Data analyst

What does a data analyst do?

A data analyst collects, cleans and studies data to help an organisation make decisions.

You turn raw figures from sales systems, surveys, web tools and databases into reports, dashboards and clear recommendations.

Data analysts work in sectors such as banking, government, retail, logistics, healthcare and technology. You usually report to an analytics manager, a head of data, or a business function such as finance or marketing.

You work with managers who ask the questions, data engineers who maintain the data pipelines, and data scientists who build models. You also explain findings to people who do not work with data.

What the work involves

• Meet business teams to understand what they need to know, and turn the request into a clear analysis question.
• Pull data from internal databases, spreadsheets and outside sources, then check it for errors, gaps and duplicates.
• Clean and combine the data using SQL, Excel or Python so it is ready to analyse.
• Run the analysis, such as trends, comparisons or A/B test results, and note what the numbers show and where they fall short.
• Build dashboards and reports in a tool such as Power BI or Tableau, and update them on a schedule.
• Present findings in everyday business language and suggest the next action.

What you might build in your first months

These are our examples of first projects for a new data analyst.

Real tasks depend on the employer.

Pay in Singapore

Back to top

Skills a data analyst needs

Back to top

How to become a data analyst in Singapore

These are the main routes into data analysis in Singapore.

Check each official page for current intakes and entry rules.

Data analyst courses and funding in Singapore

These official schemes help pay for data courses in Singapore.

Check each page for current eligibility.

Where the jobs are

Data analysts work in most sectors that keep business records.

These official facts relate to data work in Singapore.

• The SkillsFuture Skills Framework for Infocomm Technology places the data analyst role in its Data and Artificial Intelligence track. SkillsFuture
• National AI Strategy 2.0 was launched on 4 December 2023. MDDI
• The launch speech gave advanced manufacturing, financial services, healthcare, education and public services as examples of areas for scaling AI use. MDDI

Rules and frameworks in Singapore

The Personal Data Protection Act (PDPA) is the law on personal data.

The other items are guidance from the Personal Data Protection Commission (PDPC).

Back to top

Data analyst job description

A data analyst gathers data from internal systems and outside sources, checks its quality and analyses it to answer business questions.

The role produces reports, dashboards and recommendations for teams such as marketing, operations, finance and product.

Responsibilities

• Work with business teams to define questions and turn them into reporting and analysis requirements.
• Recommend which data and data sources are needed to answer each question.
• Extract data from databases, files and external sources using SQL and other tools.
• Clean, transform and validate data, and run quality checks before analysis.
• Analyse trends, segments and experiments using statistics and suitable methods.
• Design and maintain dashboards and automated reports that people can read without help.
• Explain results in clear language and recommend actions to managers and other decision makers.
• Document data definitions, calculations and assumptions so others can reproduce the work.
• Handle personal and sensitive data according to company policy and Singapore data protection rules.
• Support data engineers and data scientists on shared datasets and data models.

Requirements

• Diploma or degree in a field such as statistics, mathematics, computer science, information systems, business or economics. Some employers accept equivalent experience or a data analytics programme.
• Typically 0 to 3 years of experience for an analyst role. Entry-level roles often accept fresh graduates with project or internship work.
• Working SQL for querying, joining and summarising data, and strong Excel skills.
• Experience with at least one dashboard tool such as Power BI, Tableau or Looker Studio.
• Basic statistics, and Python or R for larger or repeated analysis, are often listed as useful or required.
• Clear communication, and the ability to explain numbers to people outside the data team.

Back to top

Data analyst career path

Most data analysts start in an associate or junior analyst role and move up through senior analyst to team lead.

Some later move into management or into more technical data roles.

1. Associate data analyst / associate data engineer

   Entry · SkillsFuture title

   You prepare and clean data, build basic reports and learn the business and its data sources under guidance.

2. Data analyst

   Mid · SkillsFuture title

   You own analyses from request to presentation, design dashboards and advise business teams on what the data says.

3. Senior data analyst

   Senior

   You handle the hardest questions, set standards for data quality and coach junior analysts.

4. Analytics manager

   Lead

   You lead a team of analysts, agree priorities with business heads and manage reporting and delivery.

5. Head of analytics

   Head

   You set the data and analytics direction for the organisation and report on results to senior leaders.

Steps marked "SkillsFuture title" follow the career pathway in the SkillsFuture Skills Framework. The others are titles commonly used in Singapore.

Related jobs and their pay

Back to top

Data analyst interview questions

We wrote these questions from the skills this job needs. They are not an official question bank.

