The nine areas it reads for
A data analyst CV is read for proof of everyday analyst work, not for a list of tools. The checker looks for these areas:
- SQL fundamentalsAlmost every data analyst job asks for SQL, and a reader wants to see what you used it for, not just the word.
- Joins and data qualityJoining tables is where totals silently double. A line showing you check duplicates, NULLs or totals tells a reader your numbers can be trusted.
- Window functions and CTEsWindow functions and multi-step queries separate everyday SQL from analyst-level SQL: rankings, running totals, the previous value.
- Metrics and definitionsAnalysts are trusted for definitions: what exactly is counted, for whom and over which period. A line where you defined or owned a metric is strong evidence.
- Experiments and statisticsProduct and marketing teams decide with A/B tests. Reading one correctly — sample size, significance, a broken split — is a frequent interview topic.
- Product analytics: cohorts, retention, funnelsCohorts, retention and funnels are the daily work of product and growth analysts, and a common case-interview question.
- BI and dashboardsMost analyst roles ship results as dashboards. Name the tool and what the dashboard changed for the people who used it.
- Python, R and spreadsheetsMany roles list Python or R next to SQL, and spreadsheets are still everywhere. Show what you automated or analysed with them.
- Recommendations and stakeholdersA number becomes useful when someone acts on it. A line with your recommendation and the decision it informed shows judgement, not just tool skills.
Mentioned is not the same as shown
“SQL” in a skills list tells a reader you have heard of it. A line with what you did and what changed tells them you can do it:
The second line is an illustration of the shape, not text to copy: write your own result, and only one that really happened. The guide A portfolio project in the AI era explains why the trace of your judgement — the question, the definition, the checks, the decision — now matters more than polish.
Why the result is a plan, not a score
Adding keywords you can’t explain doesn’t survive the first follow-up question. So every gap comes with a short lesson, a 30-second check from Atterna’s practice and one place to practise it, free first: Start days 1–2 for SQL from zero, case 1 for a real request in about 17 minutes, and the free guides. After that, a line on your CV describes work you actually did.
FAQ
Is my CV sent anywhere?
No. The check runs in your browser with fixed rules: the text is not uploaded, not stored and not included in analytics. If you allow analytics, the page counts that a check ran and which 30-second exercises you open — never the text, its length or your map. Close the tab and the text is gone.
Does it give an ATS score?
No. A percentage would be fake precision. The checker reports three plain states for each of nine areas — shown with evidence, mentioned only, not found — and quotes the lines it read, so you can check every verdict yourself.
What counts as evidence?
A line that names the skill and carries something a reader can check: a number (a percentage, money, time, a count), an outcome (reduced, increased, automated, launched), a link to a repository or portfolio, a named project, or an action on a concrete artefact, such as a dashboard built for a specific team. A skill in a skills list counts as mentioned.
Should I add the missing keywords to my CV?
Only for skills you can back up. Interviewers ask follow-up questions, and a keyword you can’t explain does more harm than a gap. The plan under the result is about learning a skill and then showing proof of it.
Can I upload a PDF or Word file?
Paste the text instead: open the file, select all, copy and paste. A plain .txt file can be uploaded and is read on your device. PDF parsing is left out on purpose, so nothing has to leave your browser.
Which roles is it for?
Junior and mid-level data analysts, including product, marketing and BI analyst roles. It reads English and Russian CVs. It is not tuned for data science or data engineering roles.
Will a good result get me interviews?
Nobody can promise that. A text check is a starting point, not a hiring verdict: it reads words and patterns, it can miss a skill phrased unusually, and it can’t judge how good the work was.