| The occupation, named precisely | The US Bureau of Labor Statistics calls this occupation Management Analysts, SOC code 13-1111. Use that code for pay data, employment counts and projections. "Management consultant", "strategy consultant" and "associate consultant" are job titles inside it, not separate occupations. Firms use their own ladder names: Business Analyst (McKinsey), Associate Consultant (Bain), Associate (BCG) for undergraduate entry; Associate (McKinsey), Consultant (BCG and Bain) for post-MBA entry. |
|---|---|
| License and certification | None. There is no license, no board exam, no registration and no certification that gates this job. Nobody checks a credential before you advise a client. What gates it instead: an application deadline, an algorithmically scored online assessment, and four to six live case and fit interviews. Certifications (PMP, Six Sigma, CFA) carry weight inside specific practices and are worth nothing as a substitute for case performance. |
| Four different hiring processes | Strategy firms (McKinsey, BCG, Bain, plus Kearney, Oliver Wyman, L.E.K., OC&C, Strategy&, EY-Parthenon, Roland Berger): calendar-driven, assessment plus live cases. Big 4 and large technology consultancies (Deloitte, PwC, EY, KPMG, Accenture): higher volume, organized by service line, situational judgment test plus recorded video plus a lighter case. Specialist boutiques (pricing, turnaround, healthcare, supply chain, energy): domain first, off-cycle, two or three interviews and often a written case. Internal strategy groups: an ordinary corporate loop. The documents and the preparation are not interchangeable. |
| How the calendar works, and how long the process takes | Strategy firms run one campus cycle per entry level per year. Undergraduate windows for the following year's start dates and summer internships open in late spring and close across the summer, with first rounds in late summer and fall and finals a few weeks later. MBA internship deadlines cluster in late fall and early winter; MBA full-time deadlines run through the fall. Experienced hiring is rolling, with volume concentrated early in the calendar year and again in September and October. From application to offer is typically four to ten weeks at a strategy firm, often two to three weeks at a boutique. Every date must be confirmed on the firm's own careers page; prep-industry trackers are secondhand and go stale. |
| The online assessment, by firm | McKinsey Solve: a game-based assessment, commonly including the Ecosystem Building and Redrock Study scenarios, scored on your answers and on process data from how you worked. The scenario mix varies by market and changes between cycles, so the practice material in your invitation is the only authority on what you will see. BCG: an online case experience delivered through a chat interface (Casey), roughly 25 to 30 minutes, or an office numerical and cognitive test; names differ by office, so ask the recruiter which one you are getting. Bain: many offices send the SOVA assessment, covering verbal, numerical, logical and situational judgment. Big 4: a situational judgment test plus a recorded one-way video interview. |
| What the interviews look like | First round: typically two interviews of 40 to 60 minutes, each roughly 10 to 15 minutes of structured experience questions, 25 to 30 minutes of case, then your questions. Final round: two or three interviews with partners, less structure, more judgment, sometimes a written case or a short presentation. McKinsey runs interviewer-led cases plus the Personal Experience Interview; BCG and Bain run more candidate-led cases. No calculator in a live case interview, ever. Some online assessments do provide an on-screen calculator, which is why assessment math and case math need separate practice. |
| What AI changed, stated narrowly | The junior deliverable layer compressed: literature and market research, benchmarking, interview-transcript coding, first-draft slides and model skeletons. The core did not: scoping an ambiguous problem, getting data that is not written down anywhere, and owning a recommendation in front of a client whose budget dies if you are right. Firms have started testing AI judgment directly. Press reporting says McKinsey has piloted a stage in which candidates use its internal assistant, Lilli, during a case and are graded on how they challenge and contextualize what it gives them. |
| Where real pay numbers live | Four sources, in this order. BLS OES and the Occupational Outlook Handbook for SOC 13-1111 give the occupation-wide picture; the most recent median sits a little above six figures, which mainly tells you that most people carrying this code do not work at a strategy firm. MBA employment reports published by business schools break out median base and signing bonus by industry and often by employer, reported to a common standard and in many cases audited, and are the best public consulting pay data that exists. Pay-transparency postings in Colorado, California, New York, Washington, Illinois and a growing list of other states carry a salary range on the posting itself. H-1B and LCA disclosure filings give base pay by employer, title and worksite. Blog salary tables are self-reported offer collections: directional, not evidence. |
Four different hiring processes share this job title
"Management consultant" describes who pays you (a client, for advice about how their organization should operate) and almost nothing about how you get hired. Four distinct hiring machines sit under the title, and candidates lose months by preparing for the wrong one. Work out which machine you are entering before you write a single resume bullet.
The first is the strategy firm. McKinsey, BCG and Bain define it, and a long tail of strategy houses runs a near-identical process: Kearney, Oliver Wyman, L.E.K., OC&C, Strategy&, EY-Parthenon, Roland Berger, the strategy arms of restructuring firms such as Alvarez & Marsal. The process is calendar-locked, gated by an online assessment, and decided by live case interviews. It is the most standardized hiring process in professional services, which is good news: it is knowable, practiceable and almost entirely free of mystery.
The second is Big 4 and large technology consulting: Deloitte, PwC, EY, KPMG, Accenture, Capgemini, Infosys Consulting. These firms hire an order of magnitude more people, organize around service lines rather than a general pool, and screen with a situational judgment test plus a recorded video interview before a human spends much time on you. The case is real but lighter on arithmetic gymnastics and heavier on "can we put you on a client project next month". Practice fit matters enormously: Human Capital, Technology Strategy, Deals, and Supply Chain and Operations are effectively different employers sharing a brand.
The third is the specialist boutique, and it is the most underrated door in the industry. Pricing firms, turnaround and restructuring shops, healthcare and life-sciences consultancies, supply chain specialists, energy and utilities advisors, government consultancies. They hire off-cycle, interview in two weeks instead of four months, weigh domain knowledge above school name, and often replace one interview round with a written case or a take-home. A partner who needs someone who understands pharmacy benefit managers or utility rate cases will hire the person who understands them.
