Get hired for the analysis, not your tool list
HRLens reads your analyst CV the way a hiring manager does, then shows you which bullets prove impact and which just name software.
Why strong analysts get filtered out
Most data analyst CVs read like a software inventory. SQL, Python, Power BI, Excel, a dbt mention, maybe Snowflake. Every other applicant lists the same stack, so the hiring manager has nothing to compare. The tools tell them you can run a query. They do not tell them whether anyone acted on the answer, or whether the business was any different afterwards.
Hiring managers are buying judgement. They want to know which question you were handed, how you scoped it, what you found, and what somebody decided because of it. That is a different sentence from "built dashboards in Tableau". It sounds like "replaced three manual weekly reports with one self-serve dashboard, so the ops team stopped waiting on me for churn numbers". Same work, and now it has a consequence attached.
The second filter is the ATS. Recruiters search by skill and by title, so a CV that hides SQL inside a paragraph, or calls the role "Insights Ninja", can miss the shortlist before a human ever opens it. You need both: keywords a parser can read cleanly, and bullets a person actually wants to finish.
Career changers and bootcamp grads have the opposite problem. They pad. Six course projects with the same public dataset, each written as a tutorial summary, signal that nobody has ever depended on your numbers. Two well-chosen projects with a real question, a messy data source and a defensible conclusion beat six every time.
What HRLens checks on an analyst CV
Impact, bullet by bullet
Flags every line that stops at the tool and suggests the missing outcome, metric or decision.
ATS parsing and keywords
Confirms your stack is machine-readable and matched to the wording in the job ad, not buried in prose.
Project credibility
Tells you which portfolio or bootcamp projects earn their space and which ones read like coursework.
How to write the work you actually do
Dashboards
Name who uses it and what they stopped doing manually. A dashboard nobody opens is not an achievement.
Recurring reporting
Lead with the time you gave back or the errors you removed, then say how you automated it.
Ad hoc analysis
State the question you were asked, the finding, and the call it supported. Three clauses, one line.
Data quality work
Unglamorous and highly valued. Quantify the rows reconciled, the pipeline fixed or the definition you finally settled.
Bootcamp and self-taught projects
Keep two or three. Use messy real data, show the decision you would defend, and skip the tutorial datasets.
Your skills section
Group by SQL, languages, BI tools and warehouses. Drop anything you would not survive a live screen on.
Before and after
The same job, rewritten so it lands
| Your current bullet | What the hiring manager hears | Rewrite that proves the analysis |
|---|---|---|
| Built dashboards in Tableau for the sales team | You can use Tableau | Built a pipeline dashboard the sales team checks daily, replacing a spreadsheet three reps updated by hand |
| Wrote complex SQL queries against the data warehouse | You know joins | Rebuilt the revenue query behind monthly reporting after finding refunds were double counted, correcting the figure finance reported to the board |
| Performed customer churn analysis using Python | You took a course | Segmented churn by onboarding path and showed accounts without a kickoff call left far sooner, which moved kickoff into the standard rollout |
| Responsible for weekly and monthly reports | You are a report factory | Automated the weekly ops pack, cutting a full day of manual work and freeing the team for ad hoc requests |
| Bootcamp capstone: Titanic dataset classification model | Everyone submitted this | Analysed two years of a local charity's donation records, found lapsed donors clustered after one gift, and proposed a follow-up window |
Rewrite your analyst CV around the decisions you changed
Start from your current file or from scratch. HRLens drafts ATS-friendly bullets that keep your stack and add the outcome.
Data analyst CV questions, answered
How do I list SQL, Python and BI tools without sounding generic?
Keep a short skills section grouped by type: query languages, programming, BI and visualisation, warehouses. Then prove each one inside a bullet where it did something. SQL appears in the line about fixing the revenue query, Python in the churn segmentation. The section gets you through keyword search, the bullets get you through the interview.
How do I describe dashboard work as business impact?
Answer three questions in one sentence: who uses it, what it replaced, and what changed as a result. "Built a Power BI dashboard" becomes "gave the ops team a self-serve view of fulfilment delays, ending the daily requests they used to send analysts". Adoption and time saved carry more weight than the number of charts.
Should I put bootcamp projects on my CV?
Yes, if you are changing careers and they show judgement. Keep two or three, use data that came with real mess in it, and write each as a question, a finding and a recommendation. Cut anything built on a famous teaching dataset. Reviewers have read hundreds of those and they prove only that you finished the module.
What if I do not have permission to share exact numbers?
Use relative figures or ranges. "Cut report build time by roughly half" and "analysis covering a seven figure annual spend" say plenty without exposing anything confidential. Order of magnitude works too: thousands of rows, dozens of stakeholders. Never invent precision you cannot defend, because a hiring manager will ask how you calculated it.
How do I get a data analyst CV through an ATS?
Use one column, standard section headings, a normal font and a PDF export with selectable text. Skip logos, icons and tables around your skills. Mirror the exact wording of the job ad, so "SQL" and "Power BI" appear as written. Run a free HRLens check to see how a parser reads your file before you apply.
Do I still need a portfolio link?
It helps, but only if the work is finished and readable. One tidy repository or a single write-up with the question, the data and the conclusion beats a profile of half-run notebooks. Put the link in your header next to your email, and make sure the first thing a reviewer sees explains what you were trying to find out.