Product Analyst

A product analyst works alongside the people who build and run a digital product - an app, a website, a piece of software - and uses data to tell them what users are actually doing with it and whether changes are working. They pull and interpret usage data, run and read experiments, and turn the results into recommendations for product managers, designers and engineers. The role sits between analytics and product decision-making: you're not just producing numbers, you're expected to say what they mean for what the team should build next.

Starting salary: £28k–£38k
Live graduate jobs: 13 open now
Main way in: Direct application to junior or graduate product analyst roles
They usually ask for: A degree in a numerate or analytical subject is the most common…

Approximate graduate salary

Typically around £28,000–£38,000 to start, though this varies widely. London and venture-backed tech or fintech tend to sit at the upper end or above it; regional employers, charities and the public sector tend to sit lower. Bonuses and share options are common in tech and absent elsewhere.

What you'd actually do

Writing SQL queries against the company's data warehouse to answer questions like 'how many users who started sign-up never finished it, and where did they drop out?'
Building and maintaining dashboards in a tool such as Looker, Tableau or Power BI so product managers can check key numbers without asking you every time
Analysing A/B tests - where two versions of a feature are shown to different groups of users - and working out whether the difference in results is real or just noise
Sitting in on product team ceremonies (sprint planning, stand-ups, roadmap sessions) and being the person who says what the data suggests about a proposed change

Show 3 more

Checking and chasing data quality problems: event tracking that has broken, a release that stopped logging a key action, numbers that suddenly look wrong
Writing up findings as a short deck or written summary with a clear recommendation, then presenting it to a product manager or a wider stakeholder group
Working with engineers to specify what new user actions should be tracked before a feature ships, so the data exists when you need it

How graduates get in

Direct application to junior or graduate product analyst roles - common at tech companies, fintechs, online retailers, gaming companies and digital-first banks and insurers. Titles vary: you may see 'data analyst', 'insight analyst' or 'business analyst' advertised for what is essentially this job, so search broadly.
Graduate schemes in data and analytics at larger employers (banks, retailers, telecoms, media), where you rotate through analytics teams and may land in a product-facing one. Common at big corporates, rarer at startups and scale-ups, which tend to hire directly into a specific team.
Internally, after starting in a related role - customer insight, marketing analytics, operations analysis, or a graduate rotation - and moving across once you know the product and the data. This is a very common route and worth bearing in mind if you can't get in directly.
Via a technology or digital graduate scheme in the public sector or a large non-tech organisation; government departments and the NHS increasingly run digital services with analysts attached, though the title there is often 'performance analyst' or 'data analyst'.

Show 2 more

Conversion routes for graduates from non-quantitative degrees: a data analytics bootcamp, a master's in data science or business analytics, or self-taught SQL plus a portfolio project. Bootcamps are genuinely used by employers here, but they are no guarantee - the portfolio and your ability to talk about analysis in interview matter more than the certificate.
Internships and placement years in analytics, product or consulting are a strong feeder - a year in industry on a sandwich degree is an unusually effective way in.

What employers ask for

A degree in a numerate or analytical subject is the most common background - economics, maths, statistics, computer science, engineering, physics, psychology with a quantitative component. That said, subject matters less than demonstrable comfort with data; graduates from other subjects do get in if they can show SQL and analytical work.
SQL is the near-universal practical requirement. Many employers test it at interview with a live or take-home exercise. If you learn one thing before applying, learn SQL.
Spreadsheet fluency (Excel or Google Sheets) is assumed - pivot tables, lookups, basic modelling. Python or R is often listed as desirable rather than essential at junior level, and matters more in data-science-leaning teams.
A 2:1 is the typical ask on structured graduate schemes; direct-entry roles at smaller companies are much more flexible and often care more about a technical test and a portfolio than your classification.

Show 2 more

No professional qualification is required or standard for this role. Tool-vendor certifications and short courses exist but are optional - nobody is chartered as a product analyst.
Some evidence of having used data to answer a real question: a dissertation with original analysis, a placement, a side project analysing a public dataset, a Kaggle-style exercise. Employers vary a lot in how much they weigh this versus raw academic record.