1. Write a SQL query to find the top three customers by spend in each month. How would you do it?

   What a good answer covers

   • Aggregates spend by customer and month first
   • Uses a window function such as RANK or ROW_NUMBER partitioned by month
   • Explains how ties are handled and why
   • Filters on the rank in an outer query or common table expression

2. You join an orders table to a customers table and the row count goes up. What could be wrong?

   What a good answer covers

   • Checks whether the join key is unique in each table
   • Explains that duplicate keys multiply rows
   • Compares row counts before and after the join
   • Describes how to fix it by deduplicating or aggregating first
   • Knows when to use a left join instead of an inner join

3. A manager asks what a p-value and a confidence interval mean. How do you explain them without jargon?

   What a good answer covers

   • Describes the p-value as how surprising the result would be if there were no real effect
   • Describes the interval as a range of plausible values for the true figure
   • Avoids saying the p-value is the chance the result is true
   • Mentions that a small effect can be significant with a large sample
   • Relates it to a business decision

4. You run an A/B test on a checkout page. How do you decide the sample size and when to stop?

   What a good answer covers

   • Defines one main metric and the smallest change worth detecting
   • Estimates sample size before the test starts
   • Waits for the planned sample instead of stopping when the result looks good
   • Checks that users are split randomly and evenly
   • Considers business cycles such as weekday and weekend behaviour

5. You receive a survey file with missing values, duplicate responses and inconsistent text entries. What steps do you take?

   What a good answer covers

   • Profiles the data first to see how large each problem is
   • Removes or flags duplicates with a clear rule
   • Standardises text categories and formats
   • Decides whether to drop, fill or keep missing values, and explains why
   • Keeps a log of every change so the work can be repeated

6. You need to show monthly revenue by region to senior managers. Which chart do you choose and how do you lay out the dashboard?

   What a good answer covers

   • Chooses a line chart for trend and a bar chart for comparison
   • Limits the number of colours and removes clutter
   • Puts the main message in the title
   • Adds filters only where users need them
   • Tests the layout with a real user

7. Twelve monthly Excel files arrive with the same layout every month. How do you combine them without copying and pasting?

   What a good answer covers

   • Uses Power Query to load all files from a folder
   • Applies the same cleaning steps to every file
   • Sets data types and removes header rows consistently
   • Refreshes the query when a new file is added
   • Explains how to handle a file with a changed column name

8. In Python, how would you merge two datasets, summarise them by group and spot outliers?

   What a good answer covers

   • Uses pandas merge with a clear key and checks the result size
   • Uses groupby with named aggregations
   • Checks for outliers with summary statistics, box plots or an interquartile range rule
   • Investigates whether an outlier is an error or a real value before removing it

1. Your dashboard shows a different sales total from the finance team's report. What do you do?

   What a good answer covers

   • Compares definitions first, such as gross or net sales and date cut-offs
   • Traces the numbers back to the source tables
   • Checks filters, currency and refresh timing
   • Agrees one definition with finance and documents it
   • Tells stakeholders about the gap before they find it

2. A manager asks you why sales dropped last month, with no other detail. How do you approach it?

   What a good answer covers

   • Asks what decision the answer will support
   • Confirms the drop is real by checking against past months
   • Breaks the figure down by product, region, channel and customer type
   • Forms a few possible explanations and tests each with data
   • Reports what is known, what is not, and a suggested next step

3. A colleague suggests pasting a customer dataset into ChatGPT to speed up your analysis. What do you say?

   What a good answer covers

   • Checks company policy on approved AI tools first
   • Keeps personal and confidential data out of public tools
   • Suggests using sample, masked or synthetic data instead
   • Says AI output must be checked against the real data
   • Describes where AI does help, such as drafting SQL or explaining code

1. Tell us about a time your analysis changed someone's mind or a decision.

   What a good answer covers

   • Describes the situation and the original view
   • Explains what data and method you used
   • Shows how you presented it for the audience
   • States the result of the decision

2. Describe a time you found a mistake in your own work after you had shared it.

   What a good answer covers

   • Tells the stakeholders quickly and clearly
   • Explains the cause and the fix
   • Describes the check you added to prevent it happening again
   • Takes ownership of the error

1. You are given a customer file that contains NRIC numbers, phone numbers and email addresses for an analysis. How do you handle it under Singapore data protection rules?

   What a good answer covers

   • Names the Personal Data Protection Act (PDPA) as the law that governs personal data
   • Uses only the fields the analysis needs
   • Masks or removes identifiers such as NRIC numbers before analysis
   • Limits access and storage, and follows the retention rules of the company
   • Checks with the data protection officer when unsure

2. You are analysing weekly sales for a Singapore retailer. How do you account for local events and seasons?

   What a good answer covers

   • Marks festive periods such as Chinese New Year and Hari Raya
   • Includes school holidays, public holidays and year-end sales
   • Compares with the same period in earlier years as well as the previous week
   • Separates one-off promotions from the usual trend
   • Tells the business how much of a change is seasonal

Common questions about a data analyst job

Sources