The fourth is internal strategy: corporate development and strategy teams at large companies, "Strategy and Operations" or "BizOps" at technology firms, internal consulting groups at banks, insurers and hospital systems. The hiring process here is an ordinary corporate loop (recruiter screen, hiring manager, a case or a take-home analysis, a panel), and these teams disproportionately hire people who already have a consulting firm on their resume. If you want this job and do not have consulting experience, lead with the operating result you personally delivered, not with frameworks.
Two boundary cases worth naming, because conflating them in an application reads as confusion about the job. An implementation or solutions consultant at a software vendor is paid by the vendor to configure and help sell that vendor's product; the interview is a product demo and a technical screen, not a case. An independent consultant is a business owner, and the hiring process is sales: a referral network, a proposal, a rate card.
- Strategy firm: deadline, online assessment, two rounds of live cases plus structured fit. Decided by case performance and client presence.
- Big 4 and large tech consulting: situational judgment test, recorded video, often a single day of interviews including a lighter case and sometimes a group exercise. Decided by service-line fit and whether you can be staffed quickly.
- Specialist boutique: off-cycle application, a conversation with a practice lead, a written or take-home case, partner chemistry. Decided by domain credibility.
- Internal strategy: recruiter, hiring manager, take-home or live case, panel. Decided by whether you have personally moved an operating number.
- Public sector consulting (Booz Allen Hamilton, Guidehouse, LMI, ICF and the federal practices of the large firms): the same loop plus a security clearance requirement. An active clearance is sometimes the entire reason you get the interview, and sponsorship adds months to a start date.
The calendar is the real gate, and it is unforgiving
There is no license to earn and no exam to book, so the thing that actually eliminates people is a date. Strategy firms run a single annual campus cycle per entry level, and a missed deadline is not a delay of a few weeks, it is a year. This is the most common way a qualified candidate fails to get hired as a management consultant.
Note where those deadlines fall. Undergraduate windows at McKinsey, BCG and Bain for the following year's start dates and summer internships have in recent cycles opened in late spring and closed across the summer, the earliest in June and the latest in August, ahead of first rounds in late summer and fall. That means they close while most students are away from campus and no career office is staffed. People miss them for exactly that reason, not for lack of qualification. Do not take any date from an article, including this one: open the firm's own careers page, find the posted deadline for your entry level and office, and put it in your calendar in March.
MBA candidates run a different clock. Internship deadlines cluster in late fall and early winter, with interviews in January and February. Full-time deadlines run through the fall, with first rounds in October and November and finals into December. Advanced-degree candidates (PhD, MD, JD, postdoctoral) have their own recruiting track at MBB with its own dates, its own events and a stated tolerance for candidates who have never seen a case interview, offset by the same case bar on the day.
Experienced hiring is rolling rather than seasonal, but it is not uniform across the year. Volume concentrates early in the calendar year and again in September and October, because that is when practices know their staffing gaps for the period ahead. MBB's experienced-hire category for people with roughly zero to three years of work experience often has a formal deadline of its own. Boutiques hire whenever someone is needed, which is why they are the right target if you are reading this in the wrong month.
Networking has one specific mechanical function here, and candidates consistently misunderstand it. A referral from someone inside the firm gets your resume read by a human rather than filtered, and at some firms it triggers an assessment invitation. It does not pass the assessment for you, it does not shorten the case rounds, and a referral submitted after the deadline is worth nothing. Do the conversations in the first months of the year you apply, ask people what their team is actually being asked to do, and ask explicitly near the end: "would you be comfortable referring me when applications open?" Then apply on time.
If you are rejected, find the reapplication rule before you do anything else. Firms state theirs in their careers FAQ; a wait of a year or more is normal, it varies by firm, office, role and the stage you reached, and failing the online assessment usually carries its own defined lockout. Candidates get offers on a second or third attempt routinely. Candidates who reapply before they are eligible get filtered automatically and burn the attempt.
The stages, in order, and who is deciding at each one
Stage zero is the resume screen, and it is faster than you think. Recruiting teams, often supported by trained consultant readers, move through a stack in seconds per document. For campus roles at many offices, academic performance is used as a numeric filter before anyone reads a bullet. This stage is not a judgment of your potential, it is a sort.
Stage one is the online assessment. McKinsey Solve is a game-based assessment, commonly including the Ecosystem Building and Redrock Study scenarios, scored on your answers and on process data from how you worked: how you moved, how long you deliberated, how you handled a rule you had misread. The scenario lineup differs by market and changes between cycles, so treat the practice material in your invitation as the authority rather than a blog's description of last year's version. BCG sends an online case experience delivered through a chat interface, known as Casey, running roughly 25 to 30 minutes, or an office numerical and cognitive test. Bain commonly sends SOVA, covering verbal, numerical, logical and situational judgment. Big 4 firms front-load a situational judgment test and a recorded one-way video interview.
No firm publishes a pass rate, and the figures circulated by prep vendors are marketing, not data, so ignore them. What is worth knowing is this: the assessment is a real cut, it happens before a human has formed any view of you, and it is the stage candidates most often treat as a formality because the cases feel like the real test. Nothing you learn from case practice helps you with it.
Stage two is the first interview round: typically two interviews, 40 to 60 minutes each, with consultants and engagement managers. Each splits into a structured experience conversation of 10 to 15 minutes, a case of 25 to 30 minutes, and a few minutes of your questions. McKinsey runs interviewer-led cases, where the interviewer drives you question to question, alongside the Personal Experience Interview, which takes one story from your life and stays inside it for ten minutes. BCG and Bain run more candidate-led cases, where you are expected to drive, choose the next analysis and say why.
