The short answer: Business analyst is one title for three different jobs at these companies. At Amazon it is an analytics and reporting role built on SQL, Excel and Tableau. At Capital One it is a strategy and analytics role hired through case interviews. At McKinsey it is the entry-level generalist consultant on a client team. Capital One and McKinsey publish how their BA interviews work: Capital One a mini case and a Power Day of cases, McKinsey a Personal Experience Interview paired with a case in every interview. Amazon publishes only its general process, so its SQL screen and round mix come from a few candidate reports. Read the posting first and prepare for the job it describes. Last reviewed October 2026.
Business Analyst interview: quick answers
| Question | Short answer |
|---|---|
| What does a business analyst do? | It depends on the company. Amazon BAs build metrics, reports and dashboards for a business team. Capital One BAs turn analysis into product, pricing and credit strategy. McKinsey BAs own a workstream on a client consulting team. |
| What do the interviews test? | Amazon: Leadership Principles in every interview, plus SQL and a metric case in candidate reports. Capital One: business cases with real math, a product strategy conversation and job fit. McKinsey: a PEI and a case in every interview, after the Solve assessment. |
| Can I use AI tools? | Capital One prohibits AI tools in screenings, interviews and assessments unless permission is given. McKinsey does not allow real-time AI answers in interviews or AI in assessments. Amazon publishes no candidate AI rule; ask your recruiter. |
| How much do business analysts make? | Posted US base ranges run from $66.9K (Amazon, Nashville) to $160K (Amazon Senior BA and BA III). Capital One Sr. BA lists $101.1K to $138.4K and McKinsey BA $112K to $132K. |
| Do I need case interview prep? | For Capital One and McKinsey, yes: both run case interviews. For Amazon, candidate reports describe an open-ended metric case, but SQL and Leadership Principles stories matter more. |
| How do students get in? | All three hire students: Amazon through a BA internship, Capital One through its Analyst Development Program and internships, and McKinsey through BA internships and full-time campus recruiting. |
What business analysts do
At Amazon, a business analyst owns metrics and reporting for a business team. Postings ask for SQL, 5+ years of Excel with VBA, pivot tables and power pivots, and Tableau, and describe recurring reports, ad hoc analysis, dashboards and recommendations to stakeholders. The work is spread across Fulfillment by Amazon, Amazon Logistics, Selling Partner Services, Whole Foods Market, AWS and others. Amazon keeps BA separate from business intelligence engineer, which adds data modeling, ETL pipelines and Python, and from data scientist.
At Capital One, a business analyst works where data and strategy meet. Sr. BA postings cover product and pricing strategy for lending products, marketing, credit risk models and loss forecasting, and name SQL as the one tool. The job is to frame a business problem, test hypotheses and recommend a strategy with IT, Sales, Operations and Finance. It is not an IT requirements role, and Data Analyst and Data Scientist are separate categories.
At McKinsey, the business analyst is the entry-level generalist consultant, typically within four years of an undergraduate degree. BAs join client teams of typically 3 to 5 consultants, own a workstream and combine research, data analysis, interviews and brainstorming into recommendations presented in PowerPoint decks and Excel models. Travel is part of the job, and data science and research analysts are separate job families.
How 3 companies run the business analyst interview
The short version of each process. Each company guide has the full detail, with links to the company's own pages, and says which parts come from candidate reports.
| Company | Screens | Main interviews | AI tools | What stands out |
|---|---|---|---|---|
| Amazon | Assessment for some roles; 1 or 2 phone screens, SQL (candidate reports) | Loop of 2 to 7 interviewers; SQL and case (candidate reports) | No published rule; ask the recruiter | Excel, SQL and Tableau reporting; Bar Raiser |
| Capital One | Online assessment, recruiter screen, 30-minute mini case | Power Day: 2 to 3 cases, product strategy, job fit | Prohibited unless permission is given | Strategy and analytics in cards and banking |
| McKinsey & Company | Solve game-based assessment | PEI plus case in every interview; 4 to 5 interviews (candidate reports) | No real-time answers; none in assessments | Generalist consultant; pen and paper, no calculator |
The interview types in depth
Online assessments
All three companies can start with an online step. Amazon's general assessments page names work style assessments and work sample simulations for some roles, and candidate reports describe SQL and Excel content for BAs. Capital One sends an online assessment of about 20 to 45 minutes with modules on relationships, a business case, your story and your approach. McKinsey invites most consulting applicants to Solve, a game-based assessment it says needs no preparation.
