Try a free Cornerstone Research-style case now: Brightline Ad Damages. No card, no signup.
Free, no signup: 2 worked cases with model answers. Free, no card: your first full practice case with an AI interviewer. Paid: Pro $49/month for every case.
Updated ·By Esteban Ronsin·Founder · former Deloitte and Accenture Strategy consultant
Built and reviewed with former McKinsey, BCG and Bain consultants and interviewers.
Free Cornerstone Research interview prep resources
Cornerstone Research interview prep resources: what you get, whether it is free or paid, and the link
Resource
What you get
Free or paid
Link
Cornerstone Research: what we look for in analyst candidates
Cornerstone's own interview page: two rounds with a different case in each, five tips for the case, its AI guidance, and two example cases (Zilo, a securities case, and Duff, an intellectual property case) worked on its analyst case examples page
Free. Published by Cornerstone Research, checked Nov 11, 2026
Practice cases for Cornerstone Research interviews7 cases
7 library cases matched to Cornerstone Research's case types, run out loud with an AI interviewer or step by step and scored on 7 skills; start with Thomas the Tank Rail Revenue Decline
Free + paid. The case overview is public; a free account includes 1 full AI-graded case in voice or guided mode; Pro ($49/month) adds every case
Dunmore Cargo Bikes: market sizing case with answers Industry: Mobility. Difficulty: easy.1 math question.
How big is the annual US market for electric cargo bikes sold to urban delivery couriers, and is it big enough for a $10 million revenue goal? 5 questions, 1 exhibit, a model answer for each
Free. The whole case and every model answer, no account; the practice button opens a live case like it
Northbay Grocers: competitive response case with answers Industry: Grocery retail. Difficulty: medium.1 math question.
How should a regional supermarket chain respond to a hard-discount grocer opening nearby with prices about 20% lower? 5 questions, 2 exhibits, a model answer for each
Free. The whole case and every model answer, no account; the practice button opens a live case like it
Consulting fit and behavioral questions answered out loud, with follow-up probes and structured feedback
Free + paid. 1 behavioral or Learning Mode session with a free account; Behavioral Unlimited ($39/month, behavioral and PEI only) or any Pro plan adds more
Firm-style cases are practice cases written for Road to Offer or adapted from university consulting club casebooks, not the firm's own cases. Road to Offer is independent and not affiliated with Cornerstone Research.
Try a free Cornerstone Research-style case now: no card, no signup
An original practice case written by Road to Offer in a Cornerstone Research-style format, not Cornerstone Research's own case. Everything stays in your browser.
Brightline Ad Damages: Did a rival's advertising cost a gym chain its members?
Cornerstone's own cases are litigation economics, such as isolating how much of a price drop an alleged wrong explains, which it likens to measuring the causal effect of one variable on another. This is an original exercise in that same reasoning, in a business setting and not a legal case. Our library has no litigation-economics case, so the full version below, Thomas the Tank, is the closest library case: a revenue decline to explain. Cornerstone Research (opens in a new tab), read
1Step 1: Read the case
Brightline Fitness runs forty studios in one metro area, and its membership revenue has fallen by a fifth over two years. Brightline says a rival, Nova Gyms, ran misleading advertising for the last eighteen months and pulled members away, and it is suing for the lost revenue. You are helping Nova's lawyers. How much of the fall, if any, can fairly be blamed on the advertising?
2Step 2: Ask clarifying questions
Open a question to hear the interviewer's answer.
What exactly is the claim?
Interviewer: Brightline says Nova's ads promised prices and class quality that Nova did not deliver, and that members left Brightline for Nova because of them. It wants the revenue it says it lost during the eighteen months the ads ran.
What do we know about Brightline's revenue?
Interviewer: Index it to 100 two years ago. It is 80 today, and most of the fall came in the last eighteen months.
Did anything else change at Brightline or in the market?
Interviewer: Brightline raised its monthly price by about a tenth a year ago. Ask for anything else you think matters and I will tell you whether the client has the data.
Is there a comparison group?
Interviewer: Yes. Brightline runs six studios in two nearby towns where Nova has no studios and the ads did not run. Their revenue is on the same index.
Do we have member records?
