INTERVIEW GUIDE

American Express Data Analyst Interview: Questions & Process

American Express is a deeply data-driven company (credit risk, fraud, marketing analytics), and its Data Analyst interview reflects that: expect SQL, a business or analytics case framed around a real Amex-style metric, some statistics and A/B-testing reasoning, and a values-based behavioral round. The differentiator is connecting data to a business decision, not just writing a query. Expect a recruiter screen, a technical screen, and 2-3 interview rounds.

The interview process

1. Recruiter screen — ~30 min call

Tests: background, SQL and analytics experience, the team (risk, marketing, fraud), and logistics

2. Technical screen — 45-60 min

Tests: live SQL (joins, aggregation, window functions) and a few statistics or metric-reasoning questions

3. Analytics / business case — 45-60 min

Tests: an Amex-style problem — e.g., why a metric moved, how to size an opportunity, or how to evaluate a campaign — structured end to end

4. Technical / case deep dive — 45-60 min

Tests: deeper SQL, experiment design or measurement, and sometimes Python/SAS/R or a take-home depending on the team

5. Behavioral / values — 45 min

Tests: collaboration, integrity, customer focus, and communicating insights to non-technical partners

Questions you're likely to get

Technical

Write SQL to find each cardmember's total spend in the last 90 days, including those with none.
Use a window function to find each customer's most recent transaction.
Find the top 10% of customers by spend within each segment.
What's the difference between an INNER JOIN and a LEFT JOIN, and when does the choice change your numbers?
How would you design and measure an A/B test for a new card-rewards offer?
Explain p-value and confidence interval in plain language for a business partner.

Role-specific

Card spend in a segment dropped 8% month-over-month. How would you investigate the cause?
How would you size the revenue opportunity of launching a new merchant-category bonus?
What metrics would you track to measure the health of a credit-card portfolio?
How would you decide whether a marketing campaign actually drove incremental spend?

Behavioral

Tell me about a time your analysis changed a business decision. How did you make it land?
Describe a time you had to explain a complex finding to a non-technical stakeholder.
Tell me about a time you caught an error or data-quality issue before it reached a decision.
Why American Express, and why analytics in financial services?

How to answer (worked examples)

Card spend in a segment dropped 8% month-over-month — how do you investigate?

Structure before you speculate. First check whether it's real (data/logging issue, seasonality, fewer days in the month) or a genuine drop. Then decompose: is it fewer active cardmembers, lower spend per member, or a mix shift across merchant categories or geographies? Quantify each segment's contribution to isolate the driver, form a hypothesis, and confirm it in the data. Close with a recommendation and how you'd monitor it. Amex rewards the analyst who decomposes the number methodically over one who guesses.

How would you decide whether a campaign drove incremental spend?

Lead with the word incremental — versus what would have happened anyway. Describe a test-vs-control design (randomized holdout if possible, or a matched control), define the primary metric and measurement window, and account for confounders and seasonality. Name the pitfalls — selection bias, novelty effects, spillover — and state your decision rule up front. Showing you measure causality, not just correlation, is the signal in Amex's decision-science culture.

Tell me about a time your analysis changed a decision.

STAR. Situation: the decision on the table and the assumption being made. Task: the question you owned. Action: the analysis and, crucially, how you communicated it — a clear visual and a one-line takeaway a non-technical partner could act on. Result: the decision that changed and the measurable impact. Amex weighs business influence and communication as much as technical correctness.

What American Express looks for

Strong, correct SQL: joins, aggregation, and window functions on realistic data
Business-case structure: you decompose a metric and connect data to a decision
Statistical and experimentation sense — incrementality, A/B testing, and measurement
Clear communication of insights to non-technical, financial-services stakeholders
Integrity and attention to data quality (especially given risk and fraud contexts)
Red flag: writing SQL without framing the business question, or confusing correlation with causation

FAQ

Is the Amex Data Analyst interview more SQL or more case-based?

Both. You need fluent SQL, but the differentiator is the analytics/business case — Amex's decision-science culture wants analysts who turn data into a defensible business decision, not just a query.

How much statistics do I need?

Practical stats: hypothesis testing, A/B testing, incrementality, and reading a result correctly. You won't need deep theory, but you should reason clearly about experiments and measurement.

Do I need Python, R, or SAS?

It depends on the team. SQL is the constant; some risk and marketing-analytics roles value Python, R, or SAS, and a few use a take-home. Confirm the stack with your recruiter.

How should I prep for the business case?

Practice structuring open-ended questions: clarify the metric and time frame, decompose the problem, quantify drivers, and end with a recommendation. Communicate the takeaway in one clear sentence.

How long does the process take?

Commonly 3-6 weeks from screen to decision, depending on team and scheduling.

Practice turning data into a decision

Amex's case rounds reward analysts who structure an ambiguous business question out loud and land a clear recommendation. Rehearse the metric-diagnosis and incrementality cases — plus your SQL — with OfferLoop's realtime voice coach before the loop.

Practice this interview out loud →

All American Express interviews →American Express Software Engineer interview →American Express Data Scientist interview →American Express Data Engineer interview →American Express Machine Learning Engineer interview →

OfferLoop is an independent interview-practice tool and is not affiliated with, endorsed by, or sponsored by American Express. All company names and trademarks are the property of their respective owners.

Interview formats vary by team, level and year, and this guide is compiled from general knowledge of publicly discussed hiring processes — treat it as preparation material, not an official description of American Express's current process.