Data, Analytics & AI

Skill areas

Data & Analytics

Data Analyst

A Data Analyst collects, cleans, and interprets structured and unstructured data to help organisations make better decisions. Core work includes writing SQL queries, building dashboards and reports in tools like Power BI or Tableau, conducting statistical analysis, and communicating findings clearly to non-technical stakeholders. Analysts work across departments — marketing, operations, finance, product — translating business questions into data queries and turning results into actionable insight. The role sits on a spectrum between reporting (descriptive analytics) and more advanced analysis involving predictive modelling, A/B testing, or machine learning. Modern Data Analysts are expected to be proficient in SQL and at least one scripting language (typically Python or R), comfortable with cloud data platforms (e.g. BigQuery, Snowflake, Databricks), and capable of designing analyses that are reproducible and well-documented. Data storytelling — presenting findings in a way that influences decisions — is an equally important skill.

Career path3 sample paths

Where this role can take you

Realistic progression routes, linked roles, and typical UK timelines for your next move.

Overview

Data analysts turn raw data into insight for marketing, finance, operations, and product teams—SQL, Excel, and BI tools are daily essentials. Progression leads to senior analyst, analytics engineer, or data science with further study. Lateral moves into business analyst or operations suit those who want more stakeholder facilitation or process ownership.

Progression opportunities

UK data analyst careers often start in reporting teams, FP&A support, or marketing analytics. Senior roles add metric ownership, data governance, and mentoring; some move into data engineering or science via Python and statistics depth. Business analyst paths fit when you prefer defining problems and processes over building pipelines. Operations paths fit when you want to run the teams that act on your dashboards. Public sector and NHS roles emphasise Power BI and open data; fintech emphasises speed and experimentation.

Typical timelines

Expect 2–3 years before a senior data analyst title if you deliver trusted metrics and documentation. Google Data Analytics Certificate or similar helps career changers; employers increasingly test SQL live. Move when stakeholders request you by name and you own a metric tree—not when you have only built pretty charts without adoption.

Interview prep10 questions

Practise the questions you'll actually get asked

Study the most common interview questions for this role, with structured guidance and strong sample answers.

Common interview questions

Answer guidance, sample responses, and tips for each question.

Ranked by how often this question appears across real interviews.

