Market research agencies: the six whose clients describe actual studies

Sortlist’s market research listing is crowded with digital marketing agencies that tick “market research” among a dozen services. We read the reviews on the profile of each of the 74 agencies with at least 10 reviews and no sponsored placement. For 53 of them, not a single review describes a survey, a focus group or a study – clients praise their SEO, social media, websites or branding instead. Their ratings may be deserved; they just tell you nothing about their research.

Tip: before you shortlist any “market research agency”, open its reviews and count those that mention a survey, interviews, a sample or a study. If almost none do, ask for two research reports it has delivered – with the client’s name blacked out if needed.

Most “top market research companies” lists are dominated by the same global groups. They run syndicated panels, retail measurement and multi-country trackers that a boutique cannot match – but they usually work on annual contracts with large brands, and few publish client reviews. We looked each one up on Sortlist:

Global research companies with a Sortlist profile

These companies have a Sortlist profile but no verified client review yet, so we cannot score them with the same method. The descriptions are factual summaries of what each one is known for:

• Kantar – A global data and research group known for brand equity tracking, advertising testing and innovation research.
• NielsenIQ (Chicago) – A consumer intelligence company focused on retail measurement and consumer panels; it combined with GfK in 2023.
• GfK (Woking, UK) – A market research company known for consumer and retail panels, now part of NielsenIQ.
• Gartner (Stamford, CT) – A research and advisory company focused on technology and business functions, sold mainly by subscription.
• Forrester (Cambridge, MA) – A research and advisory firm covering technology, customer experience and B2B marketing.
• Dynata (Plano, TX) – A first-party data and online panel company that many agencies use for survey fieldwork.
• Mintel (London) – A market intelligence company that publishes consumer and category reports.
• Savanta (London) – A data and market research consultancy.
• Attest (London) – A consumer research platform for running your own online surveys.
• Drive Research (Syracuse, NY) – A full-service market research company in upstate New York.

We could not find a Sortlist profile for Ipsos, YouGov, Qualtrics, Euromonitor International, B2B International, Appinio or Suzy, which also appear in other rankings. Without verified reviews on Sortlist, we cannot compare them fairly with the agencies above.

1. The sample

Sortlist lists 1,618 market research providers on this page. We evaluated the 220 agencies on the first three pages of the listing, in its relevance order. 77 of them have at least 10 verified client reviews – our threshold for a rating to be meaningful. 3 of those also pay for a sponsored placement and were set aside, leaving 74. Of these, 6 agencies with at least 2 client reviews that describe research work (surveys, focus groups, market or user studies) were ranked, with 193 verified reviews between them and an average rating of 4.94; the 68 others are listed separately.

2. The score

3. Independence

• Sortlist is a marketplace, not an agency: it does not appear in its own ranking.
• Agencies cannot pay to enter the ranking or to move up. Agencies with a sponsored placement on this listing are excluded from scoring and shown separately.
• The order of the listing page, subscriptions and badges play no part in the score – only the three criteria above.

4. Limits

Reviews reflect each agency’s work across all its services, and agencies with few reviews may be excellent but are not yet scored. That is why every entry says what it is best for – and when another agency would suit you better. The ranking is updated regularly.

“Market research company” covers very different businesses. Knowing which one you need saves you from paying a consultancy for a survey you could field yourself – or from buying raw panel data you cannot analyse.

The agencies in our ranking fall into the last three rows of a full-service spectrum: MRF and Effet Majeur are research firms, Labbrand pairs consumer research with branding, Unlock Management pairs market studies with market entry, and Adswell and Udjat Dubai use research to steer their own campaigns.

Start from the decision, not from the method. “Should we launch in Spain?”, “Which of three prices?” and “Why do trial users churn?” need three different studies.

How many respondents do you need?