Stage three is the final round: two or three interviews with partners and principals, occasionally including a written case, a short presentation, or a group exercise at Big 4 firms. The case here is often looser, sometimes drawn from a live client problem the partner is working on this month, with less scaffolding and more tolerance for ambiguity. The question being answered in this round is not "can this person do analysis". It is "would I put this person in a room with a client who is paying us a great deal of money and is in a bad mood".
Offers arrive with a deadline stated in the letter, often a couple of weeks, and with a structure rather than a single number: base, performance bonus, signing bonus, relocation, and at the pre-MBA level a possible business school sponsorship that can be worth more than the rest of the letter combined. Base pay at campus levels is banded and effectively not negotiable. Start date and signing bonus have some give, particularly with a documented competing offer.
One structural difference worth planning for: Big 4 and large technology consultancies frequently compress everything into a single interview day of back-to-back sessions, sometimes including a group case where you are observed collaborating. Being right while steamrolling three other candidates fails that exercise. Boutiques do the opposite, stretching two or three conversations over a few weeks and giving you a written case to complete at home, which rewards structure and writing quality more than performance under live pressure.
Do case interviews still run the same way? Yes, with three real changes
The direct answer, because this is the question people arrive with: yes. The live case interview remains the center of hiring at every strategy firm and most Big 4 consulting practices in 2026-27. It is still conducted by a human, still roughly 25 to 30 minutes, still requires percentage and growth arithmetic done by hand with no calculator, and still ends with a recommendation delivered in under a minute. Anyone telling you the case interview is dead is selling something.
Change one: parts of the screen are now delivered and scored by software. BCG's Casey conducts a written case through a chat interface and evaluates what you type. McKinsey Solve has been algorithmically scored since it replaced the old Problem Solving Test. That has a practical consequence for preparation. In a chat-delivered case there is no interviewer reading your intent charitably, so your written answer has to carry the structure explicitly: state the approach, state the calculation, state the implication, in that order, in plain sentences.
Change two, and the one that is genuinely new: AI inside the interview itself. Press reporting says McKinsey has piloted an interview stage in which candidates use Lilli, the firm's internal assistant, while working a case, and the reporting is consistent on what was being graded. Not the output. The interaction: whether the candidate was curious, whether they were skeptical, whether they could take a plausible-looking AI answer, challenge it, notice what it did not know about this client, and adapt it. Bain's talent leadership has said publicly that candidates should expect AI to appear in the process, and BCG already runs an AI-delivered written case. Treat this as spreading but not yet universal, and treat the invitation email as the only authority on what you are allowed to use.
That change is easy to prepare for and almost nobody does. Practice working a case with an assistant open, out loud, narrating the thing that is actually being assessed: "I asked it for industry cost benchmarks, it gave me a figure with no source, so I would check that against the client's own P&L before using it, and until then I carry it as an assumption rather than a fact." That sentence is the whole skill. Candidates who paste model output into their answer and present it as their reasoning fail the stage, which is the point of the stage.
Change three: weight has shifted toward fit, judgment and presence. That follows from the first two. When the work of assembling a structured analysis has been meaningfully assisted by tooling, what distinguishes a good hire is the part that has not been: reading a room, knowing which question to ask the client's plant manager, sensing which recommendation the organization can actually absorb, being trusted with a difficult conversation. McKinsey's Personal Experience Interview has always probed this. Expect more of it everywhere, earlier.
Here is what has not changed at all, and it is most of the test. You must size a market from first principles without a calculator. You must build a structure that fits this client rather than a textbook, and say why those are the right three buckets. You must be comfortable being wrong in front of someone and recovering. You must do arithmetic out loud, round deliberately, and sanity-check your own result before the interviewer does. And you must synthesize: the answer, two reasons with numbers attached, one risk, in thirty seconds. If your preparation consists of asking a chatbot to generate frameworks, you have prepared for the wrong test.
One hard rule: using AI during an assessment where it is not permitted is a straight rejection, and firms state the rule in the invitation email. Several firms explicitly encourage AI for preparation and prohibit unapproved use during a scored assessment. Read the email rather than assuming either way.
- Still tested, unchanged: market sizing by hand, a bespoke structure with a stated hypothesis, mental arithmetic out loud, reading a chart and saying what it implies, a 30-second synthesis naming the risk.
- New: a chat-delivered written case (BCG's Casey), where your structure has to be explicit in writing because no interviewer is inferring it.
- New and not yet universal: an AI-assisted case stage, reported at McKinsey with its internal assistant, graded on skepticism and contextualization rather than on the model's output.
- Heavier than before: the fit and personal experience interview, and the partner round's judgment of client presence.
- Unchanged and absolute: no calculator in a live case, and no unapproved AI in a scored assessment.
The resume: what gets read in seconds, and what gets ignored
One page. Not two. This is one of the few industries where the one-page convention is enforced rather than merely preferred, and it holds until you have roughly a decade of experience or you are applying as a named industry expert. Reverse chronological, single column, standard section headings, exported to PDF. Firms run real applicant tracking systems (Workday, Avature and similar), and while consulting recruiters keyword-match far less aggressively than technology recruiters do, a two-column layout still parses into nonsense.
Education goes at the top for campus and early-career applications and moves below experience once you have three or more years of work. Include academic performance if it is strong, because many offices filter campus applications on it. No firm publishes a cutoff, and in the US a GPA below roughly 3.5 at a strategy firm means something else on the page has to do heavy lifting. Include standardized test scores where they are good: SAT, ACT, GMAT or GRE. Strategy firms still read them as a cheap numeric proxy, and leaving a strong score off is a self-inflicted wound. Omit a weak one without comment.
Every bullet does one job: it proves you personally caused a result. The shape is action, approach, quantified outcome against a baseline. "Renegotiated terms with 14 regional carriers by rebuilding the lane-level cost model, cutting outbound freight cost per shipment 11% against the prior-year baseline" survives interrogation. "Responsible for vendor management and cost optimization initiatives" does not, because nothing in it happened. Five bullets per role is the ceiling and three is often better.
