Guidehouse Case Interview Examples: Health, Energy and Government Cases Solved (2026)

Guidehouse has been owned by Bain Capital since December 2023, and its cases follow its clients: hospitals, utilities, state agencies, defense programs and banks. The CaseBasix guide to Guidehouse interviews describes a 30 to 45 minute live case where you clarify, structure, analyze data and recommend, with a take-home or an Excel exercise for some roles. Guidehouse itself publishes no interview cases on the careers pages we checked, so the only way to see the format is to practice it.

This page gives you eight practice cases with the math worked step by step, every one an original case written for Road to Offer and labelled that way. They are built from the five industries Guidehouse lists on its about page, and they all end the way a Guidehouse engagement ends, in a recommendation with a number and a risk attached. The Guidehouse case interview guide owns the process, rounds, clearance and behavioral questions; this page is the practice.

Guidehouse practice cases at a glance

How a Guidehouse case works: the interviewer presents a business problem from the practice you applied to, and you clarify, structure, analyze data and recommend. Guidehouse does not publish its case format, so the facts below are labelled by source.

• Official, interview stages: Guidehouse's internship page lists an initial interview (on campus or online) and a final round with team members, then the offer, and does not mention cases. Its Gateway to Guidehouse Life Sciences page says candidates get one behavioral and one case interview with Life Sciences team members.
• Reported, case length and style: CaseBasix reports a 30 to 45 minute live case and advises practicing candidate-led cases. Guidehouse does not say who leads.
• Reported, other formats: a 24 to 48 hour take-home case for some roles, and an Excel exercise for some specialized or technical roles (CaseBasix).
• Official, industries: Communities, Energy and Infrastructure; Defense and Security; Financial Services; Health; and State and Local Government (Guidehouse about page). CaseBasix names healthcare, energy, life sciences and the public sector as the common case topics.
• Reported, math: growth rates, break-even points and basic financial metrics (CaseBasix).

Start free now: National Health System is an interviewer-led public sector operations case adapted from the INSEAD casebook: the Sydney Department of Health has found that people over 60 are unhappy with the time and quality of the hospital services they receive. It has 8 questions, 1 with math, 1 exhibit, and fewer than 10 completed runs so far, so no median time yet. An AI interviewer runs it out loud, shows each exhibit when you reach it and scores the debrief. Your first full case is free when you create an account. A free account also gets one full run in every simulator inside the app: one complete run of each of our four McKinsey Solve practice formats (Red Rock, Sea Wolf, SFL Classic and SFL Island), a full Casey-style case, a full CCA-style diagnostic and a complete Bain SOVA-style diagnostic.

Case (style, industry)	Quant difficulty	Qualitative difficulty	The math you run	Median time
National Health System (Independent, health services). 8 (1 with math), 1 exhibit.health servicesQuant difficulty mediumQualitative difficulty mediumFewer than 10 runsTry free	medium	medium	Older share, visits, specialists needed	Fewer than 10 runsFewer than 10 runs	Try free
EV Charging Hub Profitability (McKinsey, energy and retail). 8 (3 with math), 3 exhibits.energy and retailQuant difficulty mediumQualitative difficulty medium17 min· 45 voice runsTry free	medium	medium	Segment profit bridge, percentage change	17 min45 voice runs17 min, 45 voice runs	Try free
City Bike Share Concession (McKinsey, urban transport). 9 (4 with math), 3 exhibits.urban transportQuant difficulty hardQualitative difficulty medium32 min· 11 voice runsTry free	hard	medium	Revenue by zone, 15-year concession value	32 min11 voice runs32 min, 11 voice runs	Try free
New Rubber Plant Investment (BCG, manufacturing). 8 (2 with math), 2 exhibits.manufacturingQuant difficulty mediumQualitative difficulty hardFewer than 10 runsTry free	medium	hard	Train shipping capacity, monthly profit, break-even period	Fewer than 10 runsFewer than 10 runs	Try free

Two are adapted from consulting club casebooks (INSEAD and Wharton); the rest were written for Road to Offer. The library has no case in Guidehouse's own style, so these are matched by case type to the industries Guidehouse lists, and none is Guidehouse's own material. New Rubber Plant Investment runs in guided mode only. More are in the case library, 92 cases in all. Every debrief scores seven categories: structure (25% of the score), quantitative analysis (20%), business judgment (15%), and problem understanding, exhibit reading, communication and synthesis (10% each), with feedback on every question. Median times come from completed runs between June 25 and September 22, 2026, and count runs, not people.

