Skip to content
← Work from Nepal

Work from Nepal for the World

Data Analyst

Turn a company’s messy numbers into clear answers: what sells, what’s failing, and what to do next.

Pays abroad
No US figure for analysts alone; data scientists $120,230
To learn
Basics in 6 months; a job takes longer
Hours
Reports on a schedule; some meetings
Will AI close it?
Growing

The work, day to day

A data analyst answers questions with numbers: which products sell, where customers give up, whether a change worked. A day is getting data out of a database with SQL, cleaning it (missing values, duplicates, dates in three formats), finding the pattern, and showing it in a chart or dashboard a manager can read in a minute, then saying what it means and what to do. Much of the job is careful checking, because a wrong number in a report is worse than none. Data entry, copying figures into forms, is a different job, and a shrinking one.

What it pays

The US Bureau of Labor Statistics has no figure for data analysts alone. Data scientists, the more technical job above them, earned a median $120,230 in May 2025, and their number is expected to grow 35% by 2035. For Nepal there is no reliable figure for remote analysts. IIDS’s study of 2022 found that IT freelancers earned an average of US$4,723 a year, and that salaries for IT workers in Nepal averaged US$2,810 at entry level, US$6,395 at mid level and US$15,322 at senior level. It also counted 7,212 freelancers on global platforms doing clerical and data entry, the work this job grows out of.

Will AI close it?

Growing: AI makes more of this work.

The World Economic Forum’s 2025 survey of employers lists data analysts and scientists among the fastest-growing jobs to 2030, and the US expects 35% more data scientists by 2035. The job that is closing is next to it: data entry clerks come first among the fastest-declining. Analysts stay needed because AI answers are only as good as the data: Microsoft warns that without well-prepared data its Copilot in Power BI “can misinterpret the data and return generic or inaccurate results”. Knowing the business and catching the wrong answer keep you in the job.

How AI helps you do it from Nepal

An AI assistant can draft the SQL or Python for a question, suggest a chart, explain an error message and turn your findings into a clear English summary for a manager abroad. Microsoft’s Copilot in Power BI writes DAX queries and builds report pages, though it needs a paid Microsoft plan, so you’ll meet it at work rather than while learning. Be careful: check every number an AI gives you against the data, never upload a client’s data to an AI tool the client hasn’t approved, and be able to explain each step yourself, because a take-home test checks exactly that.

Data Analyst · 1 of 2 · science-playground.com/learn/remote-work/data-analyst

The skills, in order

  1. Spreadsheets: formulas, pivot tables and charts, in Excel or Google Sheets.
  2. SQL: getting, joining and summing data from a database.
  3. Basic statistics: averages, spread, samples, and why going together isn’t causing.
  4. A dashboard tool: Power BI or Tableau.
  5. Python with pandas, for data too big or messy for a spreadsheet.
  6. Telling the story: one page a busy manager will read.

Where to learn them from Nepal

Show your work

Make three or four analyses from public data and publish them on Kaggle or GitHub. Pick questions people care about: Nepal’s remittances month by month, census figures by district, a shop’s sales if the owner will share them. For each, show the question, the cleaned data, the SQL or code, the chart, and the answer in three sentences. Employers often set a take-home task: a messy dataset and a day to find something useful in it. Practise exactly that.

Finding the work

Most analyst jobs sit inside a company’s team, so the employer-of-record route matters most: Deel employs people in Nepal on companies’ behalf, and Remote can pay contractors here. Job boards’ worldwide filters (Remote OK has one) find companies that hire anywhere; skip ads that say US or EU only. Small firms abroad also want a monthly report or a dashboard kept up to date, a good first contract. A data team in a Nepali company, bank or NGO teaches the trade first.

Your first 90 days

  1. Month 1: spreadsheets up to pivot tables and charts, then start SQL, an hour a day.
  2. Month 2: one full analysis of Nepali public data, from question to cleaned data to chart to a three-sentence answer. Publish it.
  3. Month 3: learn Power BI or Tableau, rebuild your analysis as a dashboard, and make one for a local business or school, for a real reference.

Getting paid into Nepal, the 5% tax, contracts and the scams aimed at beginners: Getting Paid in Nepal (science-playground.com/learn/remote-work/getting-paid).

A business that can grow from it: Bookkeeping for Foreign Small Firms, AI Data Labelling Studio (Business Ideas for Nepal).

More than the printed sheet

A day in the work

Data work runs on deadlines more than on meetings. A request comes in writing, in a ticket or a message from a manager: why did sign-ups fall in March? You get the data out with SQL, clean it, check it, and send back a chart, a dashboard or a short note with what it means. Microsoft’s description of a Power BI analyst fits most of these jobs: you work closely with the business people to find out what they need, and with data engineers to get the data, then you prepare it, model it, and show it so that others can find answers for themselves.