Interviewer: Yes: sign-up and cancel dates, home studio and, when members gave one, the reason for leaving.
What does the client want from us?
Interviewer: A number: the share of the fall the ads could explain, how sure you are, and which data you would ask for first.
3Step 3: Structure your answer
Take two minutes. Lay out how you would test Brightline's claim and get to a number, then say what you would ask for first.
2:00
Show what a strong structure covers
Plays the claim back as a chain: the ads reached members, members left because of them, they went to Nova, and Brightline lost revenue as a result. Each link needs evidence.
Names the but-for world: what Brightline's revenue would have been without the ads. Every number is measured against that baseline.
Lists other explanations before blaming the ads: Brightline's own price rise, local job losses, Nova's lawful competition from opening studios, seasonality and Brightline's service quality.
Uses the comparison towns as the baseline: same company, same period, no ads, so anything that hit them too is not the ads.
Uses the member records to test the story: when members left against when the ads ran, where they went and the reasons they gave.
Remembers revenue is not profit: costs Brightline saved on members who left come off any damages figure.
Ends with a number and a range, not a direction, and names the first data request.
4Step 4: The interviewer's next question
In the comparison towns the revenue index fell from 100 to 92. In the metro with Nova it fell from 100 to 80. What does that tell you about the ads?
Show what a strong answer does
The comparison towns show that common factors, such as Brightline's price rise, took about 8 points, so at most the other 12 of the 20 points, roughly 60 percent, can be tied to Nova being there. A strong answer calls that a ceiling, not the answer: it still includes Nova's lawful competition, so the ads' share is smaller. It then says what would split it, such as the reasons members gave and whether cancellations jumped after the ads started.
Cornerstone says management consulting focuses on business strategy, while economic consulting provides the data-driven foundation for expert testimony in court cases and regulatory proceedings. Its two published cases ask how much of a stock drop or a lost profit an alleged wrong explains. Source: Cornerstone Research, read (opens in a new tab)
Cases written as finished engagements.
Each published case names who hired Cornerstone, the experts and team on it, the analysis and the outcome, a settlement in one and a small award in the other, with names changed. Source: Cornerstone Research, read (opens in a new tab)
Applications run through separate US, UK and Belgium portals; the brochure lists offices in Boston, Brussels, Chicago, London, Los Angeles, New York, San Francisco, Silicon Valley and Washington. Source: Cornerstone Research, read (opens in a new tab)
Each arrow opens the source: Cornerstone Research, read Nov 11, 2026.
Cornerstone Research's interview pages and practice cases, item by item
Everything Cornerstone publishes to prepare for its interviews, each opened on 11 November 2026. The two practice cases are on one page; each line is our summary of the item's own page. Read on Cornerstone Research's site .
Cornerstone Research's interview pages and practice cases, item by item: level, area and what each one is, linked to Cornerstone Research's page
The two rounds, the case in each round, five tips, its AI guidance, the two example cases and three interviewers' advice, including to weigh the behavioral part as much as the case.
A high-tech company's stock rises after its IPO, then falls when two customers worth 40 percent of sales leave; plaintiffs claim the full 40 dollar fall. Four key questions with a hint each, then how the team separated market, industry and company news. Worked below.
A court finds Duff used ACME's patented techniques and ACME claims huge lost earnings. Three key questions with a hint each, then how the team tested customer overlap, other causes of ACME's decline and costs.
The same two cases told as the team's story, with who did what and the outcome, and the line that the case interview also assesses your ability to direct a team.
The three required documents, what to do if Cornerstone does or does not recruit on your campus, and who to ask about your status. It lists no dates and no interview rounds.
Materials for MBA and PhD candidates (PhD candidates add three reference letters and a sample research paper) and the regional portals; no interview rounds described.
Show all 10
Cornerstone Research's interview pages and practice cases, item by item, continued
Cornerstone's description of the work (litigation and regulatory matters, expert testimony, antitrust, finance, healthcare, intellectual property, valuation) and how it differs from management consulting.
Two pages on the role and training, the three application documents and a line saying interviews contain both behavioral and case questions; names the nine offices.
Links to the analyst, associate, corporate and experienced-candidate tracks and the US, UK and Belgium application portals.