What does a Guidehouse case study interview look like?

Candidates report a 30 to 45 minute live case, with a take-home case or an Excel exercise for some roles, and Guidehouse does not publish the format itself. In the live version the interviewer presents a problem from the practice you applied to, and you clarify, structure, analyze data and recommend.

• Reported, live case versus take-home: CaseBasix describes the live case as a conversation of 30 to 45 minutes. It describes the take-home as due in 24 to 48 hours, returned as a slide deck or Excel output that mimics a client deliverable.
• Official, what Guidehouse confirms: its Gateway to Guidehouse Life Sciences page says invited candidates have one behavioral interview and one case interview with team members, and it gives no length or format for the case.
• Reported, tailoring by practice area: CaseBasix says to prepare for the practice you applied to and lists healthcare, energy, government and public sector, and strategy and transformation, with topics such as health systems transformation, decarbonization, public-private partnerships and enterprise change. Soreno advises running cases in the industry of your team and ending each one with a clear so-what.
• What to do with that: ask your recruiter which format your role uses, then match your reps to it. Cases 1 to 7 below are the live shape, and Case 8 is the take-home shape.

Does Guidehouse publish official practice cases?

No: the Guidehouse careers pages we checked contain no practice case, so these examples are original. The early career page describes a 10-week internship, a consulting bootcamp and team placement matched to your interests. The internship page lists five stages, from joining the talent community to the offer, with an initial interview and a final round with team members, and it does not mention cases. Only the Gateway to Guidehouse Life Sciences page says what the interviews contain: one behavioral interview and one case interview with Life Sciences team members, on a page written for the 2025 program cycle.

What Guidehouse does publish is client case studies, and they show the kind of problem your case is dressed up from:

• Grid resilience. For CenterPoint Energy, the main electric utility in the greater Houston area, Guidehouse supported a $2.7 billion resilience plan that the Public Utility Commission of Texas approved in August 2025. CenterPoint projects around $6.50 of benefit for every dollar spent. Case 4 below uses that benefit-cost logic with made-up numbers.
• Federal aid recovery. Guidehouse says it helped more than 350 providers and more than 20 state and local government entities obtain and substantiate more than $10 billion of federal COVID-19 aid. Case 8 is a grant allocation problem in that spirit.
• Internal efficiency. For a federal agency department, Guidehouse interviewed nine project managers one on one to decide whether to keep improving, combine or drop two internal resource platforms. Cases 2 and 7 are about throughput and backlog, the same kind of operational question.

Read two or three of these before a Guidehouse interview for vocabulary, and do not treat them as prompts.

How do these eight cases map to Guidehouse's industries?

Each case sits in one of the five industries Guidehouse lists, so you can start with the practice you applied to. The library cases in the table above are the ones to run out loud first.

All numbers in Cases 2 to 8 are assumptions made up for practice. Case 1 uses the reference inputs stored with the library case. State your own assumptions out loud in an interview, exactly as the cases do.

The loop to run on every case

Read only the prompt, write your structure and your questions, then open the solution. The cases below use the same five parts so you can compare each part to your own:

1. Prompt: the client, the decision and the constraint.
2. Clarifying questions: what you would ask, and what the interviewer would plausibly say.
3. Structure: two to four branches that you can test, not a template.
4. The math: every step substituted, with units.
5. Recommendation: a decision, one number, one risk and a next step.

If you have never run a case, the government and public sector case interview, healthcare case interview, energy case interview and operations case interview pages teach the frameworks. This page is for reps.

