Curated interview practice questions with answers. These are practice scenarios, not a verified list of past interview questions.
About Uber hiring for Data Analytics roles
Uber operates ride-hailing, food delivery through Uber Eats, and freight platforms in many countries. Its Bengaluru and Hyderabad tech centres contribute to global products in maps, payments, marketplace and safety. It hires in India for software engineers, data scientists, data engineers and analysts working on pricing, ETAs, marketplace efficiency and operations.
For Data Analytics roles, this Uber practice set focuses most on Dashboards and KPIs (6), Power BI (5), Python for Analysis (4), Web & Marketing Analytics (4). It has 2 easy, 29 medium, 9 hard questions, so you can start easy and work up to the harder ones.
Company type
Product
Headquarters
San Francisco, California, USA
Main India offices
Bengaluru, Hyderabad, Gurugram
Uber Data Analytics interview process
Online coding assessment
Technical phone screen
Coding and problem-solving rounds
System design or analytics case round
Hiring manager round
Typical pattern based on public candidate experiences. Rounds vary by role, team and year.
Solved for UberStart practicing
3 QUESTIONS FREESign up to unlock the full question bank and track your progress.Unlock every question and answer, then continue practising company-wise and package-wise.
Table calculations are computed on the aggregated results already in the view, not on the underlying rows. Examples are running total, percent of total, difference from previous, rank and moving average. Their result depends on the direction and partition you set with Compute Using, for example table across or by specific dimensions. They are quick to add from Quick Table Calculation, but incorrect Compute Using settings are a frequent cause of wrong numbers.
WITH touches AS (
SELECT t.user_id, t.channel, t.touch_ts,
ROW_NUMBER() OVER (PARTITION BY t.user_id ORDER BY t.touch_ts) AS first_rank,
ROW_NUMBER() OVER (PARTITION BY t.user_id ORDER BY t.touch_ts DESC) AS last_rank
FROM marketing_touches t
JOIN conversions c
ON c.user_id = t.user_id
AND t.touch_ts <= c.converted_ts
)
SELECT user_id,
MAX(CASE WHEN first_rank = 1 THEN channel END) AS first_touch,
MAX(CASE WHEN last_rank = 1 THEN channel END) AS last_touch
FROM touches
GROUP BY user_id;
04
Write SQL to incrementally load new and changed products into a dimension using MERGE.
MERGE INTO dim_product t
USING stg_product s
ON t.product_id = s.product_id
WHEN MATCHED AND (t.name <> s.name OR t.category <> s.category) THEN
UPDATE SET name = s.name,
category = s.category,
updated_at = CURRENT_TIMESTAMP
WHEN NOT MATCHED THEN
INSERT (product_id, name, category, updated_at)
VALUES (s.product_id, s.name, s.category, CURRENT_TIMESTAMP);
05
Write SQL to group raw page events into sessions using a 30-minute inactivity rule.
WITH ordered AS (
SELECT user_id, event_ts,
LAG(event_ts) OVER (PARTITION BY user_id ORDER BY event_ts) AS prev_ts
FROM page_events
),
flagged AS (
SELECT *,
CASE WHEN prev_ts IS NULL
OR event_ts - prev_ts > INTERVAL '30 minutes'
THEN 1 ELSE 0 END AS new_session
FROM ordered
)
SELECT user_id, event_ts,
SUM(new_session) OVER (PARTITION BY user_id ORDER BY event_ts) AS session_no
FROM flagged;
06
What is the difference between Goal Seek and Solver?
Goal Seek changes one input to hit a target output; Solver optimises with several variables and constraints, for example maximising profit under budget limits.
07
Your NOT IN query unexpectedly returns zero rows. Why might that happen, and how do EXISTS and IN differ?
IN checks whether a value is in a list or subquery result; EXISTS checks whether a subquery returns at least one row and stops at the first match. Results are the same in most cases, and modern optimisers often run them similarly. The big difference is NOT IN with NULLs: if the subquery returns any NULL, NOT IN returns no rows. NOT EXISTS behaves correctly, so prefer it for anti-joins.
