Data analyst job description
Most data analyst job descriptions list tools. SQL, Python, Tableau, Excel. Every credible candidate has all four on their resume, which means the description filters nobody and the screening problem arrives intact at your calendar. Below is a description you can post, and then the part that decides the hire: what this role has to settle, where it goes wrong, and what to ask for evidence of.
The job description
We are looking for a data analyst who can take an ambiguous business question, decide what would actually answer it, and get that answer in front of the people making the call. You will work directly with the teams whose decisions your analysis changes.
What you will do
- Partner with business owners to turn open questions into questions that data can settle
- Build and maintain the reporting the team actually uses, and retire the reporting it does not
- Investigate anomalies to the point of root cause, not to the point of a plausible story
- Define and document metrics so two teams asking the same question get the same number
- Present findings to non-technical stakeholders and defend the method when challenged
What you will need
- Fluency in SQL against a real production warehouse, including the joins nobody enjoys
- A scripting language for analysis and cleanup, typically Python or R
- Experience with a BI tool, and an opinion about what makes a dashboard get opened twice
- The ability to explain a caveat without burying the finding
Signals we weight heavily
- You have told a stakeholder that their hypothesis was not supported, and it stuck
- You have killed a report
- You can describe a time your first answer was wrong and how you found out
What a data analyst is actually accountable for
The tools are the floor. The job is judgement about which question is worth answering, and the confidence to say when the data does not support the request being made.
- Turning a vague business question into one that data can actually settle, and saying so when it cannot
- Choosing what not to measure, so the dashboard has ten numbers rather than eighty
- Getting a number in front of a decision before the decision is made, not after
- Being trusted enough that a stakeholder changes their mind because of the analysis
- Knowing the shape of the underlying data well enough to catch when a result is an artefact of collection
Four ways a data analyst hire looks right and is not
Every one of these passes a resume screen. Ladders eye-tracking study timed the average initial scan at just 7.4 seconds, which is long enough to check the four signals below and not long enough to check anything that would contradict them.
Looks right
A portfolio of polished dashboards
Almost every analyst portfolio is built on public datasets where the question was already chosen and the data was already clean. It demonstrates tool fluency and tells you nothing about whether they can decide what is worth measuring inside your business.
Looks right
Strong performance on a SQL test
SQL screens test recall of syntax under time pressure. They are easy to pass with practice and easy to fail while distracted, and they correlate with neither of the two things that actually break analyst hires: framing the question, and holding a position when a senior stakeholder disagrees.
Looks right
Years of experience at a company with a serious data function
A large data team often means the questions arrived pre-framed from an analytics lead and the pipelines were somebody else's job. That is a very different role from being the first analyst in a company of 80, where nobody will tell you what to look at.
Looks right
Certifications in the exact BI tool you use
BI tools are learned in weeks. Tool-matching is the cheapest filter available and the one most likely to screen out the person who would have been right, because they happened to spend three years in the other tool.
What to ask for evidence of instead
Four questions, and what the answer actually tells you. Take these into your own process whether or not you ever talk to us.
- Tell me about an analysis that changed a decision. What was the decision going to be, and what did it become?
- What it tells you: Analysts who have only ever produced reporting cannot answer the second half. If the decision was going to be the same either way, the work was decoration.
- Tell me about a time the data did not support what someone senior wanted it to say.
- What it tells you: This is the single most predictive question for the role and the one most rarely asked. You are listening for whether they raised it, how, and what happened next. "It never came up" is an answer.
- What have you chosen not to measure, and why?
- What it tells you: Distinguishes people who add metrics from people who own a metric set. Anyone who has run reporting for two years has killed something, and can tell you what broke when they did.
- Walk me through a result you got that turned out to be an artefact of how the data was collected.
- What it tells you: Tests whether they have ever gone below the table they query. An analyst who has never been burned by an instrumentation bug has probably not been close enough to the source.
How Continuity1 runs this funnel
The screening above is the job. These are the numbers it produces when a function owns it end to end, set against the published benchmarks for the same market.
- 1 in 3
- Shortlisted candidates you meet who become the hire
- Aligned engagements run nearer 1 in 2, distant ones nearer 1 in 10. The market takes about 180 applicants to make one hire, and that sifting lands on your team rather than ours.
- Continuity1 tracked engagements
- ~3
- Interviews your team sits in, per hire
- Ashby puts technical roles at 17.6 interviews per hire across the whole process, up 52% since 2021. The rest of that load sits with the function, not with you.
- Ashby talent-trends report
- 1 in 9
- Accepted offers that ghost before joining
- Indian employers report nearly 4 in 10 offers dropped. We lose 1 in 9.
- nasscom community
- 95%
- Offers that close inside your stated band
- 20 of the last 21. A flat fee earns nothing from an inflated offer; a percentage of CTC earns more.
- Continuity1 tracked engagements
Every brief becomes a success profile before sourcing starts, calibrated with the people who will manage the role. That calibration is the step most hiring skips, and it is why a shortlist either matches the job or matches the job advert.
You review a scored shortlist and make the calls. The filtering never lands on your calendar.
Questions teams ask
Should the job description ask for Python or is SQL enough?
Ask for SQL as a requirement and a scripting language as an expectation. Making Python a hard requirement screens out strong analysts from BI-heavy backgrounds, and the ones worth hiring pick it up inside a quarter. Making it invisible attracts people who will hand you a spreadsheet when they need a loop.
What is a realistic number of applications for a data analyst role?
It is one of the highest-volume roles you will post, and it attracts a very wide range of preparation. That is the reason the description filters so poorly: everyone has the keywords. The filtering has to happen after the application, which is where most hiring teams run out of hours.
How many interview rounds does a data analyst need?
Fewer than most teams run, if the screening was real. The published benchmark for technical roles is far higher than most teams assume, and rising; the comparison against our own median is in the funnel section above, with its source. The reason the number can come down is that a scored shortlist moves the interview from verification to judgement.
Should we use a take-home assignment?
Only if you would be happy to receive it back in ninety minutes and you are willing to pay for anything longer. Long take-homes select for candidates with free evenings rather than candidates with judgement, and the strongest people in the market decline them.
Related