Case 1: A public health service and older patients

Prompt. The Sydney Department of Health has found that people over 60 are unhappy with the time and quality of the hospital services they receive. Your team has been asked what the department should investigate and change. This is the library case National Health System, adapted from the INSEAD casebook, and the numbers below are its own reference inputs, which its exhibit labels as illustrative.

Clarifying questions. Which services drive the unhappiness, and who is in scope? The case answers that older patients are generally happy with community medical centres and that the problem sits in hospital services, which patients reach by a GP referral and an online booking. Ask for the objective (improve hospital services for older patients, nothing else), a waiting-time or error baseline (none supplied) and a budget (none supplied). Saying "no baseline" back to the interviewer is a good move, because it tells them your first recommendation will be about what to measure.

Structure. Connect older patients' needs to the care process and to feasibility:

• Needs: what time and quality mean to someone over 60 (waiting for a slot, waiting on the day, a correct diagnosis, follow-up).
• Process: booking and transport, waiting and treatment, follow-up and reimbursement, each with hypotheses, not conclusions.
• Feasibility: funding, hospital management, staff and union support.

When the interviewer says diagnoses are slow and error-prone, split causes into people (long hours, staff shortages, training) and system (missing medical histories, equipment, slow tests). In the case, missing histories and long doctor hours from staffing shortages are the confirmed causes.

The math. How many specialists does the department need? The reference inputs: 16 million people, ages spread evenly from 0 to 80, four visits a year for each person over 60, 30 minutes per visit, 8 hours a day, 5 days a week, 50 weeks a year, and 1,000 specialists today.

1. People over 60: ages 60 to 80 are 20 of 80 years, so 20 / 80 = 25%, and 25% x 16M = 4M people.
2. Visits: 4M x 4 = 16M visits a year.
3. Capacity of one specialist: 8 x 5 x 50 = 2,000 hours a year, and at 30 minutes a visit that is 2 visits an hour, so 4,000 visits a year.
4. Specialists needed: 16M / 4,000 = 4,000. Beyond the 1,000 you have, that is 3,000 more.

The common slip is to read 30 minutes as 1 visit per hour and then double it to 8,000 visits per specialist, which understates the need by half. Check the units before you divide.

Recommendation. Make medical histories available across the hospital system first, since that fixes a confirmed cause without hiring, and phase the staffing toward the 3,000 gap rather than committing to it at once. The estimate assumes every specialist hour goes to these visits and every existing specialist is redeployable, so validate real utilization, recruitment and funding before the department commits.

To run this case out loud with an AI interviewer, use the link in the table above. It is the lead case on this page.

Independent

Run the National Health System case

Public sectorMedium

Practice this case free

Case 2: A state Medicaid eligibility backlog

Prompt. A state Medicaid agency has 100,000 applications waiting for an eligibility decision. Federal rules give it 45 days to decide, and the median wait is about 100 days. The state's governor wants the backlog cleared within six months. All numbers are practice assumptions.

Clarifying questions. What counts as a decision, and does the 45 days apply to every application? How many applications arrive each month, how many caseworkers decide them, and how long does one take? Are there rules that need a human on every file? What does an error cost, since a wrong approval is a payment error and a wrong denial is a person without coverage? The interviewer gives you: 40,000 new applications a month, 300 caseworkers who each decide 5 applications a day, 20 working days a month, and a loaded cost of $70,000 per caseworker per year.

Structure.

• Demand: applications arriving, by type, and why this month is higher than usual.
• Capacity: caseworkers, minutes per file, and what slows each file (verification calls, missing documents).
• Backlog: how big it is and how fast it grows.
• Levers: add people, cut minutes per file, remove files from human review.

The math.

