Unite Wings

Deloitte interview questions

16 questions candidates report from Deloitte (USI) interviews, plus the general HR questions every interviewer asks. Each has a model answer and the follow-ups to expect.

Typical selection process

Online test (aptitude, verbal, reasoning), a group discussion or JAM round at some campuses, then a technical interview and an HR or partner interview.

As reported by recent candidates. Always check the latest official notification.

33 questions

HRMotivationMedium

Why consulting, and why Deloitte?

Reported at Deloitte

What they are checking: Whether you understand what consultants do every day, not just the brand.

Model answer

I enjoy problems that combine analysis with people. In my internship I analysed why a local retail chain's repeat customers were falling, and the part I enjoyed most was presenting the findings to the owner and seeing him act on them. Consulting lets me do that across many industries.

Deloitte appeals to me because it is broad: audit, tax, consulting, risk and technology under one roof, so I'd learn how different parts of a client's business connect. The Deloitte USI teams also work with large global clients, which means exposure to big problems early.

I also know the work involves long hours, travel and tight deadlines. I've handled that kind of pressure while managing our college fest budget alongside exams, and I'm ready for it.

BehaviouralStakeholdersMedium

Two stakeholders want conflicting things. How do you handle it?

Reported at Deloitte, Accenture

What they are checking: Facilitation and a focus on business goals.

Model answer

In my internship, the sales head wanted a lead form with only two fields to get more sign-ups, while the operations head wanted eight fields so her team didn't have to call every lead.

I met each separately to understand the goal behind their request: sales cared about the number of leads, operations about the time spent qualifying them. Then I brought data to a joint meeting: our past forms lost about 10% of users for every extra field, and operations spent about 6 minutes calling each lead.

We agreed on a middle path: three fields on the form, and a short optional second step after submission. Sign-ups dropped only slightly and calls per lead fell by half. When stakeholders still can't agree, I take the options and trade-offs to whoever owns the business goal to decide.

TechnicalSQLEasy

Write a SQL query to find the second highest salary.

Reported at TCS, Infosys, Accenture, Deloitte, Amazon

What they are checking: The most-asked SQL question. They want a correct answer and awareness of ties and NULLs.

Model answer

Subquery approach:

SELECT MAX(salary) AS second_highest
FROM employees
WHERE salary < (SELECT MAX(salary) FROM employees);

Window function approach, which generalises to the Nth highest:

SELECT DISTINCT salary
FROM (
  SELECT salary, DENSE_RANK() OVER (ORDER BY salary DESC) AS rnk
  FROM employees
) ranked
WHERE rnk = 2;

DENSE_RANK matters here: if two people share the top salary, RANK would skip 2 and return nothing. The first query returns NULL when there is no second salary, which is often what the interviewer wants.

Likely follow-ups

  • Find the highest salary in each department.
  • What is the difference between RANK and DENSE_RANK?
TechnicalSQLEasy

What is the difference between WHERE and HAVING?

Reported at Deloitte, Accenture, TCS

What they are checking: Understanding of query execution order.

Model answer

WHERE filters rows before grouping; HAVING filters groups after GROUP BY. So WHERE cannot use aggregates like SUM or COUNT, and HAVING can.

SELECT city, SUM(amount) AS revenue
FROM orders
WHERE status = 'delivered'        -- remove rows first
GROUP BY city
HAVING SUM(amount) > 100000;      -- then remove groups

The logical order is FROM, WHERE, GROUP BY, HAVING, SELECT, ORDER BY, LIMIT. This is also why you can't use a SELECT alias inside WHERE in most databases.

TechnicalSQLEasy

Explain the different types of SQL joins.

Reported at TCS, Infosys, Deloitte, Amazon

What they are checking: Fundamentals, usually followed by a question on what row counts a join returns.

Model answer

Using Customers and Orders:

  • INNER JOIN: only customers who have orders, and only orders that have a matching customer.
  • LEFT JOIN: all customers, with their orders if they exist; customers without orders get NULLs. Useful for "customers who never ordered": LEFT JOIN and WHERE orders.id IS NULL.
  • RIGHT JOIN: all orders, with customer details where they exist. Rarely used; rewrite it as a LEFT JOIN.
  • FULL OUTER JOIN: all rows from both, matched where possible.
  • CROSS JOIN: every customer paired with every order; the row count is the product of both tables.
  • SELF JOIN: a table joined to itself, such as employees to their managers in the same table.

A trap: joining on a non-unique key multiplies rows. If a customer has 3 orders and 2 addresses, joining all three tables gives 6 rows for that customer, and sums will be wrong.