Question 1 · #1
OpeningEasy
Tell me about yourself.
openingdata-analystintroduction
Why they ask
Data analyst interviews still open with your story to gauge communication and analytical journey.
What they want
Concise path into data, tools used, and business impact of your work or projects.
How to structure your answer
1. Current background and data focus
2. Project or work highlight with outcome
3. Why data analyst here
4. Ninety-second structure
Sample answer
I am a mathematics graduate who has completed a data analytics bootcamp and a university project analysing retail sales trends. I enjoy turning raw data into charts and recommendations non-specialists can use. My project used SQL and Python to identify seasonal products driving margin, which my tutor presented to a local shop owner. I also volunteer spreadsheet support for a charity tracking donor campaigns. I am detail-oriented, curious about data quality, and comfortable presenting findings. I want to grow as a data analyst here because your customer analytics team would let me apply those skills to real business decisions at scale.
Common mistakes
- Listing tools without business context
- No portfolio or project mentioned
- Overly academic with no practical delivery
Bonus tips
- Mention portfolio or GitHub if available
- Lead with impact then tools
- Align with domain in job advert
Question 2 · #2
MotivationEasy
Why do you want to work in data analytics?
motivationanalyticscareer
Why they ask
They want lasting interest in insight and decision support, not only coding for its own sake.
What they want
Motivation linking data to business outcomes and continuous learning of tools.
How to structure your answer
1. What you enjoy about analysis
2. Example of insight that mattered
3. Career direction
4. Fit for this role
Sample answer
I want to work in data analytics because I like evidence-based decisions. Data removes guesswork when teams debate what customers do or which campaigns work. I enjoy cleaning messy datasets, finding patterns, and explaining results clearly. A charity dashboard I built helped coordinators focus on high-value donor segments. I am motivated to deepen SQL, BI, and statistics while partnering with stakeholders who act on insights. This role offers the volume and variety of data I need to grow from junior analyst to someone trusted for recommendations that move KPIs.
Common mistakes
- Only mentioning high salaries or remote work
- No stakeholder or business angle
- Claiming data science and analytics are identical without nuance
Bonus tips
- Name a KPI relevant to the employer
- Show humility about learning domain knowledge
- Mention ethical use of data briefly
Question 3 · #3
TechnicalEasy
What is the difference between SQL and Excel?
sqlexceltools
Why they ask
Junior analysts must choose the right tool; this shows practical understanding of scale and workflow.
What they want
Excel for exploration and smaller sets; SQL for querying databases at scale; when to use each.
How to structure your answer
1. Primary purpose of each
2. Scale and collaboration differences
3. Strengths of Excel
4. Strengths of SQL
5. When to combine
Sample answer
Excel is excellent for quick exploration, charts, and smaller datasets on one machine, especially when stakeholders want interactive what-if analysis. SQL is a language for querying relational databases efficiently—filtering millions of rows, joining tables, and aggregating with repeatable scripts. SQL supports governed data sources and scheduled pipelines; Excel files can become fragile when shared. I often prototype in Excel or a BI tool, then productionise logic in SQL for consistency. Choosing depends on data size, refresh needs, and whether multiple people need the same defined metrics.
Common mistakes
- Saying SQL replaces Excel entirely always
- Unable to give a use case for either
- Ignoring data governance
Bonus tips
- Mention Power Query as bridge if you use it
- Give an example row count where you switch tools
- Note version control challenges in Excel
Question 4 · #4
TechnicalMedium
How would you clean messy data?
data-cleaningqualityetl
Why they ask
Most real-world data is dirty; cleaning approach reveals rigour and documentation habits.
What they want
Profile data, handle nulls and duplicates, standardise formats, validate, document transformations.
How to structure your answer
1. Profile and understand sources
2. Handle missing values and duplicates
3. Standardise types and categories
4. Validate against business rules
5. Document steps for reproducibility
Sample answer
I start by profiling: row counts, null rates, distinct values, and obvious outliers. I document the source system and extraction time. I fix structural issues—trim whitespace, standardise dates to ISO format, and map inconsistent category labels using a reference table. For missing values I choose imputation or exclusion based on business rules, never silently dropping without noting impact. I deduplicate using keys agreed with stakeholders. I validate totals against source reports where possible. In SQL or Python I keep transformation scripts in version control so analysis is reproducible and auditable.
Common mistakes
- Deleting outliers without investigation
- No validation or documentation
- Assuming clean exports from source systems
Bonus tips
- Mention great expectations or similar if used
- Give a brief messy data example from a project
- Discuss communicating data quality limits to stakeholders
Question 5 · #5
BehaviouralMedium
Explain a dashboard or project you worked on.
dashboardpower-biproject
Why they ask
They need end-to-end evidence: problem, data, visuals, metrics, and how stakeholders used it.
What they want
Clear narrative with tools, design choices, and impact.
How to structure your answer
1. Business question
2. Data sources and cleaning
3. Metrics and visuals chosen
4. Outcome and feedback
Sample answer
For a bootcamp capstone I built a Power BI dashboard for fictional e-commerce leadership tracking conversion funnel, average order value, and return rate by channel. I joined web analytics and order tables in SQL, handled duplicate sessions, and defined consistent date filters. I used clear KPI cards, a funnel chart, and a slicer for campaign. I wrote a one-page guide explaining definitions so marketing and finance would interpret numbers the same way. Presenting to classmates acting as stakeholders, they said the returns spike by category prompted a hypothetical supplier review. I learned definitions matter as much as charts.
Common mistakes
- Beautiful charts with no business question
- Cannot explain metric definitions
- No mention of audience
Bonus tips
- Bring screenshots if interview allows
- Explain one trade-off you made
- Mention refresh schedule for production dashboards
Question 6 · #6
SituationalMedium
What would you do if data appeared incorrect?
data-qualitytroubleshootingstakeholders
Why they ask
Wrong data drives wrong decisions; they test troubleshooting and communication discipline.
What they want
Pause consumption, trace lineage, compare sources, fix upstream, document, notify stakeholders.
How to structure your answer
1. Stop sharing flawed outputs
2. Reproduce and isolate the issue
3. Check pipeline and source changes
4. Correct and communicate impact
5. Add monitoring if needed
Sample answer
