Which Research Study Fits Your Business Decision? A Map for India
Match 38 business decisions to the right research study in India: product launch, pricing, ads, policy opinion, churn. Who to survey and which guide to read.
In shortStart from the decision, not the survey tool. Each business decision maps to one study type, one audience and one set of questions: a new product needs a concept test, a price needs a pricing study, a policy question needs a balanced opinion survey. This guide pairs 38 decisions with the study, the audience and the guide to read. Hercules Works runs these studies on the SuperJ panel in India, and its AI survey builder drafts a questionnaire from a one-paragraph brief.
Contents
- Start With the Decision, Then Find the People
- Five Answers to Settle Before You Pick a Platform
- 38 Business Decisions Mapped to a Study
- Write the Audience as a Behaviour
- How Many Responses: The Arithmetic
- Five Questions to Ask Any Platform
- When an Online Panel Is Not Enough
- Running This Study on Hercules
- Frequently asked questions
- Related guides
Start With the Decision, Then Find the People
Many research requests arrive in one sentence: "I want to survey Indian consumers about X and find a platform." That sentence hides four choices: which decision the result feeds, who to ask, which study to run and where to run it. Settle them in that order. The platform is the last choice, not the first.
This guide is a router. It gives a five-question brief, a table that maps common business decisions to a study type, the arithmetic behind sample sizes and five questions to ask any platform. The sample sizes are arithmetic, not benchmarks. The screener examples are illustrative.
Five Answers to Settle Before You Pick a Platform
A platform can only help once you can answer five questions. Write the answers in one place before you brief anyone, human or AI.
- Decision. What will you do differently depending on the result? If nothing changes, do not run the study.
- Audience. Who buys, uses or is affected? Write them as behaviours, not interests.
- Study. Which study type answers the decision? The table below maps common ones.
- Sample. How many people, in which cities, and how precise must the answer be?
- Platform. Can it reach that audience, verify the people, run the design and analyse the result?
Here is a brief you can paste into any survey tool, agency email or AI assistant. Fill every line.
| Line | Fill in |
|---|---|
| Decision | We will ___ if the result shows ___, and ___ if it shows ___. |
| Audience | Adults aged ___ in ___ (cities) who ___ (behaviour in the last ___ months). |
| Study type | ___ (from the table below). |
| Length | ___ minutes on a phone. |
| Must include | A screener, ___ core questions, an open-ended question and age, gender and city. |
| Analysis | Compare ___ by ___. Report counts, not only percentages. |
| Output | Questionnaire with routing, then a results summary. |
38 Business Decisions Mapped to a Study
Find your decision, then open the guide in the last column. The audience column is a starting description. Every audience needs a screener, as the audience guide explains.
| Decision | Study that answers it | Who to recruit | Read next |
|---|---|---|---|
| Is my product idea worth building? | Idea validation survey | People who have the problem, found with a screener | Validate a startup idea |
| Which features should version one have? | MaxDiff or Kano | Likely buyers of the category | Early-stage product survey |
| Who uses this today, and for what job? | Usage and attitude; jobs to be done | Users of the closest substitute | Usage and attitude survey |
| Which of three concepts goes forward, and will people buy it? | Monadic concept test with purchase intent | Category buyers, one concept per cell | Concept testing |
| Do people like the real product? | Product test | Category users who tried it | Product testing |
| Which pack design sells better? | Packaging test | Category shoppers | Packaging testing |
| What should I charge, and how much extra is a feature worth? | Van Westendorp, Gabor-Granger or willingness to pay | Likely buyers at your price tier | Pricing research |
| Will this ad work before I spend on media? | Ad pre-test | Your target audience | Ad testing |
| Did my campaign change awareness? | Brand lift with a control cell | Exposed and unexposed audiences | Brand lift |
| Do people know my brand, unaided and aided? | Brand awareness survey | Category buyers, screened without naming the brand | Brand awareness |
| Is my brand gaining ground over time? | Brand tracking | Category buyers, same quotas each wave | Brand tracking |
| Why do customers switch to a competitor? | Brand switching survey | Recent category buyers | Brand switching |
| Will people believe and understand my marketing claim? | Claim test with matched cells | Recent category buyers | Claim testing |
