Before you start
- A list of the work you have done in the last few years, including small projects
- The advert for one real role you want to apply to
A CV for a data role is read quickly and by more than one kind of person: someone who screens many applications against the advert, and later a person who does the work and wants to know whether you can do it too. Both are looking for the same thing in different ways: evidence that something changed because of you.
Start from the role, not from your history
Take the advert for one job you want and mark three things: the tools it names, the problems it describes (reporting, forecasting, pipelines, experiments) and the people you would work with. Your CV should answer those three, in that order of importance, before it says anything else. A CV written for every job is a CV written for none of them; keep one master copy with everything, and cut a version for each kind of role.
Write lines that say what changed
Most CVs describe duties: "Responsible for weekly sales reports. Used SQL and Excel." Duties tell the reader what the job was, not what you did with it. Rewrite each line as three parts:
- The problem or situation. What was slow, wrong, missing or unknown.
- What you did. The action, in your own words, with the tool where it matters.
- What changed. The result, with a number if you have one you can defend.
Compare:
- Before: Created dashboards for the marketing team using Power BI.
- After: Replaced a manual weekly spreadsheet with a Power BI dashboard fed by SQL, cutting the report from four hours of work to a scheduled refresh and letting the team check spend daily.
The second line can be questioned, and that is the point: it gives an interviewer something real to ask about, and you something real to answer.
Numbers you can defend
A figure makes a line believable only if it is true and you can explain how you know it. Good sources are time saved (measured, or counted in hours a week), errors removed (rows fixed, duplicate records found), scale (rows, tables, users of a dashboard) and decisions it informed. If a figure came from someone else, know where. If you do not have a number, describe the scale or the before and after in words. A made-up percentage is worse than none: it is the first thing a careful interviewer will ask about.
The skills section
Group skills so they can be scanned in seconds: languages (SQL, Python), tools (dbt, Power BI, Airflow), and methods (A/B testing, forecasting, data modelling). Put the ones the advert names first. Leave out anything you could not use on your first week. Ratings such as "Python: 4 out of 5 stars" say nothing a reader can trust; a project that uses the skill says much more.
Projects, with links
For each project, one or two lines in the same problem, action, result format, and a link to the code or a short write-up. Make sure the link opens on a phone, the README starts with the answer, and nothing in it contains personal data or a secret. If you keep your projects on GitHub, a short profile README that points to your best three makes the visit worth it.
Make it easy to read
- Plain words. Say "built" and "fixed", not "spearheaded synergies". Short sentences in the active voice are easier to scan.
- One clear layout. Name and contact details at the top, then a two-line summary, experience or projects, skills and education. No photos, tables or columns that confuse the systems that read CVs automatically.
- A file that opens everywhere. Send a PDF unless the advert asks for something else, and name it with your name and the role.
- Proofread aloud. Read it once slowly, out loud. Mistakes in a CV for a role about accuracy are noticed.
A last check
Hand the CV to someone outside your field and give them thirty seconds. Ask what job you want and what you are best at. If they can answer both, the CV is doing its work. If they cannot, rewrite the summary and the first line under each role until they can.
Expected result
A one or two page CV in which every line under a job or project says what changed because of your work, the skills listed match the role you are applying to, and every number is one you can explain in an interview.
Frequently asked questions
Should I include every tool I have ever used?
No. List the tools you could use on your first week without a tutorial, and put the ones the advert names first. A long list of tools you touched once makes the reader doubt the ones you know well, and any of them can come up in an interview.
I have no job in data yet. What goes under experience?
Your projects, any work in another field where you used data to answer a question (a report you automated, a spreadsheet that replaced a manual count), and volunteer work. Write them in the same format: the problem, what you did, what changed.
What if I cannot share real figures from my job?
Use relative figures you are allowed to give (halved, from weekly to daily, cut from four hours to twenty minutes) or describe the scale (a table of 30 million rows, a dashboard used by the sales team). Never invent a number; it will be the first thing you are asked about.
One page or two?
One page is enough for most people with under ten years of experience. Two is fine when every line earns its place. What matters is that the first half page tells the reader who you are and what you are good at.
References
From Grokem: https://www.grokem.com/career/career-paths/data-cv-that-shows-impact
