Compass Lexecon's posting for its 2027 US analyst class lists five responsibilities. Four are what you would expect: clean data, read documents, build tables, communicate findings. The fifth is "ensure accuracy of work product using established audit and quality control processes." That line is the interview.
Most candidates arrive ready to argue economics. The interviewer is not looking for a debate. She is checking whether every number you produce arrives with the assumption that drives it, the direction it could be wrong, and a trail a stranger could rerun. That is the job at a firm whose output ends up in expert reports that the other side's economists are paid to take apart.
You will leave with what the firm actually publishes about its process, a critical-loss calculation you can reproduce in four lines of arithmetic, and a line-by-line scoring of a weak answer against a strong one.
What the firm publishes, and what it does not
Compass Lexecon is unusually specific about its analyst process, at least in Europe. Its Berlin analyst posting, worded identically to the London one, says: "Our interview process for Analysts involves two stages. The first stage is one 60-minute interview with both motivational and technical questions. The second stage includes a written case study and a technical interview. All interviews are currently being conducted virtually." Analysts there typically join after a master's in economics, ideally in microeconomics, industrial organization, or econometrics.
The US 2027 analyst posting says nothing about stages but is exact about paperwork: cover letter, resume, unofficial transcript, and a writing sample, and "incomplete applications will not be reviewed." It also states the pay: a base salary of $105,000 plus overtime, with a discretionary bonus after the first year. As of that August 2026 posting, the Competition practice hires into Boston, Chicago, Los Angeles, and Washington, and the Finance practice into Chicago and New York.
| Stage | What the firm states | What it is scoring, in my reading |
|---|---|---|
| Application | Cover letter, CV, transcripts; the US adds a writing sample | Can you write a paragraph with a claim, evidence, and a caveat? The sample is your first written case |
| Stage 1 (Europe) | 60 minutes, motivational and technical | Why disputes economics rather than strategy work, and can you handle a micro question aloud |
| Stage 2 (Europe) | Written case study plus a technical interview | Can you produce a short, checkable note under time, then defend its weakest number |
| Not published | Case length, any numerical test, panel size, the PhD economist loop | Folklore. Your invitation email is the format; nothing else is |
The firm's students and graduates page describes the work ("conducting empirical analysis using software such as Excel and Stata, and drafting reports") and the firm (more than 700 employees in 23 offices) but nothing about interview content. Anything about a 45-minute pack or a three-person panel comes from candidate forums.
Why the interview is an audit
Consider what happened to the firm's own president in 2024. In FTC v. Kroger, Mark Israel, then president of Compass Lexecon, was the merging parties' economist on market definition and competitive effects. The FTC and its state co-plaintiffs moved to strike a "second corrected" version of his rebuttal report, which they said was served nineteen days late with "significant changes", along with a surrebuttal. The judge declined to strike either but let the government cross-examine him on the corrections at the hearing, and the court went on to block the merger in December 2024.
Corrected reports are routine. The point is what happens to them. A correction becomes a motion, a ruling, and a cross-examination, and the people who found, or missed, the original numbers were analysts and economists building tables. A correction is a litigation event, and the firm hires people who understand that before they start.
So when an interviewer asks "how do you know that?" for the third time, she is not bullying you and she is not testing composure. She is doing to your answer what opposing counsel will do to her tables in a deposition. Treat the question as a request for the trail, not an attack on the conclusion.
First translation: from the lawyer's question to a testable one
The client's question arrives in legal language: will this merger "substantially lessen competition"? Nobody can test that sentence. The analyst's first job is to turn it into questions data can answer, and the technical interview usually starts here.
The US agencies' 2023 Merger Guidelines give the structure: concentration screens that create a presumption, then market definition through the hypothetical monopolist test, which asks whether a single firm controlling every product in the candidate market could profitably impose a small but significant non-transitory increase in price, "often" five percent. The European Commission's 2024 market definition notice uses the same test with a range of 5 to 10 percent.
| Concept | The number the guidelines give | What candidates get wrong |
|---|---|---|
| Highly concentrated market | HHI above 1,800 | Treating it as a verdict. With a change above 100 it is a presumption, and presumptions get rebutted |
| Significant increase | Change in HHI above 100 | Forgetting the change is 2 × (share A) × (share B), so two 10-percent firms add 200 points |
| Share screen | Merged share above 30 percent, plus the 100-point change | Quoting the share without the change |
| SSNIP | Five percent in the US; 5 to 10 percent at the Commission | Treating it as a harm threshold. The guidelines say explicitly it is not one |
A London candidate should also know that the CMA takes a looser line. Its Merger Assessment Guidelines say market definition "is not an end in itself" and that the CMA expects to place "more emphasis on the competitive assessment as opposed to static market definition." (Updated 3 September 2026; only the efficiencies section changed.) An interviewer in London hears the difference between a candidate reciting US thresholds and one who knows which regulator the case is in front of.
