Stastitical

A statistician (the job title here appears to be a misspelling of "Statistician" or a "statistical" role such as statistical officer or statistical analyst) designs ways of collecting numerical data, analyses it and explains what it does and doesn't show. They work for government departments, the NHS and public health bodies, pharmaceutical companies and clinical trials units, market research and polling firms, banks, insurers, regulators and universities. Most of the job is turning a vague question from a non-technical colleague into something that can actually be measured, then producing an answer that is honest about its uncertainty.

Approximate graduate salary

Roughly GBP 27,000-35,000 to start in most sectors, though this varies a lot - public sector and NHS analyst posts sit towards the lower end and are set by published pay bands, while pharmaceutical biostatistics and finance can start higher, and London roles typically carry a weighting. Treat these as broad approximations only.

What you'd actually do

Writing and debugging code to clean, merge and analyse datasets - usually in R, Python, SAS or Stata depending on the sector. Cleaning and checking data typically takes far longer than the analysis itself.
Meeting a policy official, clinician, product manager or client who wants to know something, and working out with them whether the available data can answer it - and pushing back when it can't.
Fitting statistical models (regression, survival analysis, time series, multilevel models and so on) and checking whether the assumptions behind them actually hold for the data in front of you.
Producing charts, tables and short written summaries for people who will never look at your code, and explaining what a confidence interval or p-value means without using those words.
Quality-assuring a colleague's analysis - re-running their code, checking their logic, and documenting what you found. Peer review of analysis is routine in government and clinical work.
Writing or updating documentation: statistical analysis plans, methodology notes, data dictionaries, and version-controlled code repositories (usually Git).
In some roles, designing the data collection itself - sample sizes for a clinical trial, sampling frames for a survey, or randomisation schemes.

How graduates get in

Direct entry to an analyst or statistician post advertised by an employer - the most common route overall. These are often graded as 'assistant statistician', 'statistical officer' or 'graduate analyst'.
The Civil Service Fast Stream has a statistics-focused route, and individual government departments and agencies also recruit assistant statisticians directly outside the Fast Stream. Direct departmental recruitment is more common than the Fast Stream, though both exist.
Pharmaceutical and clinical research entry as a graduate or trainee statistician, either with a drug company or a contract research organisation (a firm that runs trials on behalf of others). A masters is close to expected here.
An MSc in Statistics, Medical Statistics, Biostatistics, Data Science or Applied Statistics - a very common route, and the usual conversion path for people with maths, economics, psychology or another quantitative degree who lack enough statistics content.
A PhD, which is normal for academic and methodological research posts and reasonably common in pharmaceutical statistics, but not needed for most applied roles.
Placement years and summer internships in government analytical services, NHS analytics teams or pharma biostatistics departments - these convert to graduate offers fairly often, but are not the only way in.

What employers ask for

A numerate degree - maths, statistics, economics, physics, engineering, psychology with a strong quantitative component, or similar. The subject matters more here than in most graduate jobs: employers usually want to see specific statistics modules on your transcript, not just general numeracy.
Many public sector statistician posts set a formal requirement that a substantial proportion of your degree was in statistics or that you hold a relevant postgraduate qualification. Read the advert carefully - these rules are applied literally.
A 2:1 is the common baseline, though a 2:2 plus a relevant masters is often accepted, and some employers focus on a technical test instead of the classification.
Coding ability in at least one statistical language - R and Python dominate in government, academia and tech; SAS remains widespread in pharmaceutical trials, though R is increasingly accepted there. SQL is useful almost everywhere.
A masters is genuinely expected in medical statistics and biostatistics, and is helpful but not essential in government, market research and finance.
There is no compulsory licence to practise. The Royal Statistical Society offers Graduate Statistician and Chartered Statistician status; chartership is valued in some sectors and largely ignored in others, so treat it as optional rather than a gate.

Skills that matter

Programming for data analysis (R, Python, SAS or Stata)

Almost nothing is done by hand or in a spreadsheet - your analysis has to be reproducible, so someone else can re-run your code and get the same numbers.

Understanding of study and survey design

Knowing why a sample is biased, or why a comparison group isn't comparable, prevents you producing a confident answer to the wrong question.

Explaining uncertainty to non-technical people

Your output usually feeds a decision made by someone who wants a single number, and your job is to convey how much confidence that number deserves without losing them.

Data wrangling and quality checking

Real datasets have duplicates, missing values, inconsistent coding and errors, and spotting these is often where the real analytical judgement happens.

Statistical writing

Analysis plans, methodology notes and published statistical releases are written documents, and in government and clinical work the write-up is scrutinised as closely as the numbers.

Scepticism about your own results

A surprising finding is far more often a coding mistake or a data artefact than a genuine discovery, and the habit of checking first saves careers.

Where it leads

Assistant statistician / graduate analyst: working on defined pieces of analysis with your code and conclusions reviewed by someone more senior.
Statistician / senior analyst: owning a workstream, deciding the method rather than being told it, and reviewing others' work. This step commonly takes a few years but timelines vary widely by sector and employer.
Senior or principal statistician: leading the statistical design of trials, surveys or official statistics outputs, and acting as the person others come to for methodological judgement.
From here paths diverge sharply. Some move into management of an analytical team; some become specialist methodologists; some move sideways into data science, health economics, epidemiology, actuarial work or quantitative research in finance; some move into policy or product roles where the statistics knowledge is background rather than the job.
In the public sector, progression is tied to formal grades and often requires applying for a posted vacancy rather than being promoted in place, which makes movement between departments common.

What people get wrong

“It's essentially the same job as data science, so the titles are interchangeable.”

They overlap heavily but emphasise different things. Statisticians tend to focus on inference - how confident can we be, given how the data was collected - and often work with small, carefully designed datasets. Data science leans more towards prediction, engineering and large-scale data. Many employers use the titles loosely, so read the job description rather than the title.

“You spend your time doing maths.”

Most of the week is coding, cleaning data, sitting in meetings working out what someone actually wants to know, and writing. The mathematical theory underpins your judgement but you rarely derive anything from scratch.

“The maths is the hard part and the communication is a soft extra.”

The commonest failure mode is a technically correct analysis that a decision-maker misreads or ignores. Being able to say plainly what the numbers do and don't support is what distinguishes a valued statistician from a code-runner.

“You need a PhD.”

A PhD is normal in academic and some pharmaceutical research roles, but a large share of statistician jobs in government, health services, market research and industry are open to graduates with a bachelors or masters.

Where this varies

The sector changes the job substantially. In government and the NHS you work to published statistics standards, your outputs may be scrutinised publicly, and there is a strong culture of peer review and documented methodology. In pharmaceuticals and clinical trials the work is heavily regulated, follows pre-specified analysis plans, and SAS is still widely used. In market research and polling the pace is much faster and the analysis usually simpler, with more emphasis on turning results round quickly for clients. In finance and tech, the title often shades into quantitative analyst or data scientist and coding expectations are higher. Geographically, statistical work is more dispersed than many graduate careers - large government analytical hubs exist outside London, and pharmaceutical and clinical trials work clusters around a few university cities - so this is not a London-only field. Note also that if the role you saw was titled something like 'statistical officer' or 'statistical assistant', it may involve more routine data compilation and production of regular outputs and less methodological design than the picture above.

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.