The global market research companies a Singapore buyer meets, ranked on what each one publishes before you brief it

If you are shortlisting a global research firm from Singapore, start with what each one publishes about itself rather than with a world ranking, because the ranking is harder to get at than the firms are. Esomar, the industry body whose global figures most top ten lists are built on, served a page title and nothing else to a plain fetch on 29 October 2026: esomar.org returned HTTP 200 at about 15 KB with 56 characters of readable text, and its global market research report URL returned the same. So this page does the thing a buyer can repeat. It reads the five firms a Singapore buyer actually meets, on their own sites, on one day, and sorts them by how much each one states about what it sells and where it works. Demand for this question here is small: Google Keyword Planner against geo 2702 returns 10 searches a month for this phrasing on 26 October 2026, taken as the highest figure in the group rather than the group added together.

Why two top ten lists never agree

Four different things get counted and all four are called size. Reported revenue puts the data and measurement businesses on top, because retail measurement and media panels are large contracts. Panel scale favours whoever owns the most respondents. Awards and methodology prizes favour the specialists. Directory position, on the platforms that sell placement, favours whoever paid. A list that does not say which of the four it used is a directory with a headline, and the honest use for it is building a long list to work through.

The criterion on this page is disclosure, which is checkable by anybody with a browser: what does the firm state, in the document its own site serves, about the work it sells and the markets it covers. Disclosure is not fieldwork quality, and fieldwork quality is settled at the briefing with the questions further down. Disclosure is what you can verify before you spend an hour on a call.

What each firm's own site states, read on 29 October 2026

Firm	Page read	How it describes what it sells	Figures or products it names	Singapore in the served document
Ipsos	ipsos.com/en-sg, HTTP 200, about 57 KB	"Ipsos is one of the world's leading market research companies", delivering information for understanding "Society, Markets and People"	"5,000+ clients in the world", "90 Markets", "6M+ authenticated, proprietary panelists"	None
Kantar	kantar.com, HTTP 200, about 91 KB	"Intelligence for Brand Growth", split into Signal Intelligence, Decision Intelligence and Strategic Intelligence	Named products include BrandZ, Kantar Marketplace and Media Reactions 2026; articles dated 5 and 6 October 2026	None
NIQ	nielseniq.com/global/en, HTTP 200, about 374 KB	"The Full View of Consumer Intelligence", organised by solutions, industries and a partner network	Names an AI section of its own and a partner network rather than client or panel counts	None
YouGov	yougov.com, HTTP 200, about 756 KB	Panel based data and analytics, with a long published solutions list covering brand health tracking, campaign measurement, creative and ad testing, persona creation and qualitative research	Publishes its own ranking, "YouGov's Most Recommended Brands 2026"	Once, as a linked article on Singapore's most considered beauty and personal care brands 2026
Mintel	mintel.com, HTTP 200, about 328 KB	Describes itself in its own page title as a global market intelligence and research agency, delivered three ways, On-Demand, Consulting and Integrations	Names 2027 Global Predictions and an industry list from beauty to financial services	None in the text, with Asia Pacific landscape pages linked

Every figure above is the firm's own claim about itself, read in the document its server returned on that date, and none of it is independently audited here. A figure a company publishes about its own scale is still the most useful starting point, because it is the number it has committed to in public.

The order, on the stated criterion

1. Ipsos publishes the most that a buyer can check, because it puts hard counts on the page: clients, markets and panel size, in numbers rather than adjectives.
2. YouGov publishes the most about the work itself. Its solutions list names the individual studies you would buy, which makes it the easiest of the five to send a one line enquiry to without a scoping call first.
3. Kantar publishes a clear product architecture and dated thinking, so you can tell which division you are walking into, which matters at a firm of its breadth.
4. Mintel publishes its delivery models plainly, which is the part buyers get wrong most often, since a subscription to a report library and a commissioned study are different purchases at different prices.
5. NIQ publishes the least of the five about scale or price on its home document, which is consistent with a business whose large contracts are sold through conversations rather than pages.

That order answers one question only, which is who tells you most before you call. None of the five publishes a price, and for commissioned work that is normal: the cost depends on sample, markets, method and timing, so the quote follows the brief.