A calculation you can reproduce: critical loss
A question I would expect in either technical round, with invented firms and figures.
Halvard Foods wants to buy Brightwater Creamery. Both sell premium ice cream pints to grocers in the same region. Counsel asks: is premium ice cream its own market, or is the market all ice cream? Your pack gives a net wholesale price of $4.50 a pint, a variable cost of $2.70, and a demand study estimating that a 1 percent price rise across all premium brands loses 1.6 percent of premium volume.
Step one, the margin: (4.50 − 2.70) / 4.50 = 40 percent.
Step two, the critical loss. For a price increase t and a margin m, the sales loss that leaves the hypothetical monopolist's profit unchanged is t / (t + m). With a 5 percent SSNIP: 0.05 / 0.45 = 11.1 percent. Lose less and the increase pays; lose more and it does not.
Step three, the predicted loss: 1.6 × 5 percent = 8.0 percent.
Step four, the comparison. 8.0 percent is below 11.1 percent, so a hypothetical monopolist of premium pints would find a 5 percent increase profitable, and premium ice cream passes as a candidate market. The 2023 Guidelines describe this "breakeven" comparison and call it informative.
| Margin | Critical loss at a 5% SSNIP | Critical loss at a 10% SSNIP |
|---|---|---|
| 30% | 14.3% | 25.0% |
| 40% | 11.1% | 20.0% |
| 50% | 9.1% | 16.7% |
| 60% | 7.7% | 14.3% |
Now the part that separates candidates. The interviewer asks where the 40 percent came from. If the $2.70 is a fully loaded cost from the management accounts, some of it is fixed, the true variable margin is higher, say 50 percent, and the critical loss falls to 9.1 percent. Your 8.0 percent still clears it, by one point rather than three. Give the elasticity a standard error of 0.5 and its upper end, 2.1, produces a predicted loss of 10.5 percent, which fails. The conclusion has not changed; its fragility has, and that is what the expert needs to know before signing.
One more check the sheet rewards. If each brand prices to maximize its own profit, a 40 percent margin implies through the Lerner condition that each brand alone faces demand with an elasticity of about −2.5. The candidate market as a whole must be less elastic than any single brand in it, because some switching is between premium brands, so −1.6 is coherent. Had the pack said −2.5, the right move is not to compute the predicted loss but to say the two numbers cannot both be right if these brands compete closely. That is the guidelines' margin-consistency requirement in one sentence.
Ned's rule. Never hand an interviewer a number without the input that moves it most. "About 11 percent, and it is mostly the margin definition" is an answer. "11 percent" is a guess with a decimal point.
Scoring a weak answer against a strong one
Same question, two candidates. This is how the difference reads from my side of the table.
| Dimension | Weak answer | Strong answer |
|---|---|---|
| Restating the question | "Is the merger bad for consumers?" | "Would a hypothetical monopolist of premium pints profitably raise price 5 percent? If yes, that is the market, and we move to effects" |
| Concentration | "HHI is over 1,800, so it is anticompetitive" | "HHI over 1,800 with a change over 100 creates a presumption. Testing it is the job" |
| Inputs | "Assume a 40 percent margin" | "40 percent from the parties' accounts; if it includes fixed overhead, the true margin is higher and the critical loss lower" |
| Direction of error | Silent | "The demand study includes promotion weeks, which overstate switching, so my predicted loss is probably high" |
| The sentence for the expert | "We recommend approval" | "A 5 percent increase is profitable unless more than 11 percent of volume is lost; best estimate 8 percent, range 5.5 to 10.5" |
| Under challenge | Defends the 40 percent | Recomputes at 50 percent aloud and reports the new gap |
The strong answers are not longer. They are traceable: every number has a parent, every conclusion comes with how it could be wrong, and nothing reads as advice to a chief executive. The product is what the economics supports, delivered to counsel and the expert; whether to merge is someone else's decision.
The written case: write for the reader who wants to catch you
The firm does not publish the length or format of its written case, so plan around your invitation. Whatever the length, the reader is a senior economist reading fast and looking for the number that does not tie out. Lead with the question and your answer in two sentences. Put the calculation in a small table with a source line under it, even if the source is "case pack, page 2." Add one paragraph on what would change your conclusion and by how much, then what you would ask for next: a longer price series, the parties' switching studies, costs split into fixed and variable. Skip the executive summary and the framework diagram.
| Block | Minutes of 60 | What it produces |
|---|---|---|
| Read, then write the question you are actually answering | 10 | One testable sentence |
| Calculation and two sensitivities | 25 | A table with a source line |
| Prose | 15 | Answer, fragility, next data |
| Reconcile every number in the text to the table | 10 | The audit |
The split is mine, not the firm's. The last block is the one candidates cut, and then the prose says 11 percent where the table says 12, which is the only thing the reader will remember about you.