If data looked incorrect I would pause publishing dashboards or reports that depend on it and note the time range affected. I would reproduce the issue with simple checks—row counts, sums against a trusted report, and recent deployments or ETL changes. I would compare staging versus production and speak with data engineering if pipelines broke. Once root cause is found I would fix or backfill, document the incident, and tell stakeholders what was wrong, what changed, and whether past decisions need revisiting. I would suggest a validation test to catch similar issues earlier.
Common mistakes
- Silently adjusting numbers to match expectations
- Blaming users without investigation
- Not informing stakeholders of past impact
Bonus tips
- Mention data contracts or SLAs if familiar
- Give an example where you caught a bug
- Show calm transparency under pressure
Question 7 · #7
SituationalMedium
How would you explain findings to non-technical stakeholders?
communicationstorytellingstakeholders
Why they ask
Analysis only matters if decision-makers understand and trust it.
What they want
Lead with insight and action, simple visuals, defined metrics, and caveats.
How to structure your answer
1. Start with headline insight
2. Use plain language and relevant KPIs
3. Show one clear visual
4. State limitations and recommended action
Sample answer
I would open with the answer they care about: returns rose fifteen percent last quarter, mainly in footwear online. I would avoid SQL jargon and define metrics once—what counts as a return. I would use one simple chart with labelled axes and traffic-light context against targets. I would explain likely drivers based on evidence, not speculation, and separate correlation from causation. I would recommend specific actions—review sizing guide, check warehouse pick errors—and note data gaps. I would send a short written summary after the meeting so finance and marketing align on definitions.
Common mistakes
- Walking through every query step
- Overloading slides with charts
- No recommended next step
Bonus tips
- Practice executive summary in three sentences
- Anticipate questions on data freshness
- Offer office hours for metric definitions
Question 8 · #8
TechnicalMedium
What is the difference between correlation and causation?
statisticscorrelationcausation
Why they ask
Misinterpreting relationships leads to bad business decisions; statistical literacy is essential.
What they want
Clear definitions, example, and how you would test or avoid causal claims incorrectly.
How to structure your answer
1. Define correlation
2. Define causation
3. Example where correlation misleads
4. How to investigate causality carefully
Sample answer
Correlation means two variables move together statistically—for example ice cream sales and sunburn incidents rise in summer. Causation means one variable directly produces change in another. Correlation does not prove causation because confounding variables or coincidence may explain the pattern. Ice cream does not cause sunburn; hot weather drives both. In business, higher app usage might correlate with revenue without causing it if marketing drove both. I would avoid causal language unless supported by experiments, time order, and domain theory, and I would suggest A/B tests or controlled analysis when decisions require proof.
Common mistakes
- Using correlation and causation interchangeably
- No example
- Claiming analysts never need causation
Bonus tips
- Mention spurious correlations humorously but briefly
- Note regression does not alone prove causality
- Show intellectual honesty with uncertain findings
Question 9 · #9
TechnicalEasy
What tools have you used for analysis?
toolssqlpython
Why they ask
Tooling fit speeds onboarding; they want honest proficiency levels.
What they want
SQL, Excel, Python or R, BI tools—with examples of deliverables.
How to structure your answer
1. Query and transformation tools
2. Visualisation tools
3. Collaboration or notebook tools
4. Learning goals
Sample answer
I use SQL in PostgreSQL and BigQuery-style environments for joins and aggregations. I use Excel and Google Sheets for ad hoc analysis and stakeholder-friendly tables. I analyse in Python with pandas and matplotlib for projects requiring scripting. I build dashboards in Power BI and have basic Tableau exposure from bootcamp. I version scripts on GitHub and document metrics in Confluence-style write-ups. I am keen to learn your warehouse, orchestration, and semantic layer tools during onboarding while delivering value with SQL and BI immediately.
Common mistakes
- Tool list with no project linkage
- Claiming expert in everything
- Ignoring git or documentation practices
Bonus tips
- Align to stack in job description
- Mention dbt or Airflow only if true
- Offer portfolio walkthrough
Question 10 · #10
ClosingMedium
Describe a time data helped you make a decision.
impactdecision-makingclosing
Why they ask
Closing with impact reinforces that you analyse to decide, not only to report.
What they want
Concrete decision, data used, outcome, and reflection.
How to structure your answer
1. Decision context
2. Data gathered and analysed
3. Choice made
4. Outcome and learning
Sample answer
Leading a university society event, we debated two ticket price points. I surveyed past attendance, modelled expected turnout at each price using last year's conversion, and estimated revenue minus costs in a simple spreadsheet. Data suggested a slightly lower price with higher volume would maximise revenue and inclusivity. We chose that option and sold out with record first-time attendees. The decision was not only intuitive—it was supported by assumptions we documented. That experience confirmed why I want a career where data guides fair, effective choices rather than gut feeling alone.
Common mistakes
- Analysis with no actual decision
- Fabricated metrics
- Cannot explain limitations of your data
Bonus tips
- Use business workplace example if you have one
- Quantify outcome
- End with enthusiasm for the role

Back to Data, Analytics & AI or all industries.

Expected salary range

Eligibility limits

History and context

Overview

Progression opportunities

Typical timelines

Tell me about yourself.

Why do you want to work in data analytics?

What is the difference between SQL and Excel?

How would you clean messy data?

Explain a dashboard or project you worked on.

What would you do if data appeared incorrect?

How would you explain findings to non-technical stakeholders?

What is the difference between correlation and causation?

What tools have you used for analysis?

Describe a time data helped you make a decision.

Expected salary range

Eligibility limits

History and context

Overview

Progression opportunities

Typical timelines

Tell me about yourself.

Why do you want to work in data analytics?

What is the difference between SQL and Excel?

How would you clean messy data?

Explain a dashboard or project you worked on.

What would you do if data appeared incorrect?

How would you explain findings to non-technical stakeholders?

What is the difference between correlation and causation?

What tools have you used for analysis?

Describe a time data helped you make a decision.

Typical qualifications

Major employers

Why this path matters now

Sample career paths

Analytics leadership track

Business analysis and product

Operations and performance management

Related roles in this library

Career tips