Your First Analysis Is the Role You Are Targeting
By the end of this you will have picked one role and know the next five things to read. That is the whole point. People spend six months learning tools in no particular order because they never decided what job they were learning them for.
Deciding is itself an analysis. You have a question, "which of these jobs should I aim at", and you have data: what the work is, what it pays, how many of these jobs exist, and how hard it is to get the first one. Work the question the way you would work any other.
How to read the numbers
Before the numbers, a question. Why would a median be more useful to you than an average? Have a go before you read on.
Because averages get dragged upward by a few very large salaries. The median is the middle. Half the people earn less, half earn more. It is a better picture of what an ordinary person in that job is actually paid.
Four things to keep in mind while you read:
- These are all-experience figures. They include people who have done the job for twenty years. Your first salary will sit well below the median. The median tells you where the road goes, not where it starts.
- The job title in the data is not the job title in the advert. Government statistics use fixed categories that lag behind what companies call things. I have named the category next to each number so you can check my working.
- Countries are not directly comparable. Different taxes, different costs, different survey methods. Compare a role to other roles inside the same country, not across the border.
- India is a different quality of evidence. The US, Canada and the UK all publish official salary statistics by occupation. India does not. There is no government table to point at, so India numbers here come from job boards and recruiter surveys where people report their own pay. Treat them as a rough signal.
Sources are listed at the end. US figures are 2025 medians from the Bureau of Labor Statistics. Canada is hourly, from Job Bank, based on the Labour Force Survey. The UK is gross annual pay from the Office for National Statistics, 2025.
The nine roles
Before you read them, guess which one you think has the most jobs in it. Hold that guess and see if it survives.
1. Data analyst, general
The work. You are handed a question in words and you answer it with numbers. Pull the data, check it is trustworthy, find the pattern, and explain it to somebody who will not read your code.
A real day. A request lands: "why did returns spike last month". You pull the orders, find that one warehouse accounts for most of it, and discover that half of those rows have a null reason code. The rest of the morning is deciding what to do about the nulls. The afternoon is one chart and six sentences.
Skills that get you hired. Excel first, because every job assumes it. Then SQL for filtering, grouping and joining. Then one chart tool. Statistics after those, not before.
Read that US number carefully. The Bureau of Labor Statistics has no separate category for data analysts. It counts them inside "data scientists", which is full of senior and specialist pay. The UK figure of £38,107 comes from a category that really is called data analyst, so it is the more honest picture of the ordinary job.
Demand. Strong everywhere. The US category is projected to grow 7% or more through 2034. The UK counts 68,000 of these jobs.
Difficulty of entry. Moderate, and crowded. There are many applicants per opening because this is the role everyone retrains into. A finished piece of work beats a certificate here.
Grows into. Senior analyst, then analytics engineer, data scientist, or analytics manager.
Read in this order: Excel, SQL, Chart Literacy, Statistics, Python.
2. Business analyst
The work. You sit between the people with a problem and the people who build things. You work out what is actually needed, write it down clearly, and make sure what gets built matches it.
A real day. Two hours of meetings where three people describe the same process differently. Then you draw the process, send it round, and watch them argue about your drawing instead of about each other. That argument is the deliverable.
Skills that get you hired. Writing, first and most. Excel. Enough SQL to check a claim yourself. Process mapping. Comfort asking a question that makes you look slow.
Demand. Very high. The US counts over a million management analysts, the largest category in this whole guide, growing 7% or more.
Difficulty of entry. Easier than it looks, and the best route in from a non-technical job. Industry experience counts as a qualification here, which is rare.
Grows into. Senior BA, product owner, programme manager, consultant.
Read in this order: Excel, SQL, Thinking Like an Analyst, Chart Literacy, the migration guides for project work.
3. Financial analyst
The work. You explain money. What did we spend, what will we spend, what happens to the year if this deal closes late.
A real day. Rebuilding a forecast because sales changed the timing on one large contract. Then a variance report explaining why a department is over budget, which is really an exercise in asking four people what they bought.
Skills that get you hired. Excel to a genuinely high level, well past what other analyst roles need. Forecasting. Enough accounting vocabulary to be trusted. SQL is a bonus, not a gate.