1. Capacity today: 300 x 5 x 20 = 30,000 decisions a month, against 40,000 arriving, so the backlog grows by 10,000 a month.
2. Today's wait: 100,000 / 30,000 = 3.3 months, which is about 100 days, so the 45-day rule is broken by more than double.
3. Target backlog: a 45-day wait is 1.5 months, and at 40,000 decisions a month that is a backlog of 60,000. So 40,000 files must be cleared in six months, or 6,667 a month on top of the 40,000 arriving.
4. Throughput needed: 40,000 + 6,667 = 46,667 a month, and each caseworker does 100, so 467 caseworkers, which is 167 more than today (56% more).
5. Alternative lever: suppose automated data matching against state wage and federal records can decide 25% of applications with no caseworker. Caseworker load falls to 75% x 46,667 = 35,000, which needs 350 caseworkers, so 50 more. In steady state, 75% x 40,000 = 30,000 files matches the 300 you have.
6. Year one cost, hiring only: 167 caseworkers for 6 months is 167 x $70,000 x 0.5 = $5.8M, then 100 for the next 6 months is $3.5M, so about $9.3M, and $7.0M a year after that.
7. Year one cost, automation plus temporary staff: a $4M one-time matching tool plus 50 temporary caseworkers for 6 months (50 x $70,000 x 0.5 = $1.75M) is $5.75M, and nothing extra after.

Recommendation. Do the automated matching and bring in temporary caseworkers for six months. It costs about $3.6M less in year one and avoids $7M a year of permanent staff. The risk is accuracy: automated approvals that are wrong become payment errors. Test the matching on a sample of 1,000 past files, set an error ceiling before launch, and keep a human on every file the match cannot confirm.

What a weak close and a strong close sound like on this case. A good close has four parts: the decision, what it means for the client, the main risk and the next step. The weak version restates the numbers and stops. The strong version uses the figures worked above.

The weak close fails because the governor cannot act on it: there is no choice, no number and no condition. Practice saying the strong version in under 30 seconds, then do the same on any other case on this page.

Practice a public sector case for Guidehouse

Public SectorHard

Start the public sector case

Case 3: A health system losing money to claim denials

Prompt. A six-hospital health system submits claims worth $1.2B a year in expected payments. Payers deny 11% of them on first submission, and the CFO wants to know how much that costs and what to fix. All numbers are practice assumptions.

Clarifying questions. Is 11% by value or by count? Which payers and which denial reasons? What share of denied dollars do appeals recover today? What does the health system spend on denial management now? You get: 11% by value, recoveries of 40% of denied dollars through appeals, and this breakdown of denied dollars by reason:

Structure. Follow the claim through the revenue cycle:

• Front end: registration, eligibility checks, prior authorization.
• Middle: clinical documentation and coding.
• Back end: billing, denial management, appeals.
• Payers: which ones deny the most and why.

The math.

1. Denied dollars: 11% x $1.2B = $132M.
2. Recovered today: 40% x $132M = $52.8M, so the permanent loss is 132 - 52.8 = $79.2M a year.
3. Read the exhibit: prior authorization and eligibility together are 32% + 21% = 53%, or $69.96M (about $70M), and both are preventable before the patient is treated.
4. Front-end program: if it prevents half of those, denied dollars fall by about $35M to $97M, which is 8.1% of claims. With the same 40% recovered, the loss is 97 x 0.6 = $58.2M, a gain of $21.0M.
5. Appeals program: on its own, raising recovery from 40% to 55% of the original $132M leaves a loss of 132 x 0.45 = $59.4M, a gain of $19.8M. On top of the front-end program, 55% of the remaining $97M leaves a loss of 97 x 0.45 = $43.65M.
6. Combined gain: 79.2 - 43.65 = $35.6M, which is less than the sum of the two alone ($21.0M plus $19.8M) because the appeals program has fewer denied dollars to work on.
7. Cost: front-end program $6.0M a year, plus an appeals team of 20 people at $90,000 ($1.8M) and $1.2M of software, so $9.0M. Net benefit is about $26.6M a year, and gross benefit is about 4 times the cost.

Recommendation. Fund both, and start with the front end, because it is the larger and more certain gain and it also improves cash. A cash side benefit is worth saying: $1.2B / 365 is about $3.3M of claims a day, so cutting 5 days from accounts receivable releases about $16M once. The risk is that the 40% and 55% recovery rates are the least certain numbers, so measure the recovery by payer for one quarter before sizing the appeals team.

