Boston Consulting Group · Online assessment

BCG Casey Practice Test 2026

Updated 2026-08-06 · All practice scenarios on this page are original CasingLab creations.

BCG Casey is BCG's chatbot based online case assessment. Instead of talking to an interviewer, you work through one business case inside a chat interface: Casey sends you the client situation, asks questions one at a time, shows exhibits, and expects typed answers under a per question timer. Most versions end with a short recorded video answer where you deliver your recommendation to the camera.

Casey compresses a full case interview into roughly 25 to 35 minutes with no human in the loop, which changes how you must prepare: there is no interviewer to steer you back, no partial credit for thinking out loud, and no second chance once a question locks.

This page covers the 2026 format, timing and question types, what BCG scores, and a free three question mini simulation you can play immediately. Every scenario here is an original CasingLab creation that mirrors only the format of Casey, never real test content.

Format and timing

BCG does not publish an official specification for Casey, and cohorts report variations in length and timing. The figures below reflect consistent candidate reporting for recent cycles, not official numbers.

SectionDurationQuestionsFormat
Chat case: structuring and analysisAbout 25 to 35 minRoughly 8 to 10 questionsOne continuous business case in a chat window: framework questions, exhibit reading, calculations and judgment calls, each with its own timer and typed or multiple choice answers.
Video synthesisAbout 1 min recording1 promptA one way recorded answer: you summarize your recommendation to the camera as if briefing the client, usually with limited preparation time.

Practice the real BCG Casey format

Full simulations, real timing, detailed scoring.

Practice the full simulationTry the free diagnostic

Question types on the BCG Casey

Structuring questions

Casey opens the case by asking how you would approach the client's problem. You either type a short structure or pick the best framing from options. The scoring logic rewards structures that are tailored to the specific client and collectively cover the problem without overlap. Generic memorized frameworks score poorly because they ignore the details Casey just gave you.

Scored skills: Tailored problem structuring · MECE thinking · Prioritizing what to investigate first

Example question

Invented example in the Casey style. Chat message: our client is Valora, a 60 store regional bakery franchise. Profit fell 18 percent this year while revenue stayed flat.

Casey asks: which structure best isolates the cause of the profit decline?

  • Analyze the bakery market, then competitors, then customers, then the company
  • Break profit into revenue and costs; since revenue is flat, drill into cost lines per store and across the network
  • Survey customer satisfaction across all 60 stores to find service gaps
  • Benchmark Valora's marketing spend against national bakery chains

Answer: Break profit into revenue and costs; since revenue is flat, drill into cost lines per store and across the network

The prompt already tells you revenue is flat, so a profit decline must sit in costs. The best structure uses that given fact to cut the tree immediately instead of restarting from a generic market overview. Casey rewards structures that react to the data in the chat, exactly like a good candidate reacts to an interviewer.

Exhibit interpretation

Casey drops a chart or table into the chat and asks what it means for the client. The right answer is the insight that changes the client's decision, not a description of the axes. Expect distractor options that are true but irrelevant, and keep an eye on units and base sizes.

Scored skills: Chart and table reading · Separating insight from description · Spotting the decision relevant number

Example question

Invented example. Casey shows a table for Helioscope, a maker of garden sensors: unit sales by channel. Retail: 40,000 units, margin 12 percent. Online direct: 25,000 units, margin 31 percent. Wholesale: 55,000 units, margin 6 percent.

Casey asks: which observation matters most for improving total profit?

  • Wholesale sells the most units of the three channels
  • The company sells through three different channels
  • Online direct earns roughly five times wholesale's margin, so shifting volume toward it has the largest profit upside
  • Retail sits between the other two channels on both metrics

Answer: Online direct earns roughly five times wholesale's margin, so shifting volume toward it has the largest profit upside

Three of the four options merely describe the table. Only one connects the numbers to an action that moves profit. On Casey exhibits, ask yourself what you would tell the client to do differently; the answer that survives that question is the scored insight.

Quantitative questions

Mid case, Casey asks for a number: a breakeven, a market size, a margin impact. You type a numeric answer against the clock. The arithmetic is case interview math, doable by hand, but the chat format is unforgiving: there is no interviewer to confirm your approach before you commit.

Scored skills: Breakeven and margin math · Structured mental arithmetic · Committing to an answer under time pressure

Example question

Invented example. Novabrew, a canned cold brew brand, sells at 3.20 per can with a variable cost of 1.90 per can. Monthly fixed costs are 9,100.

Casey asks: how many cans must Novabrew sell each month to break even?