For a survey, the margin of error depends on the sample size, not on the size of the market. At a 95% confidence level, for a result around 50%:

The trap: a sample of 1,000 split into ten segments is ten samples of 100. Decide which segments you will compare before you agree on the sample size, and ask the agency to state the margin of error for each.

B2B studies fail for one reason more than any other: the wrong respondents. The person who answers a consumer panel survey claiming to be an “IT decision-maker” may not be one. What changes when your buyers are companies:

• Small populations, small samples. There may only be a few thousand companies in your target market. Thirty to fifty well-chosen interviews often teach you more than a survey of 300 doubtful respondents.
• Verify who answers. Ask how respondents are recruited (your customer list, LinkedIn, specialist B2B panels, phone) and how job titles are checked. Insist on screener questions that a non-expert would fail.
• Map the buying group. In B2B, the user, the budget holder and procurement want different things. Quota the sample by role, not just by company size.
• Use your own customers – and your lost deals. Win-loss interviews with prospects who chose a competitor are one of the most useful B2B studies, and an outside researcher gets franker answers than your sales team.
• Budget for incentives. Senior professionals rarely answer for a voucher. Agree with the agency on the incentive per interview and who pays it, before fieldwork.

On Sortlist, Unlock Management and MRF describe B2B market entry and pricing studies in their reviews; Labbrand has published B2B market strategy work for Schindler.

Of the 220 market research profiles we evaluated, 193 state a minimum project budget. 71 accept projects from €1,000 and 13 start lower, while 51 start at €5,000 or more. Among the 6 ranked agencies, entry points range from €1,000 to €10,000.

What drives the price

Ask each agency to price the same brief line by line – design, sample, fieldwork, analysis, report – so you can see where the money goes.

Agencies quote what they understand. Send the same brief to three to five of them, with these points – copy the list as is:

1. 1

   The decision

   What will you decide with the results, and by when? (“Launch in Spain in Q2 or wait.”)

2. 2

   What you already know

   Sales data, previous studies, customer feedback – so they do not charge you to rediscover it.

3. 3

   Who to ask

   Target audience, countries, quotas (age, role, company size, customers vs non-customers).

4. 4

   Hypotheses to test

   Three to five statements you believe are true and want to confirm or kill.

5. 5

   Budget range

   A range lets agencies propose the right method instead of guessing.

6. 6

   Deliverables

   Report format, presentation, raw data file, verbatims, dashboard – and in which language.

7. 7

   Constraints

   Deadlines, data protection requirements, who must approve the questionnaire.

Online survey data can look clean and still be wrong: bots, professional respondents and careless answers are a known problem in panel research. Ask these questions before fieldwork and again when you receive the results:

Red flag: an agency that shows you only charts, never the questionnaire, the sample source or the number of responses behind each figure.

Many questions can be answered in-house with free or cheap tools – an agency adds value when you need a representative sample, hard-to-reach respondents or an outside view your team will trust. A do-it-yourself plan in six steps:

1. 1

   Write down the decision and three hypotheses

   Without them you will collect interesting facts that change nothing.

2. 2

   Mine what already exists

   Your CRM, support tickets, sales call notes, Google Search Console queries, and public statistics (national statistics offices, Eurostat, the US Census Bureau).

3. 3

   Read your competitors’ reviews

   Reviews on marketplaces, app stores and B2B review sites show what buyers praise and complain about, in their own words.

4. 4

   Interview ten customers and five lost prospects

   Ask about past behaviour (“the last time you…”), not opinions about your idea – people are poor predictors of what they would buy.

5. 5

   Run a short survey

   Ten questions maximum, sent to your customer list or a self-serve panel; test it on five people first.

6. 6

   Test with real behaviour

   A landing page, a pre-order or a pilot offer tells you more than a stated intention to buy.

Where do-it-yourself research goes wrong: only asking existing customers (they already like you), leading questions (“Would you use a faster…?”) and reading too much into small samples. If the decision is expensive to reverse, get a professional to at least review your questionnaire and sample.