Worked: Zilo (Securities), in Cornerstone's four key questions and its five tips
A high-tech company's stock rose from its 20 dollar IPO price to over 60, then fell to 16 after the company lost two customers worth 40 percent of its sales. Shareholders say officers knew the relationships had collapsed months earlier, and plaintiffs claim damages of 40 dollars a share, the fall over the month the news came out. Read the official case (opens in a new tab)(read )
How you might work it is ours, not Cornerstone Research's answer key.
1
What is wrong with the plaintiffs' number?
Say the goal first: isolate the effect of the alleged misconduct on the price. Comparing the price before and after the news assumes nothing else that affects the price changed in that month. The whole 40 dollar fall is therefore an upper bound, not a damages figure.
2
What else could have moved the stock?
Break it into parts and widen the view: market-wide moves, industry news, rivals' results, the company's own earlier decline and volatility, and news about the two customers that was not the alleged concealment. Take notes as you list them.
3
Estimate the but-for price
Damages are the gap between the actual price and what it would have been without the news about the alleged wrong. Ask for a model of Zilo's daily price in a clean period before the class period, as a function of the stock market and an industry index, then predict the class-period prices.
4
Measure the economy-wide and industry-wide effects
Compare predicted and actual returns day by day and find the days where the actual is statistically significantly lower. Cornerstone's summary says most days were explained by market and industry news, so those drops are not the alleged misconduct.
5
Match what is left to company news
For the remaining significant drops, run an event study: line each one up with announcements and analyst reports released around the same time, and keep only the ones tied to the allegations.
6
State the answer and its limit
Cornerstone's page ends with only about 3 dollars of the 40 dollar fall possibly related to the allegations. Say the number, say it depends on the model and the news matching, and say what you would check next, such as the length of the pre-period.
Behavioral as much as the case (Washington manager)
Candidates should consider the behavioral components "as much or even more than the case element" and be ready to explain how they handled similar situations, especially collaboration, conflict management and mentorship. Source: Cornerstone Research, read (opens in a new tab)
Each arrow opens the source: Cornerstone Research, read Nov 11, 2026.
The Cornerstone Research interview process, and what trains each step
1
Application: cover letter, resume and transcript
Analyst and Summer Analyst applicants send a cover letter (with any office preference), a resume and a transcript through a US, UK or Belgium portal; Cornerstone says applications with missing materials will not be considered. MBA and PhD candidates apply to the separate Associate track.
Cornerstone says the Analyst and Summer Analyst process is identical and starts with two virtual interviews of 30 minutes. Each round includes a case interview, with a different case each time, and its brochure says the interviews contain both behavioral and case questions.
The second round is a multi-hour final round in one of Cornerstone's offices, again with a case to review and discuss. Cornerstone does not say how many interviews it holds or how long each lasts, so confirm the schedule with your recruiter.
Cornerstone's cases are its own litigation and regulatory work in disguise: its two published examples ask how much of a stock drop or a revenue decline can be tied to alleged wrongdoing. Its tips: break the question into smaller parts, look at it from several angles, take notes and ask questions, say your reasoning out loud and listen, since the interviewers are there to guide you. AI is not allowed in interviews or assessments.
A Washington manager quoted on Cornerstone's page tells candidates to weigh the behavioral side as much as, or more than, the case, and to be ready to describe past situations involving collaboration, conflict management and mentorship.
For associates, Cornerstone says the case interview also assesses your ability to direct a team. PhD candidates add three reference letters and a sample research paper to the application. Its associate pages describe no interview rounds, so ask the recruiting team what to expect.
Missing from Cornerstone Research's pages: how many interviews the final round holds and how long each lasts, who leads the case, how answers are scored, whether there is any online test, the application timeline and when offers are made.
11 Cornerstone Research-style interview questions
Cornerstone names no interview steps beyond its two rounds, so these are grouped by the kinds of question its published cases and interviewers point to. They are written in our words, not Cornerstone's.
Securities and event-study reasoning (Zilo-style)
Cornerstone's securities example asks you to isolate how much of a price drop an alleged wrong explains.
1A stock fell 40 dollars in the month a company disclosed bad news. Plaintiffs say the whole fall is damages. What do you question?