Assume every number gets challenged in the interview, because it will be. An interviewer who sees 11% will ask how you calculated it, what the denominator was, what you would have found if you had excluded the one outlier carrier, and what you would do differently. Never put a number on the page that you cannot reconstruct from memory and defend. One indefensible figure contaminates the whole document.
Scope markers are the currency: money (revenue, budget, cost base, savings, capital), counts (people led, sites, SKUs, clients, countries, transactions), and time (how long it took, how long it had been broken). A result with no scope is unreadable, because saving 30% of something could be thirty dollars.
What is ignored, and in some cases counted against you: a skills block listing Excel, PowerPoint and teamwork; an objective statement; "detail-oriented" and "results-driven"; every club you joined with no role attached; anything from high school once you have a degree; a long technology stack on an application for a generalist role (two lines if it is relevant, and relevant means you used it to produce an analysis); a photograph in the US market; and visible AI phrasing, which experienced readers now recognize in a sentence and which reads as someone who outsourced the one document that is entirely about them.
Leadership with real responsibility outranks an additional internship. Something with a budget, a headcount and a consequence: a student organization that actually spent money, a varsity sport, military rank with people under you, a business with revenue, a research project you led. Consulting hires for the person who takes ownership when nobody told them to, and that is almost impossible to see in a list of internships.
If you are not from a target school, the arithmetic changes but the door does not close. Your resume has to carry more weight, which means one result so concrete that a partner recognizes the difficulty of it, plus a referral from an alum or practice lead who will vouch that you are worth 45 minutes. Apply to the boutiques and the Big 4 in the same cycle rather than treating MBB as the only outcome, and apply to the smaller offices as well as the headline ones, where the bar is the same and the candidate volume is not.
The cover letter is still read at McKinsey and at most boutiques, and ignored at some other firms. Check what each application asks for and write one where it is requested. One page, three paragraphs: why consulting and why this firm specifically, one proof story told in four sentences with a number in it, and why this office. Name a real practice and a real reason. A letter that would work for any of the three big firms with the name swapped tells the reader you did not care, and that is the only thing it tells them.
How to prepare, and how much preparation is actually enough
Case preparation has a clear diminishing-returns curve, and knowing where it flattens saves you two months. Twenty-five to forty full cases, done out loud with a live partner, is the range in which most people reach interview-ready. At least five of those should be with a current or former consultant who will tell you the unflattering truth. People who claim a hundred cases are usually avoiding the harder work: fit stories, mental math under pressure, and listening to a recording of themselves.
Start with the firms' own free material, because it is accurate and costs nothing. McKinsey publishes worked case examples and Solve practice material, BCG publishes an interactive case library, and Bain publishes practice cases and a case interview guide. The free case books from Wharton, Kellogg and Darden are dated in their business content and still excellent as structure drills. Case partner exchanges exist on every campus and in online communities; a paid coach is a real accelerator for the partner round and unnecessary for your first fifteen cases.
Solo reading does not transfer. A case is a conversation under time pressure with someone watching you, and the failure modes (going silent during arithmetic, narrating a structure with no hypothesis, drifting into a list) only appear when there is another person in the room.
Drill arithmetic fifteen minutes a day until it is boring. Percentage change in both directions, compound growth over three to five years, breakeven volume, weighted averages, unit conversions, and reading a chart to state the implication rather than the contents. Do it without a calculator, round aggressively, and say out loud that you are rounding: "I will call it 400 million to keep the numbers clean, which overstates it slightly." The interviewer is listening for whether you know what your own approximation costs you.
Build a story bank of six to eight experiences before your first interview, and build them properly. Each needs a situation, the conflict or resistance you personally faced, exactly what you said and did, a quantified outcome, and what you took from it. McKinsey's Personal Experience Interview will take one of these and spend ten minutes inside it: what exactly did you say to that person, how did they react, what did you do when that failed, how did you feel about it. Most candidates prepare a 90-second version and run out of substance at minute three. The themes to cover: personal impact (persuading someone who did not want to be persuaded), entrepreneurial drive (something you started that nobody asked for), and inclusive leadership (getting a group with conflicting interests to one decision).
Prepare the online assessment on its own terms rather than as a case warm-up. It rewards reading rules fast and accurately, allocating time, and not losing composure when the second scenario looks nothing like the first. Use the firm's own practice material first. One paid simulator is reasonable if money allows, but treat the prep industry's claims about exact scoring mechanics as inference, because the scoring is not published.
Learn the business basics cold, because a case is unworkable without them. Profit and loss mechanics, fixed against variable cost, contribution margin, unit economics, what a cost curve looks like and why, working capital, and how specific industries actually make money. Write one page of notes each on five industries you are likely to be cased on: retail, airlines or logistics, pharmaceuticals or payers, industrial manufacturing, and subscription software.
Record yourself doing three cases and watch them. It is unpleasant and it is the highest-yield hour in your preparation. Look for four specific things: whether you stated a hypothesis before diving in, whether you went silent while calculating, whether you ever said "so what this means is", and whether you sounded like someone a client would listen to.
- 25 to 40 live cases, at least 5 with someone who has done the job. Beyond 40 is usually avoidance.
- 15 minutes of no-calculator arithmetic daily, including chart interpretation.
- 6 to 8 fit stories built to withstand ten minutes of probing, not 90 seconds of telling.
- Assessment practice kept separate from case practice, starting with the firm's own material.
- One page of notes on five industries: how each makes money and what its cost base looks like.
- Three recorded cases, reviewed for hypothesis, silence during math, synthesis and presence.
Getting in without the campus door, in a crowded 2026-27 market
Be clear-eyed about the market you are entering as an experienced candidate. Through 2025 and 2026 the large firms trimmed their delivery layers while growing their AI, data and engineering practices; McKinsey's global headcount has been reported as falling over that period before the firm publicly committed to growing its North American workforce again. The consequence for you is not that nobody is hiring, it is who you are standing next to. The pool of available candidates with two to eight years of experience now contains an unusually large number of people who already have a strategy firm on their resume. Competing in that pool as a generic generalist is the hardest version of this job search that exists.