Practice a healthcare case for Guidehouse

HealthcareMedium

Start the healthcare case

Case 4: A utility's storm resilience plan

Prompt. A utility serving 2 million customers wants to ask its regulator to approve a $1.5B, five-year plan to harden its grid against storms. The regulator wants evidence that the benefits are worth the cost to customers. Should the utility file it as written? This is a practice case, and the numbers are not CenterPoint's or any real utility's.

Clarifying questions. What does the plan do (burying lines, replacing poles, automation)? What is the baseline, meaning how many storm outage minutes does a customer have a year, and how much would the plan cut? How does the regulator value an outage? What are the other options? Over what life? You get: the plan cuts 90 storm outage minutes per customer per year; the regulator values an outage at $30 per customer hour (a practice assumption); the plan avoids $20M a year in restoration cost and $10M in equipment damage; operating costs are $15M a year; assets last 20 years; the discount rate is 7%.

Structure.

• Benefits: avoided customer outage cost, avoided restoration cost, avoided damage.
• Costs: capital, operating cost, and the bill impact on customers.
• Alternatives: the whole plan, a targeted plan, doing nothing.
• Regulator tests: benefit-cost ratio, customer bill, and measurable results.

The math.

1. Customer minutes saved: 2M x 90 = 180M minutes, which is 180M / 60 = 3M customer hours a year.
2. Outage benefit: 3M x $30 = $90M a year, and with $20M and $10M it is $120M a year.
3. Discount factor for 20 years at 7%: about 10.6 (the present value of $1 a year).
4. Present value of benefits: 120 x 10.6 = $1,271M.
5. Present value of costs: $1,500M plus 15 x 10.6 = $159M, so $1,659M.
6. Benefit-cost ratio: 1,271 / 1,659 = 0.77. As written, the plan costs more than it returns on these inputs.
7. Targeted option: suppose an exhibit shows the worst 20% of circuits cause 60% of storm outage minutes. Hardening only those costs $500M (plus $5M a year) and captures 60% of the 90 minutes, which is 54 minutes. Assume every benefit scales with the minutes avoided, so the yearly benefit is 120 x 54 / 90 = $72M.
8. Targeted ratio: 72 x 10.6 = $763M against 500 + 5 x 10.6 = $553M, so a ratio of 1.38.
9. The rest of the plan: $48M of benefit a year against $1,000M plus $10M a year of operating cost gives 509 / 1,106, a ratio of 0.46.
10. Bill impact: spreading $500M over 20 years at 7% costs about $47M a year, plus $5M, so $52M, which is $26 per customer per year, against $78 per customer per year for the full plan.

Recommendation. File the targeted $500M first phase, which clears the regulator's test at a ratio of 1.4, costs a customer about $26 a year, and leaves a second phase to be argued with the results in hand. The assumption that deserves the most pressure is the $30 per customer hour, because the targeted plan's ratio falls below 1 if the value is lower than about $19 per customer hour, while the full plan needs about $42 to break even. Show the regulator how the ratio moves with that number.

Practice an energy case for Guidehouse

EnergyMedium

Start the energy case

Case 5: Sizing the fast chargers a state needs

Prompt. A state energy office has federal money for public fast chargers and asks how many charging ports the state needs by 2030. Size it. The numbers below are practice assumptions, and the library's EV Charging Hub Profitability case in the table above is the profit side of the same market.

Clarifying questions. Fast chargers only, or all public charging? Ports or sites? What is the state's population and vehicle count, and how many ports exist today? You get: fast chargers only, ports, 6M people, 4.5M cars, 1,200 fast ports today.

Structure. Demand for public energy divided by what one port delivers:

• EVs in 2030: cars x EV share.
• Public charging energy: EVs that rely on public charging (no home charging) versus EVs that charge in public sometimes (trips).
• Supply per port: power x hours x utilization.

The math.