What a strong answer does: Says the fall mixes the effect of the news with everything else that moved the price, so it is a ceiling. Asks for the price the stock would have had without the news and plans to estimate it.
2What else could have moved the stock during that month?
What a strong answer does: Sorts causes into market-wide, industry-wide and company-specific, and within company news separates what relates to the alleged wrong from what does not. Takes the lists apart instead of one long brainstorm.
3A model says the stock should have closed at 50 dollars. It closed at 44. Is the 6 dollar gap caused by the news?
What a strong answer does: Asks how much the stock's daily gap from the model is normally, so it can say whether 6 dollars is unusual, not only different. Notes the gap is only evidence of news if the model was built on a clean earlier period.
Cornerstone's intellectual property example asks whether the claimed loss was caused by the alleged conduct, and how much profit it really was.
4A company claims a rival's copying cost it customers. What would have to be true for that to hold?
What a strong answer does: Writes the causal chain: the rival's customers would otherwise have bought from the claimant. Tests it with customer data, comparing who each company sells to and through which channels.
5The claimant's revenue fell by a fifth. What other explanations would you test, and with what evidence?
What a strong answer does: Names market, technology, price and competitor explanations and pairs each with a source, such as industry reports or analyst coverage. Does not stop at the first story that fits.
6Lost sales are 10 million dollars. How do you turn that into lost profit?
What a strong answer does: Says revenue is not profit and subtracts the costs that would have changed with those sales, mostly variable, while checking which fixed costs would not have. Gives the result as a range.
Cornerstone analysts build and read quantitative analyses, so expect numbers to be produced or questioned on the spot.
7Estimate how many people in a mid-sized city bought a newly recalled product during a six-month window.
What a strong answer does: Picks one route, such as households times the share owning the product times the share who bought in the window, states each assumption, and checks the result against a second route.
8A chart shows monthly sales jumping in one month. What do you ask before calling it an effect of a new policy?
What a strong answer does: Asks what else changed that month, whether the jump repeats each year, what a comparison group did and what the axis and scale are. Treats the jump as a question, not a finding.
9Two regional pharmacy chains merge, and prices at stores where they overlapped rose 4 percent. Is that harm?
What a strong answer does: Compares overlapping stores with similar stores that did not overlap, before and after, and checks whether costs rose for everyone. Says what finding would change the view.
Cornerstone says to be ready for collaboration, conflict management and mentorship.
10Tell me about a time you disagreed with a teammate about how to do an analysis.
What a strong answer does: Sets out the situation briefly, shows listening to the other view, says what the two of you did to settle it, and what the result was. Owns its part and does not make the other person the problem.
11Describe a time you helped someone learn something technical.
What a strong answer does: Names the person and what they needed, how you adapted your explanation, and how you knew it worked. Shows patience and what you learned from teaching it.
Easy ways to lose a Cornerstone Research interview
Forcing a strategy framework onto a damages question
Cornerstone says management consulting focuses on business strategy and economic consulting on the evidence behind expert testimony. A profit tree of revenue and cost buckets does not answer whether conduct caused a harm; state the claim and the but-for world first. Source: Cornerstone Research, read (opens in a new tab)
Answer each key question out loud before opening its hint, write a one-line answer with a number or a test, then compare with Cornerstone's summary and redo the step you missed.
Run the sample case on this page, then a full revenue-decline case with an AI interviewer, and read a worked competitive-response case. End each by saying what else could explain the fall.
How do I prepare for the Cornerstone Research case interview?
Start with Cornerstone's own interview page and its two published cases, Zilo (a securities case) and Duff (an intellectual property case), and work each cold with its key questions before reading the summary of how its team solved it. Cornerstone says every round has a different case and that the behavioral part matters as much as the case, so also prepare stories on collaboration, conflict and mentorship. On Road to Offer a free account includes 1 full AI-graded case, such as Thomas the Tank, a revenue-decline case that trains the same "what else explains the drop" reasoning; Dunmore Cargo Bikes and Northbay Grocers are worked cases with a model answer for every question and need no account.
Where can I practice Cornerstone Research-style case interviews for free?