Take each one in a quiet space in one sitting, and answer work-style questions as you really work. The rules on outside help are strict at all three, so do the assessment alone.
SQL and Excel
This round matters most at Amazon. Candidate reports describe SQL in the phone screen and sometimes in the loop: joins, CTEs, window functions, GROUP BY and subqueries, written in an editor that may not run the query. Postings also ask for deep Excel, so expect questions on pivot tables, lookups and macros. Capital One postings name SQL, but its published interview is case-based. McKinsey does not test SQL.
A structure that works: restate the question, name the grain of the answer, write the query in CTE steps, then check it for nulls, duplicates and date edges out loud.
Business cases
Capital One and McKinsey both build their interviews around cases, in different styles. Capital One says its cases are realistic business problems with a good deal of math, such as break-even volume, and that you should end with a recommendation you can defend with data. McKinsey says to listen closely and give answers tailored to the question instead of standard frameworks, and candidate reports describe its case as interviewer-led. Amazon's case, in candidate reports, is an open-ended metric problem such as a rising refund rate.
Clarify the goal and the metric, lay out a short structure, work the numbers on paper and keep a running list of results. Finish with a recommendation, the main risk and what you would check next.
Case math must be fast and right. McKinsey bans calculators; Capital One publishes no rule, and prep guides disagree, so ask.
Product strategy
Capital One's Power Day includes a product strategy conversation. Expect a card, banking or payments feature: who it is for, how big it is, what it costs and how you would measure it.
Start from the customer and the business goal, size the idea with stated assumptions, pick one primary metric and a guardrail, and propose a test before a full launch.
Behavioral, job fit and the PEI
Every company tests your past experience, each in its own format. Amazon asks Leadership Principles questions in every interview and wants STAR answers with data, said as "I". Capital One runs one or two job fit interviews with "Tell me about a time when..." questions on problem solving, influence and results focus. McKinsey pairs a Personal Experience Interview with every case and asks for two examples for each of four areas.
Build a set of eight to ten stories with numbers in them, and practice telling each one at two lengths: a two-minute version and a deep version that survives ten minutes of follow-up.
Pay in US job postings
US salary ranges from each company's own postings for this role, seen October 29, 2026. They are base salary unless the company guide says the posting lists total salary. Each company guide has the postings and links.
| Company | Titles in the postings | Posted US pay range |
|---|---|---|
| Amazon | Business Analyst, Business Analyst II, Business Analyst III and Senior Business Analyst | $66.9K to $160K |
| Capital One | Business Analyst (campus Business Analyst Associate) and Sr. Business Analyst | $101.1K to $138.4K |
| McKinsey & Company | Business Analyst (Fellow in some offices), plus practice postings such as Business Analyst - Operations and Business Analyst - Private Capital | $112K to $132K |
Entry-level paths
- Amazon: The Business Analyst Intern, Summer 2027 role is 12 weeks in Bellevue for bachelor's students graduating December 2027 to June 2028, and requires SQL.
- Capital One: The Analyst Development Program is a full-time rotation through two teams in business or data analytics, and the Business Analytics internship and Analyst Early Internship lead into it.
- McKinsey & Company: McKinsey hires Business Analyst Interns for 8 to 10 weeks on client teams and runs a Sophomore Summer Business Analyst program in the US and Canada.
AI tools in business analyst interviews
Rules on AI differ by company, so ask your recruiter before every round.
Amazon. Amazon does not publish a candidate AI rule on its careers pages; ask your recruiter. Its assessments page says "All assessments must be completed solely by the applicant." A March 2025 news report described internal guidance asking candidates not to use generative AI tools during interviews unless explicitly permitted; Amazon has not published that as a rule.