The time difference can help. A report a manager in London wants at 9 am is due at 1:45 pm in Nepal in the British summer (2:45 pm in winter), so you can work through your morning and send it before their day begins. For a client in New York the report is ready while they sleep: their 9 am is 6:45 pm or 7:45 pm in Nepal. The meetings are fewer than a developer’s, but harder: a weekly call where you explain what the numbers say to people who didn’t see the work, and answer “but why?” on the spot.

Much of the week is checking. Of the courses in Google’s certificate, a whole one is called “Process Data from Dirty to Clean”: finding the duplicates, the blanks, and the dates written three ways before anything is counted. Note each step, so you can show how you got any number.

The pay, worked through

Small firms abroad often want one thing kept up to date, a monthly dashboard or report. There is no reliable figure for what this pays, so take US$400 a month only to show the arithmetic, with the client paying through Payoneer.

1. The client pays by US bank debit (ACH). Payoneer takes 1%, US$4, and US$396 reaches your Payoneer account. By card, it would take up to 3.99% plus US$0.49.

2. You withdraw to your bank in Nepal. Payoneer’s price list doesn’t name Nepal among the countries it pays out to in local currency, so it changes the money, for 1.2–4%; check the fee in your account. At 1.2% it takes US$4.75, leaving US$391.25; at 4%, US$15.84, leaving US$380.16.

3. In rupees, at about Nepal Rastra Bank’s buying rate on 12 October 2026, Rs 154.48 to the dollar (the rate used is Payoneer’s own): Rs 60,440 or Rs 58,727.

4. The bank keeps 5% advance tax as the money comes in (Republica, 2026): you receive Rs 57,418 or Rs 55,791.

5. One fee more: Payoneer charges US$29.95 a year to an account that receives less than US$6,000 in any 12 months. US$400 a month is US$4,800 a year, so it applies: about US$2.50, or Rs 386, a month.

So of every US$400, between about 3% and 6% goes in fees before the tax. Two such clients would take you past US$6,000 a year and end the yearly fee. A bank transfer can cost less on large payments, but on a small one it can cost more: each intermediary bank on the way may take US$10–30 (Grey), and on US$400 one such fee is 2.5–7.5%. Whatever the route, test it with a small payment first.

Nepalis doing it

Data work for clients abroad is already done in Nepal by companies, and they are a good first job. Cotiviti Nepal, set up in 2004, does risk assessment and decision analytics, software and data management for a US parent company, for clients in insurance, mortgages, healthcare and government. CloudFactory Nepal, set up in 2010, processes and labels data for machine learning for “a global client base, including startups and Fortune 500 companies” (both from Onlinekhabar, 2023). Leapfrog Technology, with offices in Kathmandu and Seattle, lists AI and data, from data engineering to advanced analytics, among its services. A year in such a team teaches how clients abroad expect data checked and delivered.

Your first portfolio project

Open Data Nepal has 630 datasets, many as CSV files ready for a spreadsheet or pandas. Pick one question, and answer it from start to finish:

When do tourists come? The Tourism Ministry’s figures give arrivals month by month since 1995. Find the busy and quiet months, and what the 2015 earthquake and the 2020 pandemic did to them.

Is Kathmandu’s air worst in winter? The portal has readings of PM2.5 and other pollutants from monitors at the US Embassy and in Phora Durbar, from January 2015 to March 2021. Average them by month and show the pattern in one chart.

Where do workers go, and what does it cost to send money home? The data from the Labour Migration Report 2022 has labour approvals by destination country and the average cost of sending remittances to Nepal.

For each, write down the question first, clean the data and say what you changed, make the chart, and give the answer in three sentences. Say what the data can’t tell you, too.

Questions people ask

Do I need to be good at maths?
School maths, used carefully: percentages, averages, and the basic statistics in the skills list. Google’s certificate starts from nothing: no degree or experience needed. The more technical data scientist jobs need far more statistics, and in the United States typically at least a bachelor’s degree (US Bureau of Labor Statistics).
What computer do I need?
A laptop, not a phone. Power BI Desktop is free but runs only on Windows 10 or later; Microsoft recommends 4 GB of memory or more and a screen of at least 1440 × 900. Google Sheets runs in a web browser, and so does CS50’s coding space, cs50.dev, for practising SQL and Python.
Can I practise on real company data?
Only with permission, and never with people’s names or private details. Public data, like Open Data Nepal’s, is better for a portfolio anyway, because an employer can open it and check your work.
Is my English good enough?
The hardest English in this job is the short summary a manager reads in a minute. Nepal’s strongest English skill in EF’s 2025 index was writing (547), its weakest listening (495). Practise both: write a three-sentence answer for every analysis, and explain your charts aloud.

Related

Sources for this part

The sheet’s own sources are listed on it, above.