TechnicalSQLMedium

What is the difference between ROW_NUMBER, RANK and DENSE_RANK?

Reported at Amazon, Flipkart, Deloitte

What they are checking: Window function fluency, needed for most analyst SQL rounds.

Model answer

All three number rows in a given order; they differ only in how they handle ties. For salaries 100, 90, 90, 80:

  • ROW_NUMBER: 1, 2, 3, 4. Always unique; ties are broken arbitrarily.
  • RANK: 1, 2, 2, 4. Ties share a rank, then a gap.
  • DENSE_RANK: 1, 2, 2, 3. Ties share a rank, with no gap.

Top 3 earners in each department:

SELECT *
FROM (
  SELECT name, department, salary,
         DENSE_RANK() OVER (PARTITION BY department ORDER BY salary DESC) AS rnk
  FROM employees
) t
WHERE rnk <= 3;

Use ROW_NUMBER when you need exactly one row per group, such as removing duplicates or picking the latest record per user.

TechnicalData cleaningMedium

How do you handle missing data in a dataset?

Reported at Accenture, Deloitte, Amazon

What they are checking: Judgement. There's no single answer; they want you to ask why the data is missing.

Model answer

First I find out how much is missing and why. Data missing at random can be handled differently from data missing for a reason; for example, income left blank more often by high earners.

Then I choose based on the situation:

  • Drop rows if very few are affected and they're random.
  • Drop the column if most of it is empty and it isn't critical.
  • Fill with the median for skewed numbers, the mean for roughly normal ones, or the most common value for categories.
  • Fill using related data, such as the median salary for the same job title and city.
  • Add a "was missing" flag column, because missingness itself can be informative.
  • Use forward fill for time series, such as carrying the last known stock price.

I always record what I did and compare results with and without the fix, so stakeholders know how much it affected the conclusions.

TechnicalExcelEasy

What is the difference between VLOOKUP, INDEX-MATCH and XLOOKUP?

Reported at Deloitte, Accenture, Capgemini

What they are checking: Excel fluency, still central in many analyst roles.

Model answer

  • VLOOKUP searches the first column of a range and returns a value from a column to the right. It can't look left, breaks if columns are inserted (because you give a column number), and defaults to approximate match unless the last argument is FALSE.
  • INDEX-MATCH: MATCH finds the row position, INDEX returns the value from any column. It can look left and doesn't break when columns move.
  • XLOOKUP (Excel 365 and 2021): =XLOOKUP(lookup_value, lookup_array, return_array, "Not found"). It looks in any direction, defaults to exact match, has a built-in "if not found" value, and can search from the bottom.

I use XLOOKUP when it's available, and INDEX-MATCH for files that need to work on older Excel versions.

TechnicalStatisticsEasy

When would you report the median instead of the mean?

Reported at Amazon, Flipkart, Deloitte

What they are checking: Statistical common sense.

Model answer

When the data is skewed or has outliers. The mean is pulled by extreme values; the median isn't.

Example: if nine people in a team earn ₹5 lakh and the founder earns ₹1 crore, the mean salary is about ₹14.5 lakh, which describes nobody. The median, ₹5 lakh, describes the typical person.

Order values, delivery times and incomes are usually right-skewed, so I report the median, often with the 90th percentile to show the tail. I use the mean when the data is roughly symmetric or when the total matters, such as average revenue per user used to forecast total revenue.

TechnicalAnalysisMedium

Sales dropped 15% last month. How would you find out why?

Reported at Flipkart, Amazon, Deloitte

What they are checking: Structured investigation instead of jumping to guesses.

Model answer

  1. Check the data first: is it a tracking bug, a missing data load or a change in definition? Many "drops" are pipeline problems.
  2. Check for expected causes: seasonality (compare with the same month last year), fewer days, a festival that moved, or a big sale in the previous month.
  3. Break revenue down: revenue = visitors × conversion rate × average order value. Find which factor moved.
  4. Segment the moving factor: by region, channel, device, product category, and new versus returning customers. A drop concentrated in one segment points to the cause.
  5. Look for internal changes: app releases, price changes, stock-outs, marketing budget cuts, payment failures.
  6. Look for external ones: competitor sales, news, regulation.

For example, if conversion fell only on Android after a release on the 10th, the next step is checking that release with the engineering team. I'd then quantify how much of the 15% each cause explains.

TechnicalRequirementsMedium

How do you gather requirements from stakeholders?

Reported at Accenture, Deloitte, Cognizant, Capgemini

What they are checking: Whether you know the techniques and how to turn vague asks into clear requirements.