1. EVs: 4.5M x 20% share in 2030 = 900,000.
2. Segments: 20% have no home charging, so 180,000 depend on public charging and 720,000 do not.
3. Energy per EV: 10,000 miles x 0.3 kWh per mile = 3,000 kWh a year.
4. Public energy: the 180,000 get 70% from public chargers: 180,000 x 3,000 x 0.7 = 378M kWh. The 720,000 get 10% on trips: 720,000 x 3,000 x 0.1 = 216M kWh. Total 594M kWh.
5. One port: an average of 60 kW while in use, 8,760 hours a year, used 15% of the time: 60 x 8,760 x 0.15 = 78,840, about 80,000 kWh a year.
6. Ports needed: 594M / 80,000 = about 7,400, or about 120 EVs per port.
7. Gap and cost: 7,400 - 1,200 = 6,200 ports, and at $150,000 a port that is $930M.
8. Sensitivity: at 10% utilization the need is about 11,300 ports; at 20% it is about 5,650. Utilization moves the answer more than any other input.

Recommendation. Plan on about 7,400 ports and fund the first wave on highway corridors and in areas where residents lack home charging, because those drive the 70% public share. Do not fund all 6,200 on the 15% assumption: release money in tranches and check utilization of the first wave against the 15% before the next, because if real utilization is 20%, the state would build about 1,750 more ports than it needs.

Practice an energy market entry case for Guidehouse

EnergyMedium

Start the energy case

Case 6: An aircraft depot that is falling behind

Prompt. A defense agency runs a maintenance depot with 12 repair bays. Each aircraft takes 120 days. The fleet is 300 aircraft and each must go through the depot every 6 years. Readiness is slipping. What should the agency do? All numbers are practice assumptions.

Clarifying questions. What does the depot need to do each year, and what does it do? What takes the 120 days? What would more bays cost, and what does a faster process cost? You get the exhibit below and the costs in the math.

Structure. Required throughput versus actual, then three levers:

• Demand: aircraft that must pass through each year.
• Capacity: bays and days per aircraft.
• Levers: more bays, faster steps, less waiting.

The math.

1. Required: 300 / 6 = 50 aircraft a year.
2. Actual: 12 bays x 365 / 120 days = 36.5 a year, so the queue grows by about 13 aircraft each year.
3. Days per aircraft needed with 12 bays: 12 x 365 / 50 = 87.6 days, so the depot must cut about 32 days.
4. Read the exhibit: waiting for parts is 38 of 120 days, the largest delay that is not the repair itself. Cut parts waiting from 38 to 10 days (28 days saved), inspection from 20 to 17 (3 days) and queue from 6 to 2 (4 days). That is 35 days, so 85 days, and 12 x 365 / 85 = 51.5 aircraft a year, which meets the 50.
5. More bays instead: 50 x 120 / 365 = 16.4, so 17 bays, 5 more. At $20M a bay that is $100M, plus 5 bays x 40 people x $100,000 = $20M a year in staff.
6. Parts route: a $60M buffer stock of critical parts, plus about 10% a year to carry it, so $6M a year.

Recommendation. Fix parts waiting first. It costs $60M once and about $6M a year, against $100M and $20M a year for five more bays, and it meets the 50 aircraft target without adding sites. The risk is that parts waiting comes from suppliers, not from the depot's stock policy, so confirm with the parts list which parts cause the waiting and whether a buffer actually shortens it before buying $60M of inventory.

Practice a defense case for Guidehouse

Aerospace & DefenseMedium

Start the defense case

Case 7: A bank's anti-money-laundering alert backlog

Prompt. A regional bank's transaction monitoring system raises 15,000 alerts a month. Forty analysts review them, and 24,000 alerts are waiting. A regulator has criticized the backlog. How should the bank clear it? All numbers are practice assumptions.

Clarifying questions. How many alerts does an analyst clear? What share lead to a suspicious activity report? Can the rules that create alerts be changed, and who must approve a change? What is the deadline? You get: 14 alerts per analyst per day, 20 days a month, about 95% of alerts closed as false positives, a six-month deadline, and a loaded cost of $85,000 per analyst. Rule changes need testing that a regulator will accept.