The move is to stop being a generalist on paper. Firms sell practices, and practices hire for what they are selling. Pick the thing you genuinely know and make the whole application about it: payer and provider economics, insurance claims operations, semiconductor or automotive supply chain, utility rate cases and grid investment, retail pricing and promotion, defense acquisition, post-merger integration, or deploying AI inside a regulated industry. A partner with a staffing gap in a named sector will interview a credible specialist from industry over a fourth generalist.
Routes that reliably work, in rough order of conversion. The advanced-degree track (PhD, MD, JD, postdoc) remains a real and separate door at MBB with its own calendar, its own events, and an acceptance that you have never seen a case, offset by exactly the same case bar on the day. Military transition programs at every large firm, which value demonstrated leadership under consequence and are practiced at translating rank into scope. Big 4 to strategy firm lateral moves after two to three years, which is the most common path for people who did not get an offer the first time. Industry expert hires into a specific practice. Internal strategy into consulting, which reads well because you have been the client. And implementation, operations and turnaround firms, which hire on what you can run rather than on where you studied.
Go at boutiques first and deliberately, not as a consolation. A forty-person pricing firm will take your call off-cycle, decide in two weeks, and have you in front of a client in your first month. Two years of that changes the MBB conversation entirely, because you arrive with engagement experience instead of an aspiration.
The networking that converts is specific. You want twenty minutes with a principal or partner in the practice you are targeting, not a recruiter and not a first-year analyst. Ask one question only they can answer: what is the work your team is being asked to do that it was not being asked two years ago. Listen. Then send a one-page note that shows you understood the answer and connects one thing you have personally done to it. That note is the artifact that gets forwarded. A connection request with a paragraph of flattery is not.
If you are targeting public-sector consulting, the clearance question dominates everything else. Booz Allen Hamilton, Guidehouse, LMI, ICF and the federal practices of the large firms post roles requiring an active Secret or Top Secret clearance, and an active clearance can be the single reason you are hired over a stronger candidate without one. Sponsorship exists but adds months to a start date, and "clearable" is not the same as cleared on a posting that says cleared.
Pay, progression, and where to get numbers you can rely on
Do not take a pay band from a blog, including this one. Consulting compensation varies by firm tier, level, office city and country by a factor of three or more, and the self-reported tables that dominate search results mix MBB offers with Big 4 offers and independent contractor rates. Here is where to look instead, in order of reliability.
Start with the BLS. The Occupational Employment and Wage Statistics series and the Occupational Outlook Handbook cover Management Analysts under SOC 13-1111, with national and state percentiles, industry breakdowns and a ten-year employment projection that has been running above the all-occupation average. The most recent median sits a little above six figures, and the most useful thing that figure tells you is what it excludes: the large majority of people carrying this occupation code do not work at a strategy firm, so the median describes the occupation, not the job you are interviewing for.
For strategy-firm pay specifically, the best public data is the MBA employment report. Business schools publish median base salary and median signing bonus by industry and frequently by employer, compiled to a common reporting standard and in many cases audited, with the sample size stated. Read the two or three most recent reports from schools that feed your target firms and you have the post-MBA band for consulting at that tier.
Then use the pay-transparency laws. Postings in Colorado, California, New York, Washington, Illinois and a growing number of other states must carry a salary range, and that includes consulting roles based in those states. This is the only source that gives you a firm-specific, level-specific, current number attached to an actual open role. Search the firm's careers site with the location filter set to one of those states even if you intend to work elsewhere. H-1B and LCA disclosure data is the fourth source, giving base salary by employer, title and worksite for sponsored roles, which is useful for calibrating levels at large firms.
Understand what the offer is made of, because base is not the only number that matters. A campus offer typically bundles base, a performance bonus that depends on your rating, a signing bonus, relocation, and at the pre-MBA level a potential business school sponsorship whose cash value can exceed everything else in the letter. Base at campus levels is banded by policy and effectively not negotiable; signing bonus and start date have a little give, particularly with a documented competing offer.
Progression is explicit and fast, and it is the industry's defining feature. Pre-MBA analyst tenure runs roughly two to three years before business school or an exit. Post-MBA, the ladder is consultant, engagement or project leader, principal, partner, with named promotion windows and continuous written feedback. Up-or-out is real and is not a scandal: most people leave, the firm expects them to, and the exits (private equity, corporate development, startup operating roles, industry strategy, portfolio company leadership) are a genuine part of the compensation. Weigh a firm's exit pattern in your practice area as seriously as its salary.
Finally, the shape of 2026-27, stated as narrowly as the evidence allows. The entry pyramid is narrower than it was: the research, benchmarking and first-draft deck work that filled a first-year's day is the work AI tooling absorbed first, and client pressure toward outcome-linked pricing reduces the appetite for billable junior hours. At the same time, the large firms report that a substantial and growing share of their staff carry AI, data or engineering titles, and those practices are where headcount is being added. A generalist analyst offer is still a real and good offer. But the work you will be judged on starts higher up the value chain than it did for the person three years ahead of you, and the interview process now looks for that earlier.
What a management consultant must know about AI in 2026-27
Say the narrow true thing first: AI has not automated the core of consulting, and it has substantially automated the layer of work that juniors used to be paid to do. Both halves of that sentence matter, and most writing on this topic gets one of them wrong.
What has not moved. Sitting with a chief operating officer who cannot articulate why their fulfillment costs have drifted, and leaving the room with the actual question defined. Walking a plant floor and noticing the thing that is in no dataset. Getting a number out of a regional manager who has a reason to keep it quiet. Deciding which of three defensible recommendations the organization can actually absorb. Standing behind an answer in front of a board when someone's budget dies if you are right. None of that is a text-generation problem, and nobody has shipped a product that does it.