| Do my users really need the product? | Product market fit survey | Your own active users | Product market fit |
| What is changing in my market and among competitors? | Market intelligence plus a consumer survey | Public sources, then category users | Market intelligence |
| How do I reach Gen Z or health and wellness buyers? | Audience study with a screener | Age band 18-24, or wellness buyers by recent purchase | Gen Z survey |
| How do Gen Z and Millennials differ as customers? | Cohort comparison with matched life stages | Age bands 18-24, 24-35 and 35-45, split by birth year | Gen Z vs Millennials |
| How do I understand Gen Z behaviour over time? | Survey plus interviews and diaries | Adults by life stage and city | Gen Z behaviour research |
| Where does my brand stand against competitors? | Brand positioning research | Category buyers rating several brands | Brand positioning |
| How do customers see me compared with each rival? | Competitor benchmarking survey | Recent buyers of each brand | Competitor benchmarking |
| Which name or logo should I use? | Name and logo test with matched cells | Recent category buyers | Logo and name testing |
| How do my employees feel right now? | Pulse survey | Your own staff | Pulse survey |
| How do I reach doctors, tradespeople or other professionals? | Professional audience study | Role by screener, with a feasibility count | Surveying professionals |
| I need a number for all Indian adults. | National omnibus or probability sample | Adults 18+, representative design | Omnibus vs online panel |
| My results may be biased. How do I check? | Bias review and a pilot | Your sample and questionnaire | Survey bias |
| I am doing a thesis or research paper. | Academic survey with a stated sampling route | A defined population and ethics approval | Academic research respondents |
| Who exactly is my customer? | Segmentation | Category buyers, broad sample | Segmentation |
| How big is the market? | Market analysis survey | Adults in the target geography | Consumer market analysis |
| How do shoppers choose at the shelf or app? | Shopper path to purchase | Recent buyers | Shopper path |
| Are customers satisfied after a purchase or support call? | CSAT survey | Your own customers, or category users | CSAT survey |
| Why do customers leave? | Churn survey | Your own lapsed customers | Churn survey |
| Where does the customer journey break? | Journey survey | Customers by stage | Customer experience survey |
| Should I enter smaller cities or rural India? | Tier 2 and 3 or rural study | Residents of named cities or districts | Tier 2 and 3 research |
| What do Indians think of a government policy? | Policy opinion survey | Affected adults in named cities | Policy opinion survey |
| What do business buyers need? | B2B decision-maker survey | Role and buying authority, by screener | B2B decision-makers |
| Is my app or website easy to use? | Usability survey | Real or likely users | Usability testing |
| How do my employees feel? | Employee survey and eNPS | Your staff | Employee surveys |
| I need depth, not percentages. | Moderated interviews | A small, well-screened group | Moderated interviews |
If your decision is not here, the study types guide lists all 22 study types, and the research design chooser helps you pick between monadic, sequential and paired designs.
Write the Audience as a Behaviour
A panel cannot find people who are interested in electric cycles. It can target adults by city, age band, gender and NCCS, then use a screener to confirm what they do. Turn every wish into a behaviour you can ask about. The examples are illustrative.
| What you say | What to screen on |
|---|---|
| People who care about E20 petrol | Bought petrol for a car or two-wheeler in the last 3 months |
| Tier 1 city people who might ride an electric cycle | Lives in a named city, travels for work or study on 2 or more days a week, and has used a bicycle, two-wheeler or public transport |
| Parents of young children | Has a child in your age band, picked from a household list with decoys |
| Small business owners | Role and buying authority, picked from an occupation list with decoys |
Ask for a feasibility count for your exact audience before you commit. Incidence decides how many people must be screened to find enough of them.
How Many Responses: The Arithmetic
The margin of error for a share near 50% is about 1.96 times the square root of 0.25 divided by n, assuming a simple random sample. The figures below are that arithmetic, not a promise. An app-based panel is not a random sample of India, so treat them as the best case.
| Responses | Margin of error (95%, share near 50%) |
|---|---|
| 30 | ±17.9 points |
| 100 | ±9.8 points |
| 250 | ±6.2 points |
| 400 | ±4.9 points |
| 1,000 | ±3.1 points |
Each subgroup has its own margin. Thirty people in one city give ±17.9 points for that city, so use city cells to spot direction, not to rank cities.
