Latest US poll: 63% say gas prices are affecting their household budget a great deal or a moderate amount
See the pollOur methodology
Every client study goes through the same six stages, from the decision you need to make to the file your team reads. This page explains each one, and helps you find the method that answers your question.
Six stages
The rules below come from the methods we run. Each one links to the study it comes from.
Start from the decision, then pick the method that answers it.
Entering a market, defending share and changing your position each need different questions. We agree the decision first.
From: Competitor researchFor pricing: Van Westendorp when you need a range, Gabor-Granger when you have a few candidate prices and want demand at each one. Both in one survey, with the open price questions first, so no price we show sways the answers.
From: Pricing researchIn a concept test, each person sees one idea only, so the ideas never colour each other's scores.
From: Concept testingIn a brand funnel, brand names come first without help, then from a list, then consideration, first choice, recent purchase and recommendation. This order stops earlier questions from leading later answers.
From: Brand funnel & KPI snapshotClaims are made alike in length and style, so a claim does not win just because it is better written.
From: Claims testingMany questions come from our library of 309 ready-made templates, then are edited to fit your category.
From: Usage & attitudeDecide exactly who answers, and how many.
A short screener lets in only the people the study is about, such as people who bought the category recently.
From: Concept testingTargets by age, gender, region and other profile answers, so the sample matches your plan.
From: Audience samplingFor a concept test we suggest at least 200 people per idea, and 300 if you want to compare groups.
From: Concept testingEach segment needs enough people to be profiled, so a segmentation sample is larger than for most studies.
From: SegmentationTell us the audience, the country and the number of completes. If the audience is too small, we say so before we quote, not after fieldwork starts.
From: Audience samplingRun the survey the same way for everyone.
Studies run on PocketsFull, the panel we recruit and pay ourselves. Members answer a survey's qualifying questions in the app before it opens. For audiences the panel cannot reach, we can work with partner panels and tell you when we do.
From: PocketsFull panelFor a survey hosted elsewhere, we set up the return links for completes, screen-outs and full quotas, so each outcome is recorded on both sides.
From: Audience samplingInvitations go to members who match your criteria, and our team follows each quota until it is full.
From: Audience samplingIn a tracker, each wave uses the same screener, the same quotas and the same weighting.
From: Brand health trackingA video ad plays to the end before any questions appear, and an audio spot plays once, as it would on air.
From: Video ad testingCheck the people and the answers before anyone reads them.
Studies on our panel run with attention checks on every response, and checks that open answers stay on topic.
From: Usage & attitudeVideo and audio answers are checked for length, words and topic, and video for a live person in enough light. A recording that fails goes back with the reason.
From: AI video interviewsWe review the data for incomplete answers, failed trap questions, speeders, straight-liners and answers that contradict each other, and agree any removals with you before the final count.
From: Audience samplingSeparate real differences from noise, and say what they mean.
Up to five audience groups side by side, with markers where differences are significant. In a tracker, changes between waves are tested before anyone acts on them.
From: Brand health trackingRim weighting to the targets you set for your survey.
From: Reports & exportsPeople say they will buy more often than they do, so in market sizing we scale stated intent down and show that adjustment as an assumption you can change.
From: Market sizing & demand forecastingOpen, audio and video answers are grouped into themes: the main points people made, how many raised each one, and the tone.
From: AI video interviewsA chart for every question type plus 36 more chart views, Ask the study for plain-English questions, a PowerPoint deck, an Excel workbook, CSV files, chart images and the AI report as a web page.
From: Reports & exportsNamed standards and codes, not adjectives.
The international standard for market, opinion and social research. It sets requirements for how research projects are planned, run, checked and reported.
From: AccreditationsQuality management, and information security management under a managed, audited system.
From: AccreditationsAs an ESOMAR member, we work to the ICC/ESOMAR International Code.
From: AccreditationsAs an MRS Company Partner, we commit to the MRS Code of Conduct.
From: AccreditationsWhich method?
Start from the decision you need to make. Each answer opens the method that fits it.
Which of our ideas should we take forward?
The method
Find the winning idea before you build it
How it runs
What we measure
We have a long list of ideas. Which few deserve a full test?
The method
Find the few ideas worth a full test
How it runs
What we measure
What should we charge?
The method
Find the price your buyers will accept
How it runs
What we measure
Which claims, features or messages matter most?
The method
Rank a long list by what matters most
How it runs
What we measure
Which line-up of products reaches the most buyers?
The method
Find the line-up that reaches the most people
How it runs
What we measure
Will our pack get seen, read and picked?
The method
Check your pack gets seen, read and picked
How it runs
What we measure
Will this ad work before we spend on media?
The method
Test your video ad before the media spend
How it runs
What we measure
Is our brand growing, wave by wave?
The method
See how your brand moves, wave by wave
How it runs
What we measure
Who are our customers, really?
The method
Find the groups inside your market
How it runs
What we measure
Where does the customer journey break?
The method
See where the customer journey breaks
How it runs
What we measure
How satisfied are customers, and what should we fix?
The method
Measure satisfaction, then find what to fix
How it runs
What we measure
Why do people do what they do, in their own words?
The method
Real people, on camera, in their own words
How it runs
What we measure
Glossary
The words you will meet in a proposal or a report, explained the way our method pages use them.
Public polls
Catalyse Pulse, our public polls, is a different job from client research, and it has its own method page.
Tell us the decision you need to make, and our researchers will suggest a method and explain why.
Offices: London (head office) · Orlando, FL · Gurgaon, Haryana