What has moved, concretely. Literature and market research synthesis. Competitor and cost benchmarking. Coding and clustering a stack of expert interview transcripts. First-draft slides, including the formatting labor that used to consume analyst evenings. The skeleton of a financial model. Cleaning and reshaping a client data extract. That was a large share of a first-year's billable day, and it now takes hours instead of weeks.
The firms built their own tooling for exactly this work, and the tools are part of the job now. McKinsey's Lilli is firm-wide and pitched internally on the research and synthesis time it saves. BCG runs internal assistants, including one aimed at the slide-formatting grind, alongside a large AI build practice. Bain has a services alliance with OpenAI and internal assistants of its own. Assume that on day one you will be issued an internal assistant, a policy document about what you may put into it, and an expectation that you already know how to use one well.
The commercial consequence is the one candidates miss. Clients know what these tools cost and what they can do, which pushes engagement pricing toward outcomes rather than staffed hours, which in turn reduces the number of junior hours a firm wants to sell. That is the mechanism behind a narrower entry pyramid, not a general collapse in hiring: McKinsey has publicly committed to growing North American headcount while the same period saw its delivery layers shrink. The job is still there. It starts further up.
The second consequence is what you will be selling. A large share of what strategy clients are buying in 2026-27 is AI-related: where to deploy it, what it will actually cost, what it will break, how to reorganize around it, and how to govern it. You do not need to build models to work on this. You do need to be able to say, credibly and specifically, what a retrieval system costs to run at this client's document volume, why their data is not ready, which three processes are worth automating first, and what the failure mode looks like in a regulated environment. A candidate who can discuss that concretely is worth more than one who has memorized an extra framework.
And the thing that is now tested directly: your judgment about machine output. Press reporting says McKinsey has piloted an interview stage where candidates use Lilli during a case and are assessed on how they interrogate what it gives them. The behavior being graded is skepticism applied out loud. Where did that number come from, does it fit what we know about this client, what would I check before putting it on a slide. Candidates who treat the assistant as an oracle fail that stage. Candidates who treat it as a fast junior analyst whose work must be checked pass it.
Tracing an AI-generated number back to a primary source
An unsourced figure in a client deck is a career event in consulting. The tooling produces plausible numbers with no provenance, which makes verification the junior skill that now carries the most weight. It is also the specific behavior reported as being graded in McKinsey's AI-assisted interview pilot.
Show it: In the case, say it out loud: "that benchmark has no source attached, so I would carry it as an assumption and verify it against the client's own P&L before it goes in front of anyone." On the resume, one bullet where you caught an error in an analysis, with what the error would have cost.
Prompting against a client-specific fact base rather than general knowledge
Generic output is worthless in consulting, because the answer always depends on this client's cost structure, contracts and constraints. The consultants who get leverage from these tools feed them the engagement's own documents and data, then interrogate the result.
Show it: Describe one analysis where you grounded a model in a specific document set or dataset, and what you got that a general answer would have missed. Be able to explain how you checked that it was actually using your material rather than improvising.
AI-assisted deck drafting with the storyline still yours
Tools now handle layout, formatting and first drafts, which has raised the floor on polish and left the storyline as the differentiator. A deck whose pages are beautiful and whose argument does not progress reads worse now, not better, because the polish removes the excuse.
Show it: Bring a work sample if you have one you can share: an action-titled page sequence where the titles alone read as a complete argument. In interview, when asked how you would present, give the title sequence rather than the page contents.
Analysis in code rather than only in spreadsheets
The analyses junior consultants are asked for routinely exceed what is comfortable in Excel: client extracts with millions of rows, repeated reshaping, a sensitivity run across dozens of scenarios. AI assistance makes Python and SQL reachable for non-engineers, which has moved the baseline expectation up rather than removing it.
Show it: One resume line naming the tools and the analysis they produced ("rebuilt the lane-level freight model in Python across 2.4m shipment records"), plus the ability to describe your data cleaning decisions when asked. Naming a language you cannot discuss is worse than omitting it.
Costing and scoping an AI deployment honestly
This is a large part of what clients are currently buying, and the market is full of advisors who cannot answer the hard questions. Being able to say what a system costs per query at volume, where the data work actually sits, what the human review step costs, and which use case fails first is immediately commercial.
Show it: Pick one AI use case in your target industry and walk through it unprompted in three minutes: the process, the current cost, the realistic reduction, the implementation risk, the governance constraint. Use it as your answer to "what interests you about this practice".
Client data handling and confidentiality discipline
Pasting client material into a consumer AI tool breaches the engagement contract at every serious firm, and it is the most common way a junior creates a real problem. Firms deploy internal assistants specifically so this does not happen, and they are listening for whether candidates understand why.
Show it: When AI comes up in interview, name the boundary without being asked: approved internal tooling only, no client-identifying material in external tools, and check the engagement's own data agreement. One sentence establishes that you can be trusted with a dataset.
Working a case with an assistant open, out loud
McKinsey has reportedly piloted this, Bain's talent leadership has said to expect AI in the process, and BCG already runs an AI-delivered written case. The mechanics of narrating your reasoning while using a tool are unfamiliar and visibly awkward if you have never practiced them.
Show it: Practice five cases with an assistant open and a partner listening, narrating what you asked, what you got, what you doubted and what you would check. The grading is on the interrogation, so make the interrogation audible.
Synthesis, which is the part no tool does for you
A machine will give you ten defensible observations. Choosing the one the client should act on this quarter, stating it in a sentence, and naming the risk is the judgment the whole hiring process screens for, and it became more decisive as the analysis layer got cheaper.
Show it: End every practice case with a 30-second recommendation: the answer, two supporting reasons with numbers, the biggest risk, and the first thing you would do next week. Do it even when the interviewer does not ask.
What a screen is looking for
These are the terms that a resume screen, human or automated, is matching against for this role. Use the ones that are true of you, in the words the posting uses.