Cornerstone's own two practice cases are the closest thing and are free on its site. Road to Offer has no litigation-economics case, so the sample case on this page is an original causal-attribution exercise you can run in your browser with no account, and Thomas the Tank is the nearest full case: you run it out loud with an AI interviewer or step by step, and your first full case is free with a free account. Dunmore Cargo Bikes (market sizing) and Northbay Grocers (competitive response) are free to read with a model answer for every question. They are business cases picked for the same habits, not Cornerstone's cases. Firm-style cases are practice cases written for Road to Offer or adapted from university consulting club casebooks, not the firm's own cases.
What are Cornerstone Research's practice cases and how should I use them?
Two, on its analyst case examples page: Zilo, a securities case (a stock fell from over 60 dollars to 16 after the company lost two customers worth 40 percent of sales, and plaintiffs claim 40 dollars a share), and Duff, an intellectual property case (a court found Duff used ACME's patented techniques, and ACME claims hundreds of millions of dollars of lost earnings). Each has key questions with a hint and a summary of the team's analysis. Answer each key question out loud before opening its hint, then compare your plan with the summary.
Does Cornerstone Research run case interviews?
Yes, but not strategy cases. Cornerstone says each round of its Analyst and Summer Analyst interviews includes a case, a different one each time, and its published examples are litigation problems: how much of a stock price drop, or of a patent holder's lost profit, can be tied to alleged misconduct. It says its cases are designed to be accessible to many academic backgrounds, so you need not be an economist, but you should be able to reason about causation and use numbers.
How many interview rounds does Cornerstone Research have?
Two for Analyst and Summer Analyst roles, per Cornerstone's interview page: two virtual first-round interviews of 30 minutes each, then a multi-hour final round in one of its offices. A case is part of each round. The page does not say how many interviews the final round holds or how long each lasts, and its associate pages describe no rounds, so confirm the schedule with your recruiter.
Is the Cornerstone Research case interview candidate-led or interviewer-led?
Cornerstone's pages don't say. Its tips point to a conversation: you are "presented with a different case at each round to review and discuss", you should ask questions and say your reasoning out loud, and you should listen because the interviewers are there to guide you. Practise both opening with your own plan and answering one directed question at a time.
Do I need economics or regression to pass the Cornerstone Research interview?
Not formally: a Cornerstone manager says its cases are designed to be accessible to a wide variety of experiences and academic backgrounds. Its own Zilo solution does use a regression of the stock price on a market index and an industry index, so be ready to explain in plain words what a comparison group or control does and why a stock's usual behavior is the baseline for what it would have done without the news.
Can I use AI during a Cornerstone Research interview or assessment?
No. Cornerstone says candidates may use AI to brainstorm, improve application materials and practise answers, but that during interviews and assessments it does not permit AI and expects your own thoughts, analysis and logic. Its associate application page repeats that your application should reflect your own work and voice.
What behavioral questions does Cornerstone Research ask?
Cornerstone does not publish a question list. A Washington manager on its interview page says to be ready to explain how you handled similar situations before, focusing on collaboration, conflict management and mentorship, and a Boston principal says to be an active listener. Prepare one specific story for each, with what you did and what changed, and say them out loud; Road to Offer's fit practice asks consulting fit questions with follow-up probes, 1 session free with a free account.
What do I send to apply to Cornerstone Research as an analyst?
A cover letter that says why you are interested and any office or location preference, a resume with your degree and expected completion date, and a college transcript, official or unofficial. Cornerstone says applications with missing materials will not be considered. US, UK and Belgium each have their own portal, and US-based students are not considered for UK or Belgium openings.
Is there an online test before the Cornerstone Research interviews?
Cornerstone's careers pages name none, and they say the process is two interview rounds. Its AI guidance mentions "interviews and assessments" without describing an assessment, so follow your invitation if you receive one and ask the recruiting team. Road to Offer has no Cornerstone test simulator, so none is listed here.
How is a Cornerstone Research case different from a McKinsey, BCG or Bain case?
Cornerstone says management consulting focuses on business strategy, while economic consulting supplies the data-driven foundation for expert testimony in court cases and regulatory proceedings. In practice the question changes from what the company should do to whether the conduct caused the harm someone claims and how much: you state the claim, the but-for world, the data and the other explanations. Bates White, another economic consulting firm, describes the same idea on its case interview tips page, with merger and cartel pricing topics.
