What Does a Data Analyst Do? Daily Tasks, Skills, Tools, Salary Path and How AI Is Changing the Role

A data analyst collects, cleans, analyses and visualises data to answer business questions and help people make better decisions. Day to day, that means writing SQL queries, building spreadsheets and dashboards, spotting trends and problems, explaining what the numbers mean, and recommending actions to managers and teams in sales, marketing, finance, operations or product.

"Data analyst" is one of the most common entry points into a data career, and one of the most misunderstood. It's less about advanced maths and more about asking good questions, getting reliable data, and communicating clearly. This guide explains what data analysts really do, the skills and tools they use, how the role compares with similar jobs, and how AI is reshaping the work.

What a data analyst does, in one example

A marketing manager asks: "Our sales dropped 12% last month. Why?"

A data analyst would:

1. Clarify the question: which sales, which regions, compared with what period?
2. Get the data: query orders, website traffic, campaign spend and pricing from the data warehouse or source systems.
3. Clean it: fix missing values, duplicates, mismatched product codes and time zones.
4. Analyse: break sales down by channel, region, product and customer type; compare with last month and last year; check for seasonality.
5. Find the cause: e.g., paid search traffic fell 40% after a budget cut, while other channels were stable.
6. Visualise and explain: a simple chart and a short summary with the impact in money.
7. Recommend: restore the search budget on the best-performing campaigns, and set up an alert for future drops.

That cycle (question → data → clean → analyse → communicate → recommend) is the heart of the job.

Typical daily tasks

• Writing SQL queries to pull and combine data
• Cleaning and preparing data (often 40–60% of the work)
• Building and maintaining dashboards and reports in Power BI, Tableau, Looker or Excel
• Ad-hoc analysis to answer questions from stakeholders
• Monitoring KPIs and investigating unusual changes
• A/B test analysis and campaign performance reviews
• Defining metrics with business teams (what counts as an "active customer"?)
• Presenting findings in meetings, slides or written summaries
• Documenting data sources, definitions and logic
• Automating repetitive reports (see How to automate tasks with AI)

The data analysis process

(A deeper walk-through: How to analyze data.)

Types of data analysis

• Descriptive: what happened? (monthly sales report)
• Diagnostic: why did it happen? (root cause of a drop)
• Predictive: what's likely to happen? (forecasting demand)
• Prescriptive: what should we do? (recommend budget allocation)

Most data analyst work is descriptive and diagnostic, with some predictive work.

Core skills

Technical

• SQL: the most important technical skill; joins, aggregations, window functions, CTEs
• Excel / Google Sheets: pivot tables, lookups (XLOOKUP), Power Query, charts
• BI tools: Power BI (DAX), Tableau or Looker
• Statistics basics: averages, distributions, variance, correlation, significance testing, sampling
• Python or R (increasingly expected): pandas, visualisation libraries, notebooks
• Data modelling basics: how tables relate; star schemas (see What is a data warehouse?)

Business and soft skills

• Asking the right questions and understanding the decision behind them
• Domain knowledge (marketing, finance, operations, healthcare)
• Data storytelling: turning numbers into a clear narrative
• Attention to detail and scepticism about data quality
• Stakeholder management: setting expectations and saying no to unclear requests

Tools data analysts use

Types of data analysts

• Business analyst / business data analyst: operational and strategic questions
• Marketing analyst: campaigns, attribution, customer acquisition
• Financial analyst: budgets, forecasts, profitability
• Product analyst: user behaviour, feature adoption, experiments
• Operations / supply chain analyst: efficiency, inventory, logistics
• Healthcare analyst: clinical and operational data
• HR / people analyst: hiring, retention, workforce planning
• BI analyst: dashboards and reporting infrastructure

Data analyst vs similar roles

How AI is changing the data analyst role

AI assistants now help analysts:

• Write and debug SQL and Python faster
• Clean data and suggest transformations
• Summarise results and draft reports
• Explore datasets through natural-language questions
• Build charts and dashboard drafts (e.g., Copilot in Power BI and Excel)
• Document queries and metrics

What AI doesn't replace: understanding the business, judging whether the data is right, choosing the right question, spotting misleading conclusions, and persuading people to act. Analysts who use AI well can handle more questions, faster, and spend more time on the high-value work. Routine report-building is the part most exposed to automation, so the role is shifting toward analysis, judgement and communication.

Tip: never paste confidential or personal data into public AI tools; use your organisation's approved tools.

How to become a data analyst

1. Learn Excel/Sheets and SQL well (2–3 months of consistent practice).
2. Learn one BI tool (Power BI is very common in job ads).
3. Learn statistics basics and, ideally, Python.
4. Build 3–5 portfolio projects with real public data: a cleaned dataset, a dashboard, a written analysis with recommendations.
5. Get a certificate if useful for structure (e.g., Google Data Analytics, Microsoft Power BI Data Analyst).
6. Apply for entry-level roles or move internally from a business role where you already use data.

(Is it hard, and is it a good career? See Is data analytics hard?.)

Frequently asked questions

What does a data analyst do?

A data analyst collects, cleans, analyses and visualises data to answer business questions, builds reports and dashboards, investigates trends and problems, and communicates findings and recommendations to decision-makers.

What skills does a data analyst need?

SQL, Excel or Google Sheets, a BI tool such as Power BI or Tableau, basic statistics, often Python or R, plus business understanding, critical thinking and clear communication.

Do data analysts need to code?

Most use SQL daily, which is a form of coding. Python or R is increasingly expected for more advanced roles, but many analyst jobs rely mainly on SQL, spreadsheets and BI tools.

What is the difference between a data analyst and a data scientist?

Data analysts mainly explain what happened and why using existing data, reports and dashboards. Data scientists focus more on statistical modelling, machine learning and prediction.

Will AI replace data analysts?

AI automates parts of the work, such as writing queries, cleaning data and drafting reports, but analysts are still needed to frame questions, validate data, interpret results and influence decisions. The role is evolving rather than disappearing.

What does a day in the life of a data analyst look like?

Checking dashboards for unusual changes, writing SQL to answer stakeholder questions, cleaning data, building or updating reports, meeting with business teams, and presenting findings with recommendations.