- Management consulting
- Strategy consulting
- Management analyst
- Case interview
- Market sizing
- Hypothesis-driven problem solving
- MECE
- Issue tree
- Driver tree
- Synthesis
- Executive summary
- Action titles
- Storylining
- Client management
- Stakeholder management
- Engagement management
- Workstream lead
- Commercial due diligence
- Market entry strategy
- Growth strategy
- Cost reduction
- Cost transformation
- Operating model design
- Target operating model
- Post-merger integration
- Carve-out
- Synergy assessment
- Benchmarking
- Profitability analysis
- Unit economics
- Contribution margin
- P&L analysis
- Financial modeling
- Three-statement model
- Scenario analysis
- Sensitivity analysis
- Business case development
- Pricing strategy
- Price elasticity
- Go-to-market strategy
- Customer segmentation
- Churn analysis
- Sales force effectiveness
- Supply chain optimization
- Demand forecasting
- Capacity planning
- Make-or-buy analysis
- Working capital improvement
- EBITDA improvement
- Zero-based budgeting
- Process improvement
- Lean
- Six Sigma
- Change management
- Program management
- PMO
- Workshop facilitation
- Executive presentation
- Board-level communication
- Expert interviews
- Primary research
- Secondary research
- Competitive analysis
- Voice of customer
- Survey design
- Data analysis
- SQL
- Python
- Excel
- PowerPoint
- Alteryx
- Power BI
- Tableau
- Generative AI
- Large language models
- AI strategy
- AI use case prioritization
- Retrieval-augmented generation
- Digital transformation
- Vendor selection
- RFP
- Statement of work
- Proposal writing
- Recruiting and mentoring
- Security clearance
- Secret clearance
- Healthcare payer and provider
- Life sciences commercial strategy
- Private equity due diligence
- McKinsey Solve
- BCG Casey
- SOVA assessment
- Personal Experience Interview
- MBA
- CFA
- PMP
Mistakes that cost people this job
Discovering the application window after it closed, because strategy-firm campus deadlines fall across the summer while campus is empty.
In March, open the firm's own careers page, find the posted deadline for your entry level and office, and put it in your calendar with reminders a month and a week out. A missed strategy-firm deadline costs a full year, and no amount of networking reopens it.
Treating networking as the gate, and spending two months on coffee chats while under-preparing the assessment.
Understand the mechanic: a referral gets your resume read by a human, and at some firms triggers an assessment invitation. That is all it does. Budget roughly a fifth of your effort to networking and the rest to the assessment, the cases and your story bank.
Under-preparing the online assessment because the case interviews feel like the real test.
It is a real cut that happens before any human has a view of you, and nothing you learn from case practice helps with it. Use the firm's own practice material, learn the time structure cold, and practice staying composed when the second scenario looks nothing like the first.
Preparing for last cycle's version of the assessment, from a blog description of the scenarios.
The scenario mix in McKinsey Solve differs by market and changes between cycles, and BCG's assessment differs by office. Treat the practice material and rules in your invitation email as the only authority, and ask the recruiter which assessment you are being sent.
Opening a case with a memorized framework (the four Ps, Porter's five forces, a generic profitability tree) regardless of what was asked.
Build a structure from the specific question using the client's own economics, then say why those are the right buckets and which one you would look at first and why. An interviewer can hear a memorized framework in four seconds, and it reads as someone who is not thinking.
Going silent while doing arithmetic, then announcing a number.
Narrate it. State the approach, say what you are rounding and in which direction, carry the units, then sanity-check your own answer out loud before the interviewer does. The reasoning you say out loud is what is being assessed; the number alone is not.
Finishing the analysis and listing what you found instead of recommending something.
Always close in 30 seconds: the recommendation, two supporting reasons with the numbers attached, the biggest risk, and the first action next week. Do it unprompted, and do it even if you did not finish the math.
Bringing fit stories that survive 90 seconds of telling and collapse at minute three.
Build six to eight stories deep enough for ten minutes of probing: exactly what you said, how the other person reacted, what you did when the first attempt failed, the quantified outcome. McKinsey's Personal Experience Interview goes further into a single story than almost any candidate expects.
A two-page resume full of responsibilities, with no number that could be challenged.
One page, five bullets per role maximum, each one action plus approach plus a quantified result against a stated baseline. Then make sure you can reconstruct every number on the page, including the denominator and the exclusions, because you will be asked.
Letting an AI assistant write the resume or cover letter.
Write it yourself and use a tool only to cut words. Experienced readers identify generated phrasing in a sentence, and the document that is entirely about you is the worst possible thing to outsource. A letter that would work with the firm name swapped out is a wasted page.
Using an AI tool during an assessment that prohibited it, or pasting model output into an AI-assisted case as your own reasoning.
Read the rule in the invitation email and follow it exactly. Where AI use is invited, the grading is on how you challenge the output, so make your skepticism audible: where the number came from, whether it fits this client, what you would verify first.
Applying only to McKinsey, BCG and Bain, and treating anything else as failure.
Apply to strategy boutiques, Big 4 strategy practices and specialist firms in the same cycle. Two years at a forty-person pricing or turnaround firm, with client exposure in month one, changes the MBB conversation entirely because you arrive with engagement experience instead of an aspiration.
Applying as an experienced hire with a generalist pitch into a market full of ex-strategy-firm candidates at the same level.
Lead with a domain the firm sells: payer economics, semiconductor supply chain, utility rate cases, retail pricing, defense acquisition, post-merger integration, AI deployment in a regulated industry. A practice lead with a staffing gap hires the credible specialist over the fourth generalist.
Reapplying before you are eligible and getting filtered automatically.
Find the firm's stated reapplication window in its careers FAQ, note that failing the online assessment usually carries its own lockout, and apply on the first eligible cycle with evidence of what changed. Second-attempt offers are routine; premature applications burn the attempt.
Questions people ask
Are case interviews still used in 2026, or has AI replaced them?