Structure.

• Inflow: alerts created, and which rules create most of them.
• Capacity: analysts, alerts per analyst, and time per alert.
• Backlog: size and age.
• Levers: add analysts, make fewer low-value alerts, speed each review.

The math.

1. Capacity: 40 x 14 x 20 = 11,200 alerts a month, against 15,000, so the shortfall is 3,800 a month and the 24,000 backlog is 2.1 months of work.
2. Clear the backlog in six months: 24,000 / 6 = 4,000 extra a month.
3. Hire only: (15,000 + 4,000) / 280 per analyst a month = 68 analysts. Steady state: 15,000 / 280 = 54, so 14 permanent hires.
4. Tune the rules: if low-risk customer segments are changed to produce 30% fewer alerts, inflow is 10,500, and steady state needs 10,500 / 280 = 37.5 analysts, which the 40 already cover.
5. Backlog burn with tuning: (10,500 + 4,000) / 280 = 52 analysts, so 12 temporary analysts for six months.
6. Year one cost, hire only: 14 permanent analysts at $85,000 is $1.19M, and 14 more for six months is $0.6M, so about $1.8M, then $1.2M a year.
7. Year one cost, tune plus temporary: 12 x $85,000 x 0.5 = $0.51M plus $0.6M for the tuning study, so about $1.1M, then nothing.
8. Testing the tuning: review 500 alerts the new rules would have dropped. If none is suspicious, the rule of three puts the 95% upper bound on the miss rate at about 3 / 500 = 0.6%.

Recommendation. Do both: bring in 12 temporary analysts now, because the backlog is the finding the regulator cited, and tune the rules after a documented sample test. Do not suppress any alert type before the test, since a missed suspicious activity report is worse than a late one. The bank should tell the regulator the plan, the sample result and the dates, which is also the strongest evidence it can give.

Practice a financial services operations case for Guidehouse

Financial ServicesMedium

Start the banking case

Case 8: An Excel exercise on allocating a repair grant

Prompt. A county has $12M of grant money to repair low-income homes. It sends you a spreadsheet of five neighborhoods and asks how to allocate it, with a one-page recommendation. This is the shape of the Excel and take-home exercises that CaseBasix reports for some roles, with practice numbers.

Structure. Add columns before you reach for a pivot table:

• Cost to repair everything: homes x cost per home.
• Need per dollar: poverty rate (as points) divided by cost per home in thousands.
• Rank by need per dollar, then fund down the list until the $12M is spent.
• Check results against two goals: the most homes, and the most repairs in poor households.

The math.

1. Cost to repair all homes: A 900 x $8,000 = $7.2M; B $7.2M; C $7.2M; D $6.0M; E $6.3M. Total $33.9M, so the $12M covers 35%.
2. Need points per $1,000: A 30 / 8 = 3.75; B 42 / 12 = 3.50; C 18 / 6 = 3.00; E 25 / 9 = 2.78; D 38 / 15 = 2.53. The ranking is A, B, C, E, D.
3. Option 1, most homes: fund the cheapest repairs first. C is $7.2M for 1,200 homes, and the remaining $4.8M fills 600 homes in A. That is 1,800 homes.
4. Option 2, need first: A in full is $7.2M for 900 homes, and the remaining $4.8M repairs 400 homes in B ($12,000 each). That is 1,300 homes.
5. Weight each option by the poverty rate of the neighborhood it funds. Option 1: 1,200 x 0.18 + 600 x 0.30 = 216 + 180 = 396. Option 2: 900 x 0.30 + 400 x 0.42 = 270 + 168 = 438.
6. Option 3, the middle path: A in full and C with the rest ($4.8M / $6,000 = 800 homes). That is 1,700 homes and 270 + 800 x 0.18 = 270 + 144 = 414.

In Excel, with the table in columns A to D, put =B2*C2 in E2 for the cost to repair everything, =(D2*100)/(C2/1000) in F2 for need points and =RANK(F2,F$2:F$6) in G2 for the ranking. The funding step is then a running total of cost down the ranked list that stops at the budget. Say how you built it in the memo, because the interviewer reads the method as much as the answer.