Answer: 7,000 cans

Contribution per can: 3.20 minus 1.90 equals 1.30. Breakeven: 9,100 divided by 1.30 equals 7,000 cans. Write the contribution margin first, every time; candidates who jump straight to division under the timer are the ones who flip the numerator and denominator.

Judgment and synthesis

Casey closes by testing whether you can turn findings into a recommendation: pick the best next move, or record a short video summarizing your advice. Strong answers lead with the recommendation, support it with the two or three numbers that matter, and name the main risk. Rambling chronological recaps score poorly.

Scored skills: Answer first communication · Weighing evidence into one recommendation · Concise spoken delivery on video

Example question

Invented example. Findings so far on Ferrovia, a regional rail caterer: premium menu revenue is growing 9 percent a year, the standard menu is declining 4 percent, kitchen capacity is fully used, and switching one production line to premium takes four months.

Casey asks: what should Ferrovia do next?

  • Keep the current production mix to avoid the four month switching cost
  • Begin converting one line to premium now, since growth is compounding and capacity is the binding constraint
  • Cut standard menu prices to revive its declining sales
  • Commission a customer survey before changing anything

Answer: Begin converting one line to premium now, since growth is compounding and capacity is the binding constraint

The data points one way: demand is shifting to premium, capacity is the bottleneck, and the conversion lag argues for starting early, not waiting. The survey option is the classic stall answer; Casey consistently rewards candidates who commit when the evidence is sufficient.

Try three practice questions, free

Three original questions in the BCG Casey format. No account, no email, instant corrections with explanations.

Question 1 of 3

Structuring, original scenario. Chat message from Casey: our client is Valora, a 60 store regional bakery franchise. Profit fell 18 percent this year while revenue stayed flat. Which structure best isolates the cause?

Practice the real BCG Casey format

Full simulations, real timing, detailed scoring.

Practice the full simulationTry the free diagnostic

What is actually scored

BCG does not publish Casey's scoring rubric. What follows combines BCG's public descriptions of what its assessments look for with consistent candidate reporting, and it maps closely onto how live BCG case interviews are graded.

Structure quality

Whether your breakdowns are tailored to the client in the chat, cover the problem without overlap, and prioritize the branch the data points to. Recycled generic frameworks are easy for automated scoring to spot and mark down.

Quantitative accuracy

Numeric answers are right or wrong, and reported scoring leaves little room for method marks. Clean case math habits, contribution margins, chained percentages and sanity checks, convert directly into points.

Business judgment

Choosing the insight that changes the client's decision, and committing to a recommendation when the evidence is sufficient rather than stalling for more data.

Communication on video

The closing recording is reviewed for answer first delivery: recommendation up front, the two or three numbers that support it, one risk, done. Polish matters less than structure and confidence.

How the chatbot format changes the game

In a live case interview, a good interviewer nudges you back when you drift and gives you credit for recovering. Casey does neither. Each question is scored on your first committed answer, the timer is per question, and most cohorts report you cannot return to a previous question once you move on. That single difference should reshape your preparation: practice producing complete first pass answers, not thinking out loud toward one.

The typed medium also rewards concision. A three line tailored structure beats a ten line brain dump, both for the clock and for the scoring. And because exhibits arrive inside a chat window, train yourself to read a table once, extract the decision relevant number, and answer, rather than re reading under time pressure.

Casey versus a live case interview

Treat Casey as a case interview with the safety nets removed. The same skills are tested, structuring, case math, judgment and synthesis, but there is no partial credit for a promising direction, no clarifying dialogue, and no reading of your interviewer's face. Candidates who over rely on interactivity in their live practice often underperform here.

The efficient preparation split: keep doing live cases for depth, and add chatbot format simulations for the specific muscles Casey isolates, committing under a per question timer, typing tight structures, and recording a one take video synthesis.

How to prepare for Casey in 2026

Three steps. First, learn the format so nothing on test day is new: chat pacing, per question timers, the closing video. Second, drill case math to first pass accuracy, breakevens, margins, growth chains, because Casey gives no method marks. Third, run at least two full simulations under real timing and record your video answer each time; hearing your own synthesis is the fastest way to fix it.

If you have one week: two format familiarization sessions, three timed math drills, two full dress rehearsals. Prioritize the video practice if you have never recorded yourself; almost everyone is worse on camera than they expect the first time.

Frequently asked questions

Most candidates report 25 to 35 minutes of chat, with each question individually timed, plus a short recorded video of about a minute at the close. BCG does not publish an official duration, so confirm the details in your own invitation email.

Keep reading

Practice the real BCG Casey format

Full simulations, real timing, detailed scoring.

Practice the full BCG Casey simulation