Model answer

  1. Identify stakeholders: who uses the system, who pays, who approves, who supports it.
  2. Understand the current process: interviews, watching users do their actual work, and reading existing documents and data.
  3. Run workshops to agree on problems and priorities with groups, especially when departments disagree.
  4. Use questionnaires when there are many users across locations.
  5. Build prototypes or wireframes, because people often only know what they want once they see something.

Then I document requirements as user stories or a BRD, with acceptance criteria, and confirm them with stakeholders. I ask "why" a lot: a request for "an Excel export" is often really a need for a specific weekly report.

TechnicalDocumentationEasy

What is the difference between a BRD and an FRD?

Reported at Accenture, Cognizant, Capgemini, Deloitte

What they are checking: BA vocabulary and documentation practice.

Model answer

  • BRD (Business Requirements Document): what the business needs and why. Written in business language for stakeholders. Example: "Customers must be able to track their order status without calling support, to cut support calls by 30%."
  • FRD (Functional Requirements Document): how the system will behave to meet those needs. Written for designers, developers and testers. Example: "The order page shows status as Placed, Packed, Shipped, Delivered, updated within 5 minutes of a warehouse scan."

The BRD comes first and is approved by the business; the FRD comes from it and is reviewed with the technical team. In agile teams both are often replaced by epics, user stories and acceptance criteria.

TechnicalPrioritisationEasy

How do you prioritise requirements? Explain MoSCoW.

Reported at Accenture, Deloitte, Capgemini

What they are checking: Managing scope when everything is "urgent".

Model answer

MoSCoW sorts requirements into four groups for a release:

  • Must have: without these the release fails or is illegal. Example: payment for an e-commerce launch.
  • Should have: important but there's a workaround. Example: saved cards.
  • Could have: nice if time allows. Example: dark mode.
  • Won't have (this time): agreed to be out of scope for now, which prevents arguments later. Example: a loyalty programme.

The value is in the conversation: stakeholders must agree what goes in Must. A useful rule is to keep Musts to around 60% of the effort, so there's room if estimates are wrong. For comparing many features by value, I'd use a scoring method such as RICE as well.

TechnicalGuesstimateMedium

Guesstimate: how many cups of tea are sold each day in a city of 10 lakh people?

Reported at Deloitte

What they are checking: Structure, sensible assumptions stated out loud, and a sanity check. The final number matters less than the approach.

Model answer

I'll estimate demand from the population, then check it against supply. I'll count tea sold outside the home: stalls, cafés, offices and restaurants.

Demand:

  • Population: 10,00,000.
  • Tea drinkers: excluding young children and people who don't drink tea, say 60%, which is 6,00,000.
  • Cups per drinker per day: about 2, so 12,00,000 cups in total.
  • Bought outside the home: most tea is made at home; assume 25% is bought, so 3,00,000 cups a day.

Sanity check from supply:

  • A typical tea stall sells about 300 cups a day.
  • 3,00,000 ÷ 300 = about 1,000 stalls, or one for every 1,000 people. That seems reasonable for an Indian city when you include office canteens and restaurants.

Answer: roughly 3 lakh cups a day. The biggest uncertainty is the share bought outside, which would be higher in a commercial or student-heavy city.

Tips

  • Say your structure before calculating.
  • Round numbers so the maths stays easy.
  • Always finish with a sanity check.
TechnicalCase interviewHard

Case: a restaurant chain's profits fell 20% this year. How would you find out why?

Reported at Deloitte

What they are checking: The profitability framework, applied with clarifying questions rather than recited.

Model answer

Clarifying questions first: over what period, in all outlets or some, did revenue or costs change, and is the rest of the industry seeing the same?

Then the framework: Profit = Revenue − Costs.

Revenue = number of orders × average order value, broken down by:

  • Channel: dine-in, takeaway, delivery apps.
  • Outlet and city.
  • Menu category.

Costs:

  • Variable: ingredients, packaging, delivery-app commission (often 20–30% of the order).
  • Fixed: rent, salaries, utilities.

Suppose the data shows revenue is flat but costs are up. Digging into costs, delivery orders grew from 30% to 55% of sales, and delivery commissions are eating the margin. Dine-in, which has a higher margin, fell.

Recommendations: raise menu prices on delivery apps slightly, push customers to direct ordering through the restaurant's own app with a small discount, design delivery-specific combos with better margins, and give dine-in customers reasons to come back, such as weekday offers. Then I'd estimate the profit impact of each.

TechnicalCase interviewHard

Case: should a premium coffee chain open stores in tier-2 cities?