Five Questions to Ask Any Platform
- Reach. Can you reach my exact audience in my named cities? What is the feasibility count, and how did you get it?
- Verification. How do you confirm each respondent is a real, unique person? See verified survey respondents.
- Quality. What happens to speeders, straight-liners and inconsistent answers, and are screened-out people billed? See survey data quality.
- Design. Does it run the design I need: screeners, routing, rotation, MaxDiff, image or video stimuli?
- Analysis. What do I get back: raw data, cross-tabs with significance tests, coded open-ended answers?
A respondent platform and a survey form are different products. Google Forms collects answers from whoever you send it to. A panel supplies the people. Read Google Forms alternatives if you are unsure which you need.
When an Online Panel Is Not Enough
- National estimates. A figure for all Indian adults needs a probability sample. Use a method that publishes its sampling frame and weights.
- Offline households. App-based panels miss them. Add face-to-face fieldwork. See rural consumer research.
- Regulated topics. Health, finance and elections carry rules on consent, safety reporting and publication. See the pharma, BFSI and political survey guides.
- Depth. If you do not yet know what to ask, interview a few people first.
Running This Study on Hercules
If your job is a consumer survey in India among a defined audience, this is what a typical test, pilot and main study cost at list price. The AI survey builder can draft the questionnaire from a one-paragraph brief like the one above.
| Step | Plan | Responses | Price per response | Cost |
|---|---|---|---|---|
| Test the questionnaire | Free | 30 | ₹0 | ₹0 |
| Small pilot | Plus | 100 | ₹150 | ₹15,000 |
| Main study | Pro | 400 | ₹120 | ₹48,000 |
Pro is sold in bundles of 400 to 1,000 responses and costs 20% less per response than Plus. Enterprise plans come with a dedicated research team and are priced on request.
The SuperJ network has a 25M+ consumer footprint. The survey-eligible panel is a subset, and availability depends on your targeting. Members complete Aadhaar-based KYC once, and each study receives a zero-knowledge proof of verification. Surveys run in as many Indian languages as the study needs. Quality checks flag speeders, straight-lining and inconsistent answers. Read /panel-methodology and pricing.
The limit: the app reaches smartphone users only, so offline households need face-to-face fieldwork.
Frequently asked questions
How do I find real Indian respondents for a survey?
Use a panel whose members are verified and answer on their own phones, then target by city, age band, gender and NCCS and confirm the behaviour with a screener. Ask any provider how it verifies people and what it does with speeders. Social-media links reach whoever sees them, so they suit early tests, not studies that must match an audience.
Which survey platform should I use for market research in India?
It depends on the job. A survey form is enough if you already have the people. A respondent panel is needed if you do not. A national estimate needs a probability sample. Use the five platform questions in this guide and ask for a feasibility count for your audience.
Can AI draft the survey for me?
Yes, as a first draft. Give it the decision, audience, cities, study type and length, and ask for a screener, routing and the analysis plan. Review every question for leading wording and test the draft on a small pilot before the main study.
How many responses do I need?
It depends on how precise the answer must be and how many subgroups you will compare. For a share near 50% in a simple random sample, 100 responses give about ±9.8 points and 400 give about ±4.9 points. App-based panels are not random samples, so treat these as the best case.
Can I survey a niche audience such as electric vehicle buyers or petrol users?
Yes, through targeting plus a screener. Write the audience as a behaviour, such as bought petrol in the last 3 months, and ask for a feasibility count. We publish no counts for any category.
How can customer research help me grow my business?
Use it to make specific decisions cheaper to get right: which product to launch, what to charge, which message works, why customers leave. Name the decision first, then match it to a study. See the data-driven decision guide.
How do I design the questionnaire?
Map each question to a decision, order it screener, warm-up, core and profile, ask one thing at a time and pilot it. See questionnaire design.
What does it cost to run a survey on Hercules Works?
Free covers 30 responses. Plus is ₹150 per response from 100 responses. Pro is ₹120 per response in bundles of 400 to 1,000. Enterprise plans are custom. See the pricing page.
Ready to get real consumer insights?
25M+ consumer footprint. Free to start.