, avoid relying on them too heavily. We're interested in how you connect the dots when solving the problem."
5. One sentence of recommendation, answer first. Then the number under it, then the risk that would reverse it.

Then swap. Budget an hour for the pair, which is what the recording above measures. The partner's whole job is to withhold, to time and to record, and generosity destroys the exercise on purpose.

Checking a dated prompt against a current firm sequence

Deloitte publishes a five step frame and introduces it plainly: "Apply this five-step approach while working through the case or scenario interview". Its steps are understanding the issue, identifying the underlying assumptions, summarizing specific issues and findings, stating recommendations, and outlining next steps and expected results.

That last step is the useful test for an old write-up. Read the model answer in any case book and ask whether it ends on a recommendation or continues into a next step. Most student write-ups stop at the recommendation, because that is where the interview they remembered ended.

L.E.K. adds the tension a dated write-up also tends to flatten, in step six of its own seven: "Commit yourself to a course of action, but show flexibility and a willingness to rethink in the face of new data." A write-up shows you a settled answer. The room asks you to hold one and revise it.

Across seven firms we count 65 published instructions and 43 practice cases readable with no account. None of them publishes a weight: across forty-four firm research files we hold, zero contain a published scoring scheme, so any page showing a firm's rubric with numbers on it is showing you a fabrication.

Which dimension a Stern miss belongs to, and whose numbers

Record a miss only when it has a name that will recur. Losing a cost branch is a coverage failure, and it shows up again the following week in a different industry. Forgetting one line item in one prompt tells you nothing and should not go in the log.

The weights here are ours, read from src/lib/grading/rubrics.ts in our own code. No consulting firm publishes a rubric.

Coverage takes almost half the score, which encodes a judgment about which mistake costs most. Losing a whole branch of the problem cannot be repaired by precision inside the branches you did cover. The mapping onto the protocol above is direct: a branch your partner marked absent is Coverage, two branches counting the same dollar twice is Separation, and a structure delivered on time with no specific line under any branch is Depth.

Every rubric and the reasoning behind each weight 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. There is no live interviewer in any of it, none of it teaches business content, and none of our six assessment simulators has shipped, which that page states itself.

What changed on the Stern page, and the pointer deleted

30 September 2026. Rewritten after the 2016 recording of a Stern 2006 prompt was located through a peer school's library catalogue and its metadata read, and after five Stern club addresses were probed. The earlier version ran 269 words and described what a case book is good and bad for without ever asking who publishes the Stern one.

The earlier version closed by pointing readers at our own free resources page for case books from several schools. That page has never held a single one. No such file exists on this site, and the rights position above is the reason none will be added. We deleted the pointer from all eleven pages that carried it. The error itself was recorded in the shared case book research file on 29 September 2026, after a check of the repository found no PDF, no ZIP and no archive of any kind.

Three controls stay red on purpose. We reached no public Stern club page, and five documented attempts is the whole of what we know. We could not read this query's results, so we make no claim about where copies are hosted. And we have no transcript of the recording, so nothing said inside it is quoted. Every attempt and its date is in the research file.