Demand. Steady rather than booming. The US projects 5% to 6% growth, lower than the others here, from a base of 368,500 jobs.
Difficulty of entry. Harder. This is the most credential-minded role in the list. A finance degree or an accounting qualification opens doors that skill alone will not.
Grows into. Senior financial analyst, finance business partner, FP&A manager, controller.
Read in this order: Excel, Forecasting, Statistics, SQL, Chart Literacy.
4. Reporting and BI analyst
The work. You build the thing people look at every week. Dashboards, scheduled reports, the numbers the business runs on.
A real day. A dashboard disagrees with a finance report by a small amount and everyone has stopped trusting both. You spend the day finding the one filter that explains it, then write down the definition so the argument cannot happen again.
Skills that get you hired. One BI tool, deeply, either Power BI or Tableau. SQL, because dashboards are queries wearing a coat. Chart choice. Patience with definitions.
Demand. High and very visible. Nearly every job advert for this role names its tool, which makes it the easiest role to target deliberately.
Difficulty of entry. The most reachable technical route. Tool skill is testable, so a strong portfolio dashboard does a lot of work for you.
Grows into. Analytics engineer, BI lead, data platform work.
Read in this order: Excel, SQL, Power BI or Tableau, Chart Literacy, Statistics.
5. Operations analyst
The work. You make a process cheaper, faster or more reliable. Staffing, routing, stock, queues, capacity.
A real day. Working out why Thursday deliveries run late. It turns out two rounds overlap at one depot, and the fix is a scheduling change nobody needs a model to understand once you show it.
Skills that get you hired. Excel. SQL. Statistics, more than most analyst roles need. A real interest in how the physical work happens.
Demand. Growing 7% or more in the US, but a smaller pool at 112,100 jobs. Fewer openings, and fewer people chasing them.
Difficulty of entry. Moderate. Easiest if you already work in the operation. Warehouse, clinic, call centre and logistics experience is a genuine advantage.
Grows into. Operations manager, supply chain analytics, process improvement lead.
Read in this order: Excel, SQL, Statistics, Forecasting, Chart Literacy.
6. Marketing analyst
The work. You measure whether the money spent getting attention did anything.
A real day. A campaign report where the tracking broke for two days. The interesting work is deciding, and writing down, how to treat the gap so the number is not quietly wrong.
Skills that get you hired. Excel. SQL. A web analytics tool. Chart choice, because this audience reads pictures rather than tables. Statistics, especially for testing.
Demand. Very high volume. The US counts 941,700 jobs, second only to management analysts, growing 7% or more.
Difficulty of entry. The easiest door in this list. Lots of roles, lower barriers, and small companies hire people who can simply show a clear report. It also pays the least at the start.
Grows into. Marketing manager, growth analyst, product analyst, customer analytics.
Read in this order: Excel, Chart Literacy, SQL, Statistics, Forecasting.
7. Healthcare analyst
The work. Analysis inside a hospital, insurer, clinic group or health system. Waiting times, capacity, quality, cost, safety.
A real day. Readmission rates by ward, and a long conversation about whether two wards record discharges the same way. They do not. That finding matters more than the chart.
Skills that get you hired. SQL. Excel. Careful handling of private data. Patience with coding systems and clinical vocabulary.
Demand. Growing 7% or more, and unusually stable, because healthcare demand does not follow the economy.
Difficulty of entry. Moderate. Clinical or administrative experience inside health is worth more than any certificate. Note that a records or coding job is a different job from an analyst job, even though the two get advertised in similar language.
Grows into. Senior healthcare analyst, clinical informatics, quality and safety lead.
Read in this order: SQL, Excel, Statistics, Chart Literacy, Documenting Data Limitations.
8. Data migration analyst
The work. A company is moving from one system to another and their records have to arrive intact. You profile the old data, map it to the new fields, clean it, rehearse the move, and prove afterwards that nothing was lost.
A real day. Reconciling counts after a test load. You are short 412 records. Six hours later it is a date format, and you have written the check that will catch it next time.