Capital One. Capital One's candidate terms say "The use of external resources and third-party assistance, programs, or AI tools (e.g. Gemini, Chat GPT, etc.) is strictly prohibited during screenings, interviews and assessments, unless permission is specifically indicated." Transcription services are not permitted either.
McKinsey. McKinsey marks generating real-time answers during an interview, and using AI during assessments unless specifically permitted, as not allowed. It encourages AI for polishing your resume, practicing and learning frameworks. Its integrity rules also ban calculators, apps, websites and prepared notes, and ask you to disable installed AI note-taking tools.
A 6-week prep plan
| When | Focus |
|---|---|
| Week 1 | Read your target posting and company guide. Decide which job it is (reporting, strategy or consulting), note the AI rule and ask your recruiter which rounds to expect. |
| Week 2 | Behavioral stories: eight to ten with numbers, mapped to Amazon's Leadership Principles, Capital One's competencies or McKinsey's PEI areas. |
| Week 3 | For Amazon, SQL by hand and Excel. For Capital One and McKinsey, case structure on realistic problems. |
| Week 4 | Case math on paper: percentages, break-even, profit per unit and payback, timed and checked. |
| Week 5 | Metrics and product: why a metric moved, whether a change is real, and how to size and test a new feature. |
| Week 6 | Two full mock interviews in your target format, then fix the weakest round. |
Courses for each round
- SQL
- Case in Point
- Data Visualization
- Python Pandas for Data Analysis
- Product Opportunity Sizing
- Applied Statistics
Practice questions
Questions from our bank, then practice prompts from the company guides.
Questions from our bank
From the DataInterview question bank, picked for business analyst interviews. Open one to answer it.
Order Value Significance
Interpreting P Value
Sales Forecast Model
First Year Financial Impact
Debit Card Market Size
Capital One Shopping Value
College Value Assessment
Strategic Stakeholder Management
Median vs Mean
SQL and Excel practice
One prompt per company, from the company guides. These are our prompts, not reported interview questions.
Table shipments(shipment_id, fc_id, ship_date, promised_date, delivered_date). Write a query for the on-time delivery rate by fulfillment center for shipments sent in the week starting October 19, 2026.
Sample Answer
SELECT fc_id, COUNT(*) AS shipments, AVG(CASE WHEN delivered_date <= promised_date THEN 1.0 ELSE 0.0 END) AS on_time_rate FROM shipments WHERE ship_date >= DATE '2026-10-19' AND ship_date < DATE '2026-10-26' GROUP BY fc_id ORDER BY on_time_rate. Say how you treat shipments with no delivered_date: here they count as late, because the CASE returns 0 when the date is NULL. If the stakeholder wants them left out until they arrive, filter them and report how many were excluded. Use a half-open date range so the last day is not dropped.
Business case practice
One prompt per company, from the company guides. These are our prompts, not reported interview questions.
The refund rate for a category on Amazon rose from 2% to 3% week over week. How do you find out why?
Sample Answer
Check the data first: did the refund definition, the data pipeline or the category mapping change? Then size it: on 100,000 orders a week, one point is 1,000 more refunds. Split by seller, item, fulfillment method and refund reason, and separate mix (more orders from high-refund items) from rate (the same items refunding more). Line up the start date with launches, policy changes or a new seller. End with a recommendation and the metric you would watch to confirm the fix.
Capital One is considering a new card perk that costs $4 million a year to run. Each new account it brings in earns $120 a year in profit before the perk. How many new accounts does the perk need to break even?
A private equity client is thinking about buying a regional chain of 200 gyms. How would you structure the diligence?
Metrics and statistics practice
One prompt per company, from the company guides. These are our prompts, not reported interview questions.
A dashboard shows checkout conversion up from 10.0% to 10.4% week over week, with 50,000 sessions each week. Is the change real?
Sample Answer
Use a two-proportion z-test. The pooled rate is 10.2%, so the standard error of the difference is sqrt(0.102 x 0.898 x (2 / 50,000)) = about 0.0019. The change is 0.004, so z = 0.004 / 0.0019 = about 2.1, which is significant at the 5% level, but only just. Before calling it, check for a holiday or traffic mix change between the two weeks, and see whether the lift holds next week.
