Becoming a Data Analyst in the UK (2026): Skills and Portfolio

To become a data analyst in the UK in 2026, build three things in this order: working SQL, a Power BI report you can explain line by line, and a small portfolio of projects on real UK open data. Certification such as Microsoft's PL-300 then proves the Power BI part. Employers hire on evidence, and UK public data lets you build that evidence for free.

At a glance

• Core tools: SQL, Power BI with Power Query and DAX, Excel, and enough statistics to explain a trend honestly.
• Certification: PL-300 for Microsoft Certified: Power BI Data Analyst Associate. You need 700 or more to pass.
• Free UK data: the ONS API, HM Land Registry Price Paid Data, Department for Transport road safety data and data.police.uk.
• Licence: these publishers use the Open Government Licence v3.0, which allows reuse with attribution.
• Skills map: the Level 4 Data analyst apprenticeship standard (ST0118) lists what UK employers expect.
• Portfolio: three projects, each with a question, a method, findings and limitations.

What UK employers mean by "data analyst"

Use the apprenticeship standard as your map. Skills England's Data analyst standard (ST0118, Level 4) is written with employers, and its knowledge, skills and behaviours read like a job description. They include current legislation and its application to the safe use of data, principles of statistics, descriptive, predictive and prescriptive analytics, using data systems securely, and collating and interpreting qualitative and quantitative data. You do not need to be an apprentice to use it as a checklist.

Notice what sits alongside the tools: law and security. An analyst who can explain why a dataset is published at street level rather than by address, or why a report should not expose small groups of people, stands out from one who only knows the buttons.

Job titles vary: data analyst, BI analyst, reporting analyst, MI analyst, insight analyst. Before you choose tools, collect ten adverts for jobs you would actually apply for, in your region or remote, and count which tools they name. Let that count decide what you learn second and third.

The skills, in the order to learn them

SQL first

SQL is how you get data out of the systems you will work with. Learn SELECT, WHERE, GROUP BY and JOINs, then common table expressions and window functions. Window functions are where interview questions often go, so practise them on real data:

-- Year-on-year change in average price by district (SQL Server)
WITH yearly AS (
    SELECT district,
           YEAR(transfer_date)        AS sale_year,
           COUNT(*)                   AS sales,
           AVG(CAST(price AS bigint)) AS avg_price
    FROM   dbo.price_paid
    GROUP BY district, YEAR(transfer_date)
)
SELECT district, sale_year, sales, avg_price,
       avg_price - LAG(avg_price) OVER (PARTITION BY district ORDER BY sale_year) AS change_vs_prior_year
FROM   yearly
ORDER BY district, sale_year;

Power BI: model before visuals

The PL-300 study guide (skills measured as of 20 April 2026) splits the exam into four areas:

Spend real time on modelling: fact and dimension tables, a proper date table, relationships and their cross-filter direction. Good DAX is easy on a good model and painful on a bad one. Learn CALCULATE and time intelligence early:

Collisions YoY % =
VAR CurrentPeriod = [Collisions]
VAR PriorPeriod =
    CALCULATE ( [Collisions], SAMEPERIODLASTYEAR ( 'Date'[Date] ) )
RETURN
    DIVIDE ( CurrentPeriod - PriorPeriod, PriorPeriod )

The current skills list also includes Copilot tasks, such as creating a report page and summarising a semantic model, and designing reports for accessibility. If you want public sector work, read our guide to accessible Power BI reports.

Excel and statistics

Excel is still everywhere, so be fluent with tables, PivotTables and lookups. On statistics, aim to explain averages versus medians, spread, seasonality and why a change might be noise. If you are deciding when a team should move beyond spreadsheets, see Excel or Power BI.

Three portfolio projects on UK open data

Project 1: Housing market by area

HM Land Registry's Price Paid Data covers property sales in England and Wales sold for value and lodged for registration, from 1 January 1995, as CSV or text, updated on the 20th working day of each month. Load a few years into SQL Server, clean the property type and date fields, and answer one question well, such as how sales volumes and prices changed by district. Explain why a median tells a fairer story than a mean for house prices. The page gives the exact attribution wording to copy.

Project 2: Road safety star schema

The Department for Transport's road safety open data provides three linked record-level files, collisions, vehicles and casualties, for Great Britain from 1979. The files are coded, with a data guide to decode them, which is ideal practice: turn the codes into dimension tables, relate the three fact tables, build a date table and write measures like the one above. This is the project to talk through in an interview.

Project 3: Local economy brief using the ONS API

The ONS API at api.beta.ons.gov.uk/v1 is open and needs no API key, and its content is under the Open Government Licence v3.0. Pull a dataset with the Power Query web connector or Python, join it to official area codes from the ONS Open Geography Portal, and write a two-page brief for one local authority, as a council analyst would. Writing is the skill most portfolios skip.

For a fourth project, data.police.uk publishes street-level crime, outcomes and stop and search data for England, Wales and Northern Ireland as CSV and through an API. The catalogue at data.gov.uk helps you find more UK public data.

How to present your portfolio

• One page per project: the question, the source with its licence and download date, the method, three findings and the limitations.
• Show the work: put SQL scripts, Power Query steps and a README in a public repository. Saving your report as a Power BI project file makes changes readable in source control.
• Publish safely: Publish to web makes a report and its underlying data visible to anyone without sign-in. That is acceptable for open data like these projects and never for anything from an employer.
• Keep personal data out: never use work data, even anonymised, without written permission.
• Prepare to explain: one DAX measure, one data quality problem you fixed and one finding you rejected because the data did not support it.

On your CV, describe each project in a single line with the tool and the result, and link to the portfolio page.

Mistakes that weaken a portfolio

A common mistake is a dashboard with ten visuals and no question. Start every project with one sentence a manager might ask, and cut any visual that does not help answer it. Another is hiding the cleaning work: employers want to hear how you found duplicate rows, odd dates or codes that did not match the data guide, and what you decided to do about them. The third is reusing the tutorial datasets everyone else uses. UK open data gives you a local, current angle that makes your work easier to remember.

A realistic study plan

Microsoft offers a free PL-300 practice assessment on Microsoft Learn. Associate exams without labs run for 100 minutes, and the certification renews each year through a free online assessment. Check the exam page for the current price in pounds before you book.

How CertTulen Academy can help

If SQL is your starting point, Microsoft SQL Server Fundamentals (CT-MSSQL-B) covers querying, joins and aggregation with hands-on labs. When you are ready for Power BI, Microsoft Power BI Data Analyst (PL-300) is taught live online by a Microsoft Certified Trainer and includes an exam voucher, mock practice tests and lab access, with details on the course page. Contact us for a batch that suits your time zone.