I would first agree on definitions: gross or net, including tax or returns, order date or invoice date. Then compare one small slice, such as one month and one region, in both. I check filters on the report and page, relationships that drop unmatched rows, refresh time versus the Excel extract time, and duplicate rows. Once the gap is explained, I document the definition in the report so the question does not repeat.
10
Write pandas code to fill missing values in a column with the median of each group.
# Fill missing delivery_days with the median for that city
df['delivery_days'] = (
df['delivery_days']
.fillna(df.groupby('city')['delivery_days'].transform('median'))
)
# Rows in a city with no values at all stay NaN; fall back to overall median
df['delivery_days'] = df['delivery_days'].fillna(df['delivery_days'].median())
11
What is method chaining in pandas and why is it useful?
Method chaining writes a sequence of transformations as one expression, each step returning a new DataFrame, for example df.query('sales > 0').assign(margin=lambda x: x.profit / x.sales).groupby('region').agg(total=('sales', 'sum')). It avoids many temporary variables, reads top to bottom like a recipe, and reduces SettingWithCopy issues. Wrap the chain in brackets with one step per line, and use pipe for custom functions.
12
Write a DAX measure for cumulative sales from the start of the data to the selected date.
Cumulative Sales =
VAR maxDate = MAX('Date'[Date])
RETURN
CALCULATE(
[Total Sales],
FILTER(ALL('Date'[Date]), 'Date'[Date] <= maxDate)
)
-- for a running total that resets each year use TOTALYTD([Total Sales], 'Date'[Date])
13
Users say a Power BI report still shows yesterday morning's data, although a refresh is scheduled every few hours. What do you check?
Open the semantic model's refresh history in the Service to see whether refreshes failed, for example due to expired credentials, an offline gateway or a timeout. Check whether the report is connected to a different copy of the model, whether the source itself was updated, and whether incremental refresh only covers some partitions. Also confirm the schedule time zone. Finally, add a 'data as of' timestamp on the report.
14
What is statistical significance vs practical significance?
A z-score tells how many standard deviations a value is from the mean: z = (x minus mean) divided by standard deviation. A delivery time with z = 2.5 is unusually slow. Analysts use z-scores to flag outliers, often beyond plus or minus 3, to compare values on different scales, such as marks in different exams, and to look up probabilities when data is roughly normal.
16
How would you design an A/B test for a new checkout button?
Define one primary metric, such as checkout conversion, plus guardrails like refund rate. State the hypothesis and minimum detectable effect, then calculate sample size and duration for chosen power and significance. Randomise users, not sessions, into control and variant, check the split is balanced, and run for full weeks to cover weekday patterns. Do not peek and stop early. Finally analyse with a proportion test and check key segments.
A transaction fact stores one row per event, such as each payment. A periodic snapshot stores the state at regular intervals, such as daily stock per store or month-end loan balance. An accumulating snapshot has one row per process instance, like an order, updated as it reaches milestones such as placed, packed, shipped and delivered, with a date for each. Choose based on whether the question is about events, states or process durations.
Broadcasting lets NumPy perform operations on arrays of different shapes without copying data, by virtually stretching the smaller array. For example, subtracting a 1D array of column means from a 2D array centres every column in one line: X - X.mean(axis=0). Shapes are compatible when trailing dimensions are equal or one of them is 1. It is a key reason vectorised code is much faster than Python loops.
20
Sales data from different countries mixes currencies and some weights are in grams while others are in kilograms. How do you handle it?
Never add mixed units. Identify the unit for each row from a currency or unit column, or infer it from the source system. Convert to one standard, for example INR using a dated exchange rate table joined on transaction date, and kilograms using a conversion factor. Keep the original value and unit columns for audit. Document which rates were used, and flag rows where the unit could not be determined.
21
A manager asks for a dashboard that refreshes every minute. How do you decide the right refresh frequency?
