Each one is the summary from a complete resume, with a note on the specific choice that makes it work. Take the structure — keep your own numbers.
Entry level / fresher
entry
Data analyst with a statistics degree and hands-on SQL, Excel and Tableau experience from a summer internship and four academic projects. Built reporting that replaced manual spreadsheet work for a 12-person team. Comfortable owning a question end to end: pull the data, check it, visualise it, and explain what it means to people who do not write queries.
Why it works. The middle sentence is the whole summary: replacing manual spreadsheet work for 12 people is a real outcome with a real beneficiary, and it happened during an internship - proof that you do not need a career to have one. "Check it" inside the end-to-end list is a small, deliberate word. Data quality is the thing hiring managers most doubt about a graduate, and naming it costs two characters.
Self-taught, no analyst job yet
no-experience
Self-taught data analyst with three certifications and five end-to-end analysis projects in SQL, Python and Power BI. Currently automating inventory and sales reporting in a retail operations role, where I cut a weekly stock report from 5 hours to 15 minutes. Looking for a first dedicated analyst position where the work is the analysis rather than a side effect of it.
Why it works. 5 hours to 15 minutes is doing all the work here, and note where it happened: a retail operations job, not a portfolio project. Analysis performed inside a real business under real constraints outranks a Kaggle notebook every time, even when the job title says something else. The closing line is honest about the gap without asking for sympathy.
Career change into analytics
career-change
Data analyst who came from seven years of secondary-school maths teaching, now two years into analytics work. I own the reporting for a 40-person operations team: SQL models, a Power BI layer, and a monthly review that the department actually uses to make decisions. The teaching background is the reason non-technical stakeholders keep asking me to present the numbers.
Why it works. Teaching is not treated as the thing to get past. It is the closing argument: explaining quantitative material to people who did not choose to hear it is the single most transferable skill from that career, and analytics teams are chronically short of it. "A monthly review the department actually uses" quietly separates work that is consumed from work that is merely produced.
Mid-level, 4 years
mid
Data analyst with 4 years owning subscription and revenue reporting at a 300-person SaaS company. I run the models behind the weekly revenue review, and my cohort work is what moved the team off a headline churn number that had been hiding a 9-point difference between two plan tiers. Strong SQL, dbt and Looker; comfortable being the person who says the number is wrong.
Why it works. The churn sentence is a complete analytical story in one line: a metric everyone trusted, a segmentation that broke it, and a specific size to the error. That is what a hiring manager is trying to find out and almost never sees. The final clause is a character claim rather than a skill claim, and at mid-level it is the more useful of the two.
Senior, 7 years
senior
Senior data analyst with 7 years in marketplace and subscription businesses, currently leading analysis for a 4-person analytics team. I set the metric definitions the company reports on, led the experimentation programme from ad-hoc tests to a reviewed pipeline of 60+ a year, and am usually the person brought in when two teams disagree about what a number means.
Why it works. Every clause is about influence rather than output. Owning the metric definitions the company reports on is authority; taking experimentation from ad-hoc to 60+ reviewed tests a year is building a system rather than answering tickets; being the tiebreaker between teams is organisational standing. Note there is no tool list at all - by this level the stack is assumed and the space is better spent.
Specialist (marketing analytics)
specialisation
Marketing data analyst with 5 years measuring paid, lifecycle and organic performance across a EUR 6M annual budget. I built the incrementality testing programme that replaced last-click attribution as the basis for budget decisions, and I am comfortable telling a channel owner their channel is not working. Strong SQL, GA4, and marketing-mix modelling in Python.
Why it works. A specialist summary should be unmistakably domain-specific by the second line, and "replaced last-click attribution" could not appear on any other kind of analyst's resume. It also names a political reality of the job - telling a channel owner their channel is not working - which signals you have survived that conversation rather than only run the model.