Reported at Deloitte

What they are checking: The market-entry structure: market, competition, capabilities, economics, entry mode.

Model answer

I'd structure it around five questions.

  1. Is the market attractive? Size and growth of out-of-home coffee spending in tier-2 cities, the number of young working professionals and students, and rising incomes. Willingness to pay a premium price is the key question.
  2. What is the competition? Local cafés, tea shops, other chains already present, and how they price.
  3. Can we win? Brand recognition in those cities, supply chain reach, and whether our menu suits local taste and price points.
  4. Do the economics work? Lower rent and salaries help, but lower prices and footfall hurt. Estimate break-even per store: fixed costs ÷ (average bill × margin) = cups per day needed, and compare with realistic footfall.
  5. How should we enter? Pilot two or three company-owned stores in the most promising cities, smaller formats, or kiosks in malls and colleges, before committing to a large rollout.

Recommendation: likely yes, but via a pilot in cities with many students and IT workers, with a smaller store format and a cheaper menu range, and with clear targets for sales per store before expanding.

HRIntroductionEasy

Tell me about yourself.

What they are checking: Can you give a short, structured story that connects your background to this job? It sets the agenda for the rest of the interview, so whatever you mention here becomes the next question.

Model answer

I'm [name], a final-year B.Tech Computer Science student at [college], graduating in 2026 with a CGPA of [8.1].

Over the last two years I have focused on backend development. My main project is a hostel complaint tracker built with Node.js and PostgreSQL that 400 students in my hostel now use. I also did a two-month internship at [company], where I wrote REST APIs and cut a slow report query from 9 seconds to under 1 second by adding the right indexes.

Outside class I solve problems on LeetCode, around 300 so far, and I coordinate our college coding club, where I run weekly contests for juniors.

I'm applying here because I want to work on systems that real users depend on, and this role gives me that from day one.

Tips

  • Use present, past, future: who you are, what you have done, why this job.
  • Keep it under 90 seconds. Mention only things you are happy to be questioned on.
  • Do not repeat your resume line by line, and skip family details unless asked.

Likely follow-ups

  • Tell me more about the complaint tracker. What was hardest?
  • Why did you add indexes rather than caching?
HRSelf-awarenessEasy

What are your strengths?

What they are checking: Whether you know what you are good at and can prove it with an example, rather than listing adjectives.

Model answer

My biggest strength is that I break unfamiliar problems into small pieces and keep going until they work.

In my second year our team had to build an attendance app using face recognition, and none of us had used OpenCV. I split the work into three parts: capturing images, training the model, and matching faces in real time. I took the matching part, read the documentation and two tutorials over a weekend, and had a working version in five days. We demoed it at the department expo and it marked attendance for a class of 60 with about 95% accuracy.

I'd bring the same habit here when I'm handed a codebase or tool I haven't used before.

Tips

  • Pick one or two strengths that matter for the role.
  • Every strength needs a short story with a result.

Likely follow-ups

  • What would your teammates say is your strength?
  • What did you do about the 5% the app got wrong?
HRSelf-awarenessMedium

What is your greatest weakness?

What they are checking: Honesty and self-improvement. They want a real weakness that will not stop you doing the job, plus what you are doing about it.

Model answer

I used to find it hard to say no when people asked for help, and I would take on more than I could finish.

In my third year I agreed to help with three club events in the same month as my mid-semester exams, and my marks in one subject dropped. Since then I have started keeping a simple weekly plan in Notion. Before I agree to something new I check it against what is already there, and if it does not fit I offer a smaller piece or a later date.

This semester I managed the coding club contest and my project deadlines without missing either. I still have to watch it, but I catch it much earlier now.

Tips

  • Avoid fake weaknesses like "I'm a perfectionist" or "I work too hard".
  • Never name a weakness that is core to the job, such as weak coding for a developer role.
  • Spend most of the answer on what you changed.

Likely follow-ups

  • Give me another weakness.
  • How would you handle a manager who keeps adding work?
HRCareer goalsEasy

Where do you see yourself in five years?

What they are checking: Whether your plans fit this company and whether you are likely to stay long enough to repay the training.

Model answer

In five years I want to be someone my team relies on for a whole area of the product, not just individual tasks.

For the first two years I want to get really good at the technology I'm assigned, earn a certification in it, and understand how our client's business works. After that I'd like to lead a small module, review juniors' code and talk directly with the client.

This company's training and the range of projects here make that path realistic, which is a big reason I applied.

Tips

  • Show growth inside the company.
  • Do not say you plan to leave for an MBA or a startup.
0students
taking off