Match refresh to how quickly decisions are made. A monthly board pack needs daily refresh at most, while a warehouse dispatch board may need 15-minute or live updates. Faster refresh costs more in source load, licensing and complexity, and can show incomplete data. Ask how quickly someone will act on a change, check when upstream systems finish loading, and always display the last refreshed time on the dashboard.
A factless fact table records that an event or relationship happened but has no numeric measures, only dimension keys. Examples are student attendance (student, class, date) or which products were on promotion in which store each day. You analyse it by counting rows. A promotion coverage table is useful for questions like which promoted products did not sell, by comparing it with the sales fact.
25
You launched a dashboard a month ago and usage logs show almost nobody opens it. What do you do?
Talk to the intended users before redesigning. Find out whether they know it exists, whether it answers their real questions, whether they trust the numbers, and whether it fits their routine. Common fixes are fewer, sharper KPIs, faster loading, a scheduled email snapshot, or embedding it in a meeting that already happens. Then measure usage again. Sometimes the honest answer is that the dashboard is not needed and can be retired.
26
Write SQL to find the median order value overall and per city.
-- Works in PostgreSQL, Snowflake, SQL Server (with OVER for SQL Server)
SELECT
city,
PERCENTILE_CONT(0.5) WITHIN GROUP (ORDER BY order_value) AS median_value,
AVG(order_value) AS mean_value
FROM orders
GROUP BY city
ORDER BY median_value DESC;
27
Tell me about a time you worked with messy data under a tight deadline.
Situation: for a college placement project, I received survey data from three forms with different column names two days before submission. Task: combine and analyse it. Action: I mapped columns to a standard list, cleaned obvious issues like duplicate responses and mixed date formats in Power Query, and noted assumptions. I focused analysis on the five questions that mattered most. Result: we submitted on time, and the faculty praised the documented cleaning steps.
28
Write an Excel formula to look up a price using two criteria, product and city.
=XLOOKUP(1, (A2:A500=G2) * (B2:B500=H2), C2:C500, "Not found")
-- A = product, B = city, C = price
-- G2 and H2 hold the product and city to find
-- each condition returns TRUE/FALSE; multiplying gives 1 only where both match
29
Write a DAX measure for a KPI card showing achievement against target and a colour for conditional formatting.
Achievement % =
DIVIDE([Total Sales], [Sales Target])
KPI Colour =
VAR a = [Achievement %]
RETURN
SWITCH(
TRUE(),
ISBLANK(a), "#9E9E9E",
a >= 1, "#2E7D32",
a >= 0.9, "#F9A825",
"#C62828"
)
-- apply KPI Colour via Format > Callout value > fx > Field value
This is an encoding mismatch: a UTF-8 file was read as a different encoding such as Latin-1 or Windows-1252. Re-read it with the right encoding, for example pd.read_csv(path, encoding='utf-8') or 'utf-8-sig' for files with a byte order mark, or set the file origin in Excel and Power Query import. Avoid re-saving through tools that change encoding, and agree on UTF-8 for all exchanges, especially for Hindi or other Indian language text.
31
What is the order of operations in Tableau and why does it matter?
Tableau applies filters in a fixed order: extract filters, data source filters, context filters, then FIXED LOD calculations, dimension filters, INCLUDE and EXCLUDE LODs, measure filters, and finally table calculations and table calc filters. It matters because a calculation only sees data filtered before it. For example, a FIXED LOD ignores a normal region filter unless that filter is added to context. Many wrong totals come from not knowing this order.
32
Site conversion rate rose from 2.0% to 2.4%, but total revenue stayed flat. What would you check?
Check the components: sessions may have fallen, for example after cutting paid campaigns, which removes low-intent traffic and lifts the rate without new sales. Average order value may have dropped because of discounts or a mix shift towards cheaper products. Also check whether conversion definitions or tracking changed. Revenue equals sessions times conversion rate times AOV, so break down the change across those three factors by channel.