Skills that get you hired. SQL for counting, grouping and joining. Excel. Meticulous record keeping. Client conversation skills, because this work happens on someone else's deadline.
Demand. Read this one carefully. The US projects database administrator jobs to decline through 2034. Migration itself is not declining, but it is project work inside software vendors and consultancies, not a database job. Target the vendors and the consultancies.
Difficulty of entry. Moderate, and the SQL needed is genuinely basic. Counting rows, finding duplicates, finding orphans. The hard part is the client conversation, not the query.
Grows into. Implementation consultant, solutions architect, professional services lead.
Read in this order: the migration series, SQL, Excel, then Documenting Data Limitations.
9. Product analyst
The work. You measure what people do inside a product, and help decide what gets built next.
A real day. A new sign-up screen shipped last week. You check whether more people finish sign-up, and find that they do, but fewer of them come back on day seven. Now the team has a real decision to make.
Skills that get you hired. SQL, used constantly. Statistics for experiments. Chart choice. Enough product sense to know which question is worth asking.
Demand. Concentrated in software firms, so it follows technology hiring closely, up and down.
Difficulty of entry. Hardest of the nine as a first job. Most product analysts arrive from another analyst role or from inside the company.
Grows into. Senior product analyst, product manager, growth lead.
Read in this order: SQL, Statistics, Chart Literacy, Excel, Python.
The nine side by side
Which column would you sort this table by, if you could only pick one? Answer, then read on.
Most people should sort by difficulty of entry, not by pay. The first job is the hard one. Pay follows once you are inside, and every one of these roles pays more after two years than it does on day one.
| Role | Entry | Volume | US median | UK median |
|---|---|---|---|---|
| Marketing analyst | Easiest | Very high | $78,760 | £30,479 |
| Business analyst | Easy | Highest | $101,860 | £51,729 |
| Reporting and BI | Easy | High | $105,850 | £59,593 |
| Data migration | Moderate | Niche | $104,620 | £36,015 |
| Healthcare analyst | Moderate | High | varies widely | £55,879 |
| Operations analyst | Moderate | Lower | $88,940 | £39,941 |
| Data analyst, general | Moderate, crowded | High | $120,230 | £38,107 |
| Financial analyst | Harder | Steady | $102,740 | £47,776 |
| Product analyst | Hardest first | Concentrated | no category | no category |
What to read, in order, once you have picked
Every role above ends with five things in order. Here is why the order barely changes.
Excel comes first in most of them because every job assumes it and no job teaches it. SQL comes second because it is how you get the data without asking someone. A chart tool comes third because work nobody understands does not count as work. Statistics comes fourth because it stops you being confidently wrong. Python comes last because it is the one you can do the job without.
The kits follow that order already. Start with Excel, then SQL, then the chart tool your target adverts name. Type the queries rather than reading them, which is what SQL Drill is for, and build the chart yourself in Viz Drill.
Alongside the kits there are more than forty short guides. Read them when their subject is a problem you actually have, not in order. The guides index lists all of them, including the sixteen on data migration if that is the road you picked.
The kits run in the order an analyst actually learns, and they run in the browser with the data in front of you.
Open the Analyst Prep Kit →The one habit
Write your target role at the top of whatever you are learning in. Every time you are unsure what to study next, the answer is whatever that role's adverts keep asking for.
Which of the nine did you pick, and what did the last five job adverts for it actually ask for?
References
- US Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2025 medians, and Employment Projections 2024–2034, as published on O*NET OnLine for codes 15-2051, 13-1111, 13-2051, 13-1161, 15-2031, 15-1211, 15-1242 and 29-2072.
- Employment and Social Development Canada, Job Bank wage data, drawn from the Statistics Canada Labour Force Survey, 2023–2024 reference period, updated 19 November 2025. NOC codes 21223, 21221, 11101 and 21211.
- Office for National Statistics, Annual Survey of Hours and Earnings 2025 provisional, Table 14.7a, gross annual pay by four-digit SOC 2020, released 23 October 2025.
- Info Edge, Naukri JobSpeak monthly hiring index, used for Indian demand rather than pay. India has no official occupational wage statistics, so Indian pay figures in this guide come from self-reported job board and recruiter data and are marked as such.