What does this job really want?

How it works. You paste in an advert. Our rules read it in your browser and show the full result straight away. If you rate yourself on six skills, you also get a match score. If you choose to add AI feedback, our AI check reads the advert too and adds anything our rules missed; its additions are marked “AI”.

Must-haves and nice-to-haves. We look for the words employers use: “essential”, “required”, “must” and “you’ll need” for must-haves, and “desirable”, “nice to have”, “a bonus”, “ideally”, “a plus” and “advantageous” for nice-to-haves. We also look at the headings things sit under, such as “Requirements” or “Desirable”. Skills that only appear in the duties are shown separately, and anything the advert says you’d be taught isn’t counted against you. The company introduction and the benefits are ignored.

The real seniority. Job titles can mislead. We look at the years of experience asked for (only when the number and the word “experience” are in the same sentence) and at responsibilities that usually come with experience, such as managing a team, mentoring other analysts or setting a data strategy. We flag it clearly when an advert labelled graduate, junior or entry level asks for two or more years or senior-level ownership.

Salary. We read the pay the advert states. We keep hourly and daily rates as they’re written, label “up to” figures and on-target earnings (OTE), and ignore allowances, bonuses and course fees. “Competitive” or “depending on experience” with no figure counts as unclear. An hourly rate below the National Living Wage (£12.71 an hour for workers aged 21 and over, from 1 April 2026, according to GOV.UK) is flagged. To see what data roles usually pay, use the Salary Checker.

Warning signs. We flag unpaid or commission-only work, fees for training or to apply, a job “guaranteed” after a paid course, clauses that make you repay training costs, requests for bank details, ID or money, contact through messaging apps or a personal email address, a very long list of tools for a junior role, and hype words like “rockstar” or “ninja”. These are prompts to look closer, not proof that anything is wrong.

How you match. If you rate yourself from None to Strong on SQL, spreadsheets, Power BI or Tableau, Python, statistics and communicating findings, we compare your ratings with the level the advert asks for. “Good” meets most requirements; words like “strong” or “advanced” ask for “Strong”. Must-haves count three times as much as nice-to-haves, and skills that only appear in the duties count twice as much as nice-to-haves. Anything we can’t check for you, such as a degree class or a tool outside those six, is listed so you can check it yourself.

Likely interview questions. We build them from what the advert asks for, using questions that are common in data interviews, reworded to fit the advert’s sector, plus two that nearly every interview includes. They’re likely, not certain.

The limits. Our rules read words, not intentions, so an unusual layout or unusual wording can confuse them. The AI check can make mistakes too. We can’t tell whether an advert is genuine, or whether an employer will be flexible about what it asks for. Your match score is a guide, not a prediction.