A country URL is not a promise of Singapore fieldwork

This is the finding worth carrying into your shortlist. Ipsos serves a Singapore URL, ipsos.com/en-sg, which returned HTTP 200, and the document it served carried the firm's global description with no occurrence of the word Singapore. The NIQ, Kantar and Mintel home documents carried no occurrence of it either, and the only Singapore reference across all five served pages was one linked YouGov article about Singapore consumers. A served page is evidence about that page and about nothing else, so none of this says a firm lacks a Singapore team. What it does say is that the question has to be asked rather than read.

Ask it in this order, and ask for names. Who runs the fieldwork for this study, in which office. Which panel is the Singapore sample drawn from, and what is the incidence rate for my audience here. Is the moderator or interviewer local, and in which languages. Who signs off the final report, and in which time zone are the weekly calls. A firm with real capability here answers all five in one email, and the answers are what you are buying.

Which one to brief, by the question in front of you

• Continuous brand tracking suits the panel businesses, because the value is in the same question asked the same way for years.
• Category sizing and share suits the measurement businesses, which hold retail and media data you cannot assemble yourself.
• Concept, pack or advertising testing suits a firm with a named testing product, since the normative database it compares your result against is the thing you are paying for.
• A fast single question read suits an omnibus or a self serve panel product, and the global firms all sell one. Ask for the fielding window in days.
• A trend or market report you can read tomorrow is a subscription purchase rather than a study, and the report library businesses are the shortest route to it.

Those five jobs are also how to read a proposal that offers all of them. The five kinds of supplier behind one search, and how to judge the method they propose, are set out in what a business to business research agency delivers and how to judge the method, which is the page to read beside a quote.

When a smaller supplier or desk work is the better buy

A global firm earns its fee on multi market comparability, normative data and a defensible method you can put in front of a board. For a single Singapore market question with a budget in the low thousands, a local research consultancy or a specialist in your category will usually give you more decision per dollar, and the brief is easier to write because you can sit in the same room.

There is also a layer underneath both. Competitor traffic, search demand and review patterns are available without commissioning anybody, and they answer different questions than a survey does. What those estimates can and cannot settle is the subject of why every traffic number about a rival is a model. Doing that layer first makes the commissioned study cheaper, because you arrive at the briefing already knowing which questions you cannot answer on your own.

Questions a Singapore buyer asks about the global research firms

Which is the biggest market research company in the world

It depends on what is counted, and the firms themselves count different things. Ipsos publishes "90 Markets" and "6M+ authenticated, proprietary panelists" on its own page, read on 29 October 2026, while the measurement businesses lead on contract revenue. For a buyer, the better question is which one is biggest in your category in Singapore, and that is answered by asking each firm for its last three studies in your sector.

What does a study from one of these firms cost

None of the five publishes a price for commissioned work, and the quote follows the specification. Give every firm the same sample size, markets, method and deadline in writing, and you will get comparable numbers back.

Can a mid sized Singapore company buy from them at all

Yes, usually through the self serve and omnibus products rather than a bespoke programme. Ask directly for the entry level product and its minimum, because it is rarely the thing the website leads with.

How long does a Singapore study take

Ask for the fielding window and the reporting window as two separate dates. The gap between them is where most timelines slip, and a firm that quotes them separately is a firm that has planned the work.

Do I need a research firm to understand my own customers

For a decision that commits real money, an independent read is worth buying, because your own customer list cannot tell you about the people who did not buy. For a smaller decision, structured interviews with ten customers, run by somebody who did not sell to them, get you a long way.

How do I check a firm's claims about itself

Read the figures off the firm's own page and write down the date you read them, which is what this page did. Then ask for two references in your category and one study that did not go the client's way, since the second request tells you more than the first.

The research you did not commission, which an assistant is already running

There is a new source of received wisdom about your market that nobody briefed and nobody fielded. When a Singapore buyer asks ChatGPT or Gemini who the leading suppliers in a category are, an answer arrives, built from whatever pages those systems can read, and it functions in the buyer's head like a market summary. That is not a substitute for a study and it is not measured by one, since a survey asks people and this asks a model. It is checkable in its own right, by putting the question your customers would ask into the assistant and reading which sources the answer leans on. That is the reading AI Knows Us takes, at aiknowsus.com, and it sits beside the instruments on this page rather than against them, because it reads machines where a research firm reads people. If you are about to commission research on how your category is perceived, it is worth knowing what the machines are already telling your buyers while the fieldwork runs.