Case math practice
One prompt per company, from the company guides. These are our prompts, not reported interview questions.
A card portfolio has 2 million accounts with average spend of $12,000 a year. Interchange is 1.8%. What is annual interchange revenue?
Sample Answer
Total spend is 2,000,000 x $12,000 = $24 billion. Interchange is $24 billion x 1.8% = $432 million a year. A quick check: 1% of $24 billion is $240 million, and 0.8% is $192 million, which sum to $432 million.
A product sells for $40, costs $25 a unit to make, and the business has $3 million a year in fixed costs. What is break-even volume?
Product strategy practice
One prompt per company, from the company guides. These are our prompts, not reported interview questions.
Business Cards & Payments is thinking about virtual card numbers for small business customers. How would you size it and decide?
Sample Answer
Start with the customer problem: controlling employee spend and online fraud. Size it with stated assumptions, for example 1 million business card accounts, 10% adoption and $2,000 of extra spend per adopter at 2% interchange: 100,000 x $2,000 x 2% = $4 million a year. Add fraud savings and weigh build and support costs. Recommend a pilot with adoption, spend and fraud rate as the success metrics.
Behavioral practice
Prompts we wrote for business analyst behavioral rounds, with what a strong answer covers.
Tell me about a time your analysis changed a decision.
Sample Answer
Name the decision and who made it, the question you answered, the data and method, what you recommended and what changed, in numbers.
Tell me about a time you found an error in numbers other people relied on.
Tell me about a time you disagreed with a stakeholder or someone more senior.
Tell me about a time you had to recommend something with incomplete data.
Are you ready for business analyst interviews?
1 / 5Can you write SQL or build an Excel model that answers a business question in about 20 minutes?
Company guides
Read the guide for each company you are interviewing with:
- Amazon: Amazon publishes no business analyst loop. Its general process, the SQL and business case rounds candidates report, how BA differs from BIE, the AI question and posted pay.
- Capital One: Capital One publishes its business analyst loop: a mini case, then a Power Day of cases and a product strategy conversation. What each round tests, the AI rule and posted pay.
- McKinsey & Company: McKinsey hires business analysts as entry-level generalist consultants. How Solve, the PEI and the case work, the integrity and AI rules, and posted pay.
Frequently Asked Questions
Is a business analyst the same job at every company?
No. At Amazon it is analytics and reporting with SQL, Excel and Tableau. At Capital One it is strategy and analytics for card, banking and payments products. At McKinsey it is the entry-level generalist consultant. Read the posting before you prepare.
Do business analyst interviews include SQL?
At Amazon, candidate reports describe SQL in the phone screen and sometimes in the loop. Capital One postings name SQL, but its published interview is cases, a product conversation and job fit. McKinsey does not test SQL.
Do business analyst interviews include case interviews?
Capital One and McKinsey both publish case interviews: a mini case and a Power Day of cases at Capital One, and a case in every interview at McKinsey. Amazon candidate reports describe an open-ended metric case in the loop.
Can I use AI tools in business analyst interviews?
Capital One prohibits AI tools during screenings, interviews and assessments unless permission is specifically indicated. McKinsey does not allow AI-generated answers in interviews or AI in assessments. Amazon publishes no candidate AI rule, so ask your recruiter.
How much do business analysts make?
In US postings seen in October 2026, Amazon lists $66.9K to $160K base across BA titles and cities, Capital One Sr. Business Analyst $101.1K to $138.4K, and McKinsey Business Analyst $112K to $132K.
Can I use a calculator in case interviews?
Not at McKinsey: its integrity rules ban calculators. Capital One publishes no rule, and prep guides disagree, so ask your recruiter and practice without one either way.
How is a business analyst different from a data analyst or BI engineer?
At Amazon and Capital One these are separate job families. Amazon BI engineers add data modeling, ETL pipelines and Python. Capital One data analysts need Python, R, Spark and Snowflake and own data quality. Business analysts at both lean toward reporting, strategy and recommendations.
