How to Size the Nepali Market Without Guessing: A Sourced Workbook for Population, Household and Spending Data

Ask ten people how big the Nepali market is and you will get ten confident numbers, most of them untraceable. Some will be converted from an old dollar figure. Some will be a GDP number wearing a market-size hat. A few will be honest guesses dressed up with a decimal point.

This workbook does the opposite of what a market-size page usually does. It does not give you a number. It gives you the sources, the arithmetic and the checks, so that whatever figure you end up with is one you can defend when someone in the room asks where it came from.

Start with the denominators: population and households

Almost every sizing exercise in Nepal begins with the national census, and almost every mistake begins with forgetting which kind of count is being quoted. The most recent census was taken with a reference date of 25 June 2021, or 11 Asar 2078 in the Bikram Sambat calendar, and it is published by the National Statistics Office (NSO), the body that took over the work of the Central Bureau of Statistics.

Three things in the census matter more than the headline population:

• Households, not people. If your product is bought per household, the household count is your denominator and the average household size is your sanity check on it.
• Where they live. District and urban/rural splits are published, and a national average that hides Kathmandu Valley concentration will mislead you badly.
• Who was counted. Census counts follow residency rules. Absentees working abroad, seasonal migrants and institutional populations are handled in specific ways. Read the methodology note before you treat the total as your addressable audience.

If a source gives you a population figure without saying whether it is a census count, an official projection, or a de facto estimate, you cannot compare it with anything else. Put it down and find one that can be dated.

Then the spending side: two series that will not match

Nepal publishes household spending in two different places, built for two different purposes. The Nepal Living Standards Survey (NLSS), run by the NSO, measures consumption by asking households. Rounds are roughly five years apart, so before you use one, confirm which round is the latest published and when it was released. The national accounts estimates, also produced by the NSO, give household final consumption expenditure as part of the GDP accounts.

Pull both. They will disagree. That disagreement is normal and it is where most amateur sizing goes wrong.

The reconciliation box

Why the two series diverge:

• Coverage. A household survey reaches private households. National accounts also carry imputed rent for owner-occupied housing, own-account production, and the spending of non-profit institutions.
• Imputation. The accounts estimate some components indirectly, from commodity flows and residuals, rather than by asking anyone anything.
• Period. Survey recall periods, fieldwork dates and the accounting year are not the same window.
• Basis. One series is commonly quoted per capita or per household, the other as an economy-wide total. Dividing or multiplying across the two is where fake precision is born.

Rule of thumb: use the survey when the question is about distribution, that is, who spends, how much per household, and how that varies by province or income group. Use the national accounts when the question is about the aggregate and you need it to sit consistently alongside GDP. Do not mix a per-capita survey figure with an aggregate accounts figure and present the result as a market size.

Money crossing the border: NRB data

Nepal Rastra Bank publishes remittance inflows, the balance of payments and credit aggregates, most visibly in its periodic macroeconomic situation reports. These matter for consumer sizing because a meaningful share of household spending power arrives from outside the country. Two disciplines apply. First, cite the specific issue you used, because monthly figures get revised. Second, if you convert to another currency, state the rate and the date, and keep the NPR figure visible next to it.

The five-test checklist

Run every dataset through this before it touches your model.

1. Definition. What exactly is being counted, and does that match your category? "Food" in a survey is not "packaged snacks" in your spreadsheet.
2. Denominator. Per person, per household, per adult, or total? A per-capita number multiplied by the wrong population is the single most common error.
3. Vintage. Which edition, published when, covering which year? Write it in the sentence, not in a footnote.
4. Geography. National, provincial, district, or urban only? A national average is rarely a Kathmandu number.
5. Collection method. Survey, administrative record, or modelled estimate? Each carries a different kind of error.

A worked example, with invented numbers

Everything in this paragraph is illustrative arithmetic. None of these inputs come from a Nepali source and none of them describe Nepal.

Suppose a place has 10 million households, and suppose each spends NPR 600 a month on your category. Ten million times 600 is NPR 6 billion a month, or NPR 72 billion a year. Now change one input: if the real figure is 300 a month, your market halves instantly. That is the whole lesson. Sizing is not about finding a bigger number, it is about knowing which of your inputs is doing the work, and how much damage a wrong one does.

What this page refuses to do

No headline market-size figure. No gated download in exchange for your email. No reproducing paywalled reports. Institution names, report titles and release schedules change, and a workbook that hard-codes them without a date ages badly. Whatever edition you cite, name it and date it in the same sentence, and if you cannot confirm a detail on the day you write, say it is unconfirmed rather than smoothing it over.

Where this fits

If you have not yet settled whether you need a full sizing exercise or a smaller, sharper question, start with our overview of the Nepalese market landscape and come back here once you know what you are measuring.