Use a star schema, remove unused columns, prefer measures, reduce visuals per page, limit high-cardinality fields and use aggregations or incremental refresh.
34
The finance dashboard and the sales dashboard show different revenue for last month. How do you resolve it?
Start by listing both definitions: gross or net of returns and discounts, with or without tax, booking date or invoice date, and which entities are included. Compare at a detailed level for one region to find exactly where the difference arises. Usually both are correct for their purpose. Agree with both owners which is the official revenue KPI, rename the other clearly, such as booked sales, and document it in the metric dictionary.
It extracts and transforms data using recorded steps in M language, such as merging, unpivoting, changing types and removing errors, before loading to the model.
36
Average delivery time improved by 4 minutes, but customer complaints about late deliveries went up. How do you investigate?
An average can improve while the tail gets worse. I would look at the distribution: median, 90th and 95th percentiles and the share of orders beyond the promised time. Then break it down by city, time slot and restaurant to see where late orders cluster. Also check whether promised times changed, whether complaint tracking changed and whether a few very fast orders pulled the mean down.
37
Business teams have built hundreds of their own reports, each with its own revenue logic. How do you balance self-service with governed dashboards?
A good model is a governed core plus self-service on top. The central team builds certified datasets with agreed metric definitions and a small set of official dashboards for key KPIs. Business users then build their own views on those certified datasets, not on raw extracts. This gives speed and flexibility without twenty versions of revenue. Training, naming conventions and periodic clean-up of unused content keep it healthy.
38
Tell me about a time a dashboard you built changed a business decision.
Situation: our operations team assumed late deliveries were evenly spread. Task: I built a delivery performance dashboard. Action: I added a breakdown by courier partner and pin code cluster with a weekly trend, and reviewed it with the operations manager. Result: it showed one partner caused 60 percent of delays in two zones. The company shifted those zones to another partner, and on-time delivery rose from 84 to 92 percent in two months.
39
How do you remove duplicates in Excel but keep the latest record?
Sort by date (newest first), then Data → Remove Duplicates on the key column — Excel keeps the first occurrence. Or use UNIQUE / a COUNTIFS helper column.
40
An inner join between orders and customers returns fewer orders than the orders table has. What could be wrong?
Orphan records are child rows whose foreign key has no matching parent, such as orders with a customer_id that does not exist in the customer table. They break joins silently: an inner join drops them and totals fall short. Find them with a LEFT JOIN from the child to the parent and filter where the parent key IS NULL. Then fix the source, add an Unknown member, or report them separately.
Tips to crack the Uber Data Analytics interview
Practise case questions on marketplace balance, such as how surge pricing or driver incentives affect supply and demand.
Revise SQL and statistics for analyst roles, including cohort analysis and experiment design for app features.
Study graph and geospatial problems like shortest paths, which connect naturally to Uber's mapping and dispatch systems.
Keep a one-minute story ready for every project on your CV: problem, your part, tools, result.
Practise the hard questions on this page aloud; Uber interviewers usually ask follow-ups on your first answer.
Frequently asked questions
How many rounds are there in the Uber Data Analytics interview?
Candidates usually go through 5 stages: Online coding assessment; Technical phone screen; Coding and problem-solving rounds; System design or analytics case round; Hiring manager round. The exact rounds change with the role, team, location and your experience, so confirm the process with the recruiter.
Which topics should I prepare for the Uber Data Analytics interview?
This practice set leans on Dashboards and KPIs (6), Power BI (5), Python for Analysis (4), Web & Marketing Analytics (4). Revise those first, then practise explaining one or two of your own projects end to end, because most Uber interviewers ask you to walk through real work.
Are these the exact questions asked at Uber?
No. These are curated practice questions that match the skills and difficulty usually tested in product interviews like Uber's. They are not a leaked or verified list of past questions.
Where does Uber hire Data Analytics professionals in India?
Uber has large teams in Bengaluru, Hyderabad, Gurugram. Openings and locations change often, so check the Uber careers page and filter by city and experience.