The motivational half is technical too
Candidates treat the motivational half of stage 1 as a warm-up. "Why economic consulting rather than strategy consulting?" is a question about whether you understand the product: an argument a court or regulator can rely on, produced by a firm whose senior ranks include, in its own words, "several former chief and deputy chief economists of US and European competition and regulatory authorities."
Know the business you are joining. Compass Lexecon sits inside FTI Consulting's Economic Consulting segment, which reported in FTI's 2025 Form 10-K revenue of $720.8 million, down 16.5 percent on 2024, mainly on "lower demand for our M&A-related antitrust and non-M&A-related antitrust services," partly offset by financial economics work; billable headcount fell from 1,110 to 1,014 over the year. A candidate who can say "merger volume drives your hiring, and I still want this because the work is the same in a down year" has told the interviewer something true. One who says the firm is "growing fast" has not read the filings, a poor look for a job whose fifth responsibility is accuracy.
Prepare one story where you changed your conclusion because the data moved, and name the number that moved it. That is the behavioral version of the thesis.
Practice this today
Take 40 minutes. Invent two regional firms and a product, set a price, a variable cost, and a candidate-market elasticity, and write a one-page note answering "is this a market?" with the four steps above. Then spend 15 minutes as the other side's economist: find the input that flips the answer and write the cross-examination question that exposes it.
CoachNed's math drill covers margins and percentage changes without a calculator, most of what this interview asks of your hands. The synthesis drill trains the two-sentence answer that leads a written note, and behavioral practice with four follow-ups rehearses the motivational half of stage 1. CoachNed's cases are strategy cases, not industrial-organization cases, so use the drills for speed and for the sentence, not for the economics. Everything is open for seven days, no card; then $120 for a recruiting season or $49 a month.
CoachNed is independent and not affiliated with Compass Lexecon, FTI Consulting, or any other firm named in this guide.
Frequently asked questions
Does Compass Lexecon use case interviews like McKinsey or BCG?
No. Its European analyst postings describe a 60-minute motivational and technical interview, then a written case study and a technical interview. The content is applied microeconomics and econometrics, and the deliverable is an analysis counsel can rely on, not a recommendation.
How long is the Compass Lexecon written case study?
The firm does not publish the length or format. Specific claims about packet size or timing are candidate folklore. Your invitation email is the only reliable source, and it is reasonable to ask the recruiter.
What does Compass Lexecon pay analysts?
The US 2027 analyst posting, live as of August 2026, states a base salary of $105,000 plus overtime, with a discretionary bonus after the first year. European postings promise "competitive salary and benefits" and give no figure.
Do I need a PhD to work at Compass Lexecon?
Not for analyst roles. US analysts join with a bachelor's or master's in a quantitative field; in Europe, analysts typically hold a master's in economics and research analysts an undergraduate degree. Economist positions go to PhDs or people with several years in competition economics.
Sources
- Compass Lexecon, Students and Graduates — role types, Excel and Stata work, 700-plus employees in 23 offices, former chief economists on staff. Checked 2026-09-24.
- Compass Lexecon, 2027 Analysts - US (Workday JR261049) — application materials, the audit and quality control responsibility, $105,000 base plus overtime, locations. Checked 2026-09-24.
- Compass Lexecon, Analyst, Competition Practice, Berlin (Workday JR261307) — the two-stage interview description and master's-level profile. Checked 2026-09-24.
- FTI Consulting, Form 10-K for fiscal 2025 (SEC EDGAR) — Economic Consulting revenue of $720.8 million, down 16.5 percent, the reasons given, billable headcount 1,014 versus 1,110. Checked 2026-09-24.
- DOJ and FTC, 2023 Merger Guidelines, Guideline 1 — HHI and share thresholds for the structural presumption. Checked 2026-09-24.
- DOJ and FTC, 2023 Merger Guidelines, section 4.3 Market Definition — hypothetical monopolist test, five percent SSNIP, critical loss, margin consistency. Checked 2026-09-24.
- European Commission, Notice on the definition of the relevant market, C/2024/1645 — SSNIP normally 5 to 10 percent. Checked 2026-09-24.
- CMA, Merger assessment guidelines — market definition "not an end in itself," and the 3 September 2026 update limited to efficiencies. Checked 2026-09-24.
- FTC, Respondents' motion for leave to call one additional expert, In re Kroger/Albertsons — Mark Israel described as president of Compass Lexecon, testifying on market definition and effects. Checked 2026-09-24.
- Plaintiffs' motion to strike reports of Dr. Mark Israel, FTC v. Kroger (D. Or.) — corrected report served nineteen days late with "significant changes." Checked 2026-09-24.
- Order denying motion to strike, FTC v. Kroger (D. Or.) — court declined to strike and permitted cross-examination on the corrections. Checked 2026-09-24.
- FTC statement on the preliminary injunction, December 10, 2024 — the court blocked the merger. Checked 2026-09-24.