Skills that matter

SQL

Almost every question you're asked starts with pulling the right rows out of a database yourself, rather than waiting for someone to hand you a clean file.

Understanding of experiments and statistical significance

You will be asked whether an A/B test result is real; knowing about sample size, variance and the risk of reading too much into a small difference is what stops a team shipping a change that does nothing.

Framing vague questions into answerable ones

Stakeholders ask things like 'why is engagement down?' and your first job is to turn that into a specific, measurable question before you touch any data.

Data visualisation and dashboard design

A badly designed chart gets misread and drives a bad decision, so choosing the right view and labelling it clearly is a core output, not a nice-to-have.

Scepticism about your own numbers

Tracking breaks, definitions drift and duplicates creep in - analysts who sanity-check before publishing build credibility, and those who don't lose it quickly.

Explaining technical findings to non-technical people

Your recommendation has to land with a product manager or commercial lead in a few minutes, usually with no statistics vocabulary available to you.

Curiosity about the product and its users

The best insights come from analysts who have actually used the product and have a hunch about where users get stuck, not just from querying tables.

Putting these on a CV? We counted which skills 12,914 live graduate adverts ask for, and what to do with the answer — part of our application advice.

Where it leads

Junior or graduate product analyst: you're mostly answering questions set by others, building dashboards and learning the data model. Typically a year or two, but this varies - at a small company you may be doing senior-level work almost immediately because there's nobody else.
Product analyst, then senior product analyst: you own the analytics for a product area, set your own agenda, decide what's worth investigating and push back on badly framed requests. Senior usually comes somewhere around the three-to-five-year mark, though it depends heavily on company size and growth.
From here the path commonly forks. One route is deeper technical specialism - analytics engineer, data scientist, or lead analyst - which usually means building out Python, experimentation design or data modelling skills.
The other common route is into product management, which product analysts are well placed for because they already understand users, metrics and the development process. Some move into general management of an analytics team instead.
Longer term, senior destinations include head of analytics or analytics manager, group product manager, or consultancy. A number of people also move into growth or commercial strategy roles where the analytical background is the main asset. Timelines here vary enormously and depend far more on company stage than on years served.

What people get wrong

“It's a data science job in all but name, so you need machine learning.”

The overwhelming majority of the work is descriptive and diagnostic - what happened, to whom, and why - done in SQL and a dashboarding tool. Modelling comes up occasionally, and in many teams a separate data science function handles it.

“You hand over the analysis and the product manager decides what to do with it.”

Analysts who only deliver numbers get treated as a query service and find the role frustrating. The expectation, increasingly, is that you come with a recommendation and defend it in the room - influence is a bigger part of the job than most people expect.

“Most of the time is spent analysing.”

A large share of it goes on finding out why the data is wrong, agreeing what a metric actually means with three people who define it differently, and specifying tracking for features that haven't been built yet. Clean, ready-to-analyse data is rare.

“You need to be able to code to a software engineer's standard.”

SQL plus good spreadsheet skills covers a great deal of junior work. Python helps and is worth learning, but it's usually used for analysis scripts rather than production software, and polished engineering practice isn't the bar.

Where this varies

The job differs a lot by employer size and sector. At a startup or scale-up you may be the only analyst, which means building the reporting infrastructure, answering commercial questions and doing the product analysis all at once - broad, fast and unstructured. At a large bank, retailer or established tech company you'll typically be one analyst among many, attached to a single product area, with a defined data platform, governance rules and a clearer career ladder but less scope to set your own direction. Sector also changes what you measure: in e-commerce it's conversion and basket value, in subscription software it's retention and churn, in gaming it's session behaviour and in-game economies, in the public sector it's service completion rates and accessibility. Regulated industries add compliance constraints on what user data you can touch. Finally, job titles are genuinely inconsistent - 'product analyst', 'data analyst', 'insight analyst' and 'performance analyst' can describe the same work, so read the responsibilities rather than the title when applying.

See all 13 live Product Analyst jobsFrom employers' own careers sites, updated daily Tailor your CV for this roleFree with an account, 3 a month

General guidance about the role across the UK market, not about any specific employer. Entry routes and requirements vary — always check the individual job advert.