A data quality analyst makes sure an organisation's data can be trusted before anyone uses it. They find missing, duplicated, wrong and inconsistent records, work out why they happen, fix them, and set up checks so the problems don't come back.
What a data quality analyst does
- Profiles data: how many rows, how many blanks, which values appear, what looks odd.
- Writes rules, like "every order must have a date" or "an email must contain @".
- Runs checks against those rules, often every day, and reports what fails.
- Fixes problems, or works with the team that owns the data to fix them at the source.
- Documents what each field means, so everyone uses the data the same way.
The six things they check
- Completeness. Are required fields filled in? No blank customer IDs.
- Uniqueness. Is each record there once? No customer entered twice.
- Validity. Do values follow the rules? A date is a real date, a quantity isn't negative.
- Accuracy. Do values match reality? The price in the system is the price on the shelf.
- Consistency. Does the same thing look the same everywhere? "Lagos", "lagos" and "LAGOS " are one city.
- Timeliness. Is the data up to date enough for what it's used for?
The SQL behind the checks
Most data quality checks are short SQL queries. A few you'll write again and again:
- Count the blanks in each column. Counting NULLs per column →
- Find duplicates. Detecting duplicates with GROUP BY and HAVING →
- Profile a column. Data profiling with MIN, MAX and COUNT DISTINCT →
- Flag values outside the expected range. Range checks →
- Check nothing was lost after cleaning. Before-and-after validation →
Skills you need
- SQL. The main tool: almost every check is a query. Free SQL lessons →
- Excel or Google Sheets. For quick checks and sharing results. Free spreadsheet lessons →
- Attention to detail. Noticing that one store's sales are always exactly zero on Sundays.
- Curiosity about causes. Fixing a record is good; finding why it broke is better.
- Clear communication. Explaining a data problem to the team that owns it, without blame.
Where the jobs are
Anywhere data matters a lot or is regulated: banks and insurers, healthcare, government, telecoms, retail and large online businesses. The role is sometimes called data quality analyst, data steward or data governance analyst, and it's a common first step into data teams.
How to get started
- Learn SQL basics: SELECT, WHERE, GROUP BY.
- Then learn data cleaning: blanks, duplicates, text clean-up, dates and range checks.
- Practise on messy data. Clean data teaches you nothing about this job.
- Build one project: take a messy public dataset, list its problems, fix them, and show the before and after.
Read The Data Cleaning Mindset and A Framework for Missing Data for the thinking behind the work.
Questions people ask
Is data quality analyst an entry-level job?
Often, yes. It's a common way into data teams, because it builds strong SQL and a deep understanding of how the organisation's data works.
What's the difference from a data analyst?
A data analyst uses data to answer business questions. A data quality analyst makes sure that data is right before anyone does. The skills overlap a lot, and many people move between the two.
Do I need Python?
Not to start. SQL covers most checks. Python helps later, for automating checks across many tables.
Try it yourself
Start with Level 1: TRIM, UPPER and LOWER, in your browser.
The first 3 levels are free · no sign-up needed