Case interviews are still the center of management consulting hiring in 2026, conducted live by a human, typically 25 to 30 minutes, with the arithmetic done by hand and no calculator. Three things changed around them: part of the screen is now delivered and algorithmically scored by software (BCG's Casey chat case, McKinsey Solve), press reporting says McKinsey has piloted a stage where candidates use the firm's internal AI assistant during a case and are graded on how skeptically they interrogate its output, and the fit and personal experience portion carries more weight than it did. The structure, the hypothesis, the mental math and the 30-second synthesis are unchanged. Firms state their AI rule in the invitation email: several encourage it for preparation and prohibit unapproved use during a scored assessment.
Do I need an MBA to become a management consultant?
No. Strategy firms hire undergraduates directly into analyst-level roles (Business Analyst at McKinsey, Associate Consultant at Bain, Associate at BCG), and they hire experienced candidates from industry, the military and advanced-degree programs without an MBA. An MBA is the standard entry point for the post-MBA level, which sits one rung up and pays accordingly, and many firms will sponsor a pre-MBA analyst's business school tuition in exchange for a return commitment. There is no license or certification for this occupation at all.
What online assessment will I have to take, and how hard is it?
McKinsey sends Solve, a game-based assessment commonly including the Ecosystem Building and Redrock Study scenarios, scored on your answers and on process data from how you worked; the scenario mix varies by market and changes between cycles, so the practice material in your invitation is the authority. BCG sends an online case experience delivered through a chat interface, known as Casey, running roughly 25 to 30 minutes, or an office numerical and cognitive test. Bain commonly sends SOVA, covering verbal, numerical, logical and situational judgment. Big 4 firms send a situational judgment test plus a recorded one-way video interview. No firm publishes a pass rate and the figures circulated by prep vendors are marketing rather than data, but this is a real cut that happens before a human has formed a view of you, and nothing you learn from case practice helps with it.
Is consulting hiring recovering in 2026-27?
It is uneven rather than uniformly up or down. Large firms trimmed their delivery layers through 2025 and 2026, which put an unusually large number of experienced candidates with strategy-firm backgrounds into the market, while AI, data and engineering practices are where headcount is being added. McKinsey's global headcount has been reported as falling over that period before the firm publicly committed to growing its North American workforce again. Campus entry offers are still real and still worth chasing; the mid-level generalist market is the crowded one, which is why experienced candidates should apply with a named domain rather than a generalist pitch.
How many practice cases do I actually need?
Twenty-five to forty full cases done out loud with a live partner, including at least five with a current or former consultant who will give you honest feedback. Start with the firms' own free material: McKinsey publishes worked examples and Solve practice, BCG publishes an interactive case library, Bain publishes practice cases. The returns flatten after roughly forty, and candidates claiming a hundred are usually avoiding the harder preparation: fit stories deep enough to survive ten minutes of probing, daily no-calculator arithmetic, and watching a recording of themselves.
Can I get a consulting job without going to a target school?
Yes, and the arithmetic of how is specific. Your resume has to carry more weight, which means one quantified result concrete enough that a partner recognizes how hard it was, plus a referral from an alum or practice lead willing to vouch for 45 minutes of someone's time. Apply to smaller offices as well as headline ones, where the bar is the same and the candidate volume is not. Apply to boutiques and Big 4 strategy practices in the same cycle, and consider the lateral route: two to three years at a Big 4 or specialist firm is the most common path into a strategy firm for people who did not get in on the first attempt.
Will AI eliminate the entry-level consulting job?
It has already removed a large part of what that job used to consist of, without removing the job. Research synthesis, benchmarking, interview-transcript coding, first-draft slides and model skeletons compressed from weeks to hours, and client pressure toward outcome-linked pricing reduces how many junior hours a firm wants to bill. The firms are still hiring juniors and saying so publicly. The honest summary: the entry pyramid is narrower and the work you will be judged on starts higher up the value chain than it did three years ago, which is why interviews now test judgment, skepticism about machine output and client presence earlier.
What does a management consultant earn?
Use sources rather than a band, because pay varies by firm tier, level and city by a factor of three or more. The BLS Occupational Outlook Handbook and OES series cover this occupation as Management Analysts, SOC 13-1111, with national and state percentiles; the most recent median sits a little above six figures, which describes the occupation as a whole rather than strategy-firm pay. For strategy firms specifically, the most reliable public data is business school MBA employment reports, which publish median base and signing bonus by industry and often by employer to a common reporting standard. Pay-transparency postings in Colorado, California, New York, Washington, Illinois and other states give current, firm-specific, level-specific ranges on real open roles. H-1B and LCA disclosure filings give base pay by employer, title and worksite.
Is Big 4 consulting real management consulting?
Yes, and it is the majority of the industry by headcount. Deloitte, PwC, EY, KPMG and Accenture do strategy, operations, technology and human capital consulting for the same clients, with a higher proportion of implementation work and a much larger intake. The hiring process differs: a situational judgment test and a recorded video interview up front, often a single day of back-to-back interviews, a case that is less arithmetic-heavy, and sometimes a group exercise where you are observed collaborating. Service line matters more than the firm name when you apply, and two to three years there is the most common lateral route into a strategy firm.
How do I move from industry into consulting at thirty or forty?
Stop applying as a generalist. Pick the domain you genuinely know, target the practice that sells it, and get twenty minutes with a principal or partner in that practice rather than a recruiter. Ask what work their team is being asked to do now that it was not two years ago, then send a one-page note connecting something you personally delivered to that answer. Specialist boutiques, implementation and turnaround firms, and public sector consulting (where an active security clearance can be the deciding factor) convert far better at this stage than cold applications to generalist MBB roles. You will still have to clear the same case bar on the day, so prepare 25 to 40 live cases regardless of how senior you are.
Put this on a resume in about a minute
Paste your history once and point it at the Management Consultant posting you are looking at. No account, no card.
Build my resume free More roles