Recommendation. Fund A in full and give the rest to C. Option 3 gives 94% of the most homes and 95% of the best poverty-weighted result, and A ranks first on need per dollar and second on eligible homes. Poverty rate is a stand-in for need, and repair costs on old homes tend to overrun, so a 20% overrun would cut every option's homes by one sixth (1 / 1.2 is 83%). Ask the county whether the rule is more homes or more need, because that single choice picks between Options 1 and 2.

Practice a public sector case for Guidehouse

EducationHard

Start the school system case

How should you format a Guidehouse take-home case?

Lead with one recommendation on page one, give three supported reasons and state the risks. CaseBasix reports take-home cases due in 24 to 48 hours that mirror a client deliverable, as a slide deck or an Excel output, and the format advice from the Guidehouse case interview guide and the written case interview guide applies. Case 8 above is the model: a table of assumptions, a ranked list, three options with numbers and a recommendation that names the choice the client must make.

Use the exhibit-reading rep below to practice the step that most written cases start from, which is turning a chart into the one number that changes the decision.

How do you turn these eight cases into a prep plan?

Run each case once cold, then repeat only the step that broke. Cases 1, 5 and 6 test estimating and throughput, Cases 2 and 7 test capacity and backlog, Cases 3, 4 and 8 test benefit and cost with small data. The sequence below uses all eight in about a week.

For process, rounds, clearance and the behavioral questions, use the Guidehouse case interview guide. For peers that also hire into government and cleared work, read the Booz Allen Hamilton case interview guide and the West Monroe case interview guide. The top healthcare consulting firms page compares Guidehouse's health work with other firms, and the case interview examples hub collects worked cases from other firms. If you want more practice on the math itself, see case interview math practice and the cost reduction case interview.

Prep plans for McKinsey, BCG, Bain, Deloitte, Accenture, EY-Parthenon, Kearney, Oliver Wyman, Roland Berger, L.E.K., OC&C, Strategy&, Simon-Kucher, Arthur D. Little, Capital One and your target firm. Choose your firm in the plan builder.

Turn this guide into a weekly plan

Choose your firm and interview date and get the weeks laid out, from the first lesson to the last case.

Build my prep plan

1

Learn the format

17 steps

• Learn how to caseGuided learning · 35 min
• The Case Interview MethodLesson · 8 min
• Profitability & Turnaround: AfrihealSkill drill · 5 min
• Profitability & Turnaround: Berty BankSkill drill · 5 min

Sources and further reading (checked November 5, 2026)

• Guidehouse: Early career and internships (10-week internship, consulting bootcamp, team placement matched to interests, application link)
• Guidehouse: Interning at Guidehouse (five stages from talent community to offer, with an initial interview and a final round; no case mentioned)
• Guidehouse: Gateway to Guidehouse Life Sciences (one behavioral and one case interview with Life Sciences team members; page written for the 2025 program)
• Guidehouse: About Guidehouse (five industries, Advisory, Technology and Managed Services, about 18,000 colleagues)
• Guidehouse case study: CenterPoint Energy resilience plan ($2.7 billion plan, approved August 2025, about $6.50 of projected benefit per dollar)
• Guidehouse case study: $10 billion in federal COVID-19 aid (more than 350 providers and more than 20 state and local entities)
• Guidehouse case study: federal department resource platforms (nine project manager interviews)
• Navigant Consulting (Guidehouse) interview guide, CaseBasix (candidate-reported rounds, 30 to 45 minute case, 24 to 48 hour take-home, Excel exercise, industries and case math; a prep site, not Guidehouse)
• Guidehouse Interview: Process, Case Interview, Questions, Soreno (practice-area tailoring advice and the decision, so-what, risk, next step shape of a close; a prep site, not Guidehouse)
• Guidehouse completes transaction with Bain Capital (Bain Capital ownership, deal closed December 14, 2023)