Market Research Company in Chennai: Draw the Map Before You Hire

Studies can be targeted by city, NCCS class and age, as described on the panel methodology page. Hercules does not publish a panel size for Chennai alone, so ask for the eligible count before you commission a study.

• Study types buyers consider: Automotive brand tracking, Tamil consumer sentiment, FMCG product testing, Regional language surveys, BFSI product research, Manufacturing consumer feedback

In 2011 Chennai district and the Chennai municipal corporation covered the same people: the final mother-tongue tables give both a population of 46,46,732 (C-16 district table, Tamil Nadu; C-16 town table, Tamil Nadu). The provisional figure for the corporation was 46,81,087. The urban agglomeration is the corporation plus dozens of neighbouring municipalities, town panchayats, census towns and outgrowths, which the Census lists one by one (Census 2011, constituents of urban agglomerations).

Town and city figures come from Table 2 and the agglomeration total from Table 3 of the provisional totals. Sholinganallur, Perungudi, Pallikaranai and Medavakkam also appear as separate constituents of the agglomeration; each was below the one-lakh line that Table 2 uses, so their populations are not shown.

Two consequences follow. A sample drawn only from the corporation leaves out close to half of the agglomeration. And localities that many briefs treat as central Chennai were separate towns in the 2011 count. Municipal limits can be redrawn between censuses, so ask each vendor which boundary its Chennai sample uses and get the answer in writing. Do not mix provisional and final figures in one quota calculation.

Census 2011 recorded each person's mother tongue. For Chennai district the language groups break down like this (Census 2011 C-16, Tamil Nadu):

Note that the Census reports language groups: its Hindi row, for one, adds in Bhojpuri and other related mother tongues. Three points follow for the questionnaire. Tamil is not optional: close to four in five residents gave it as their mother tongue. Telugu, at 9.3%, is a large enough group that you may want to read it as its own cell. And mother tongue is not the same as the script someone reads most easily, which the Census does not measure. Put a language choice on the first screen and record which version each respondent took.

Read the sub-rows before you brief a vendor on minority languages. The Census files Saurashtra, the mother tongue of 11,863 people in Chennai, inside the Gujarati group, which is why that group totals 31,984 while Gujarati itself is 19,690. A request for "Gujarati speakers in Chennai" can therefore mean two different populations.

People say they live "in Chennai" whether they live in Mylapore or Tambaram, so a quota by area only works if the screener asks a sharper question. Below is an illustrative template. It codes each answer to one of three frames from the Census table above.

S1. Where do you live at present? (one answer)

1. Inside the Chennai corporation area
2. Ambattur, Avadi, Tiruvottiyur or Madavaram
3. Pallavaram, Tambaram or Alandur
4. Sholinganallur, Perungudi, Pallikaranai or Medavakkam
5. Another town or village near Chennai (please name it)
6. Outside the Chennai area

S2. Please type the name of your locality. (open answer, used to check S1)

The 17,46,696 is the sum of the seven municipalities in the first table. The 22,68,227 is what is left of the agglomeration after taking out the corporation and those seven. Shares are rounded.

Three rules make the screener usable. Code every answer 5 against the Census list of agglomeration constituents before it counts toward a quota. Decide in advance what happens to people outside the frame: screen them out, or keep them in a separate cell that never enters the Chennai result. And check S1 against S2, because people describe where they live in present-day terms, while the frames are 2011 boundaries.

This page names no Chennai firms. None of the firms we checked showed a Chennai office on its own website, and we do not list firms from directory entries alone.

1. Ask where the field team sits. A sales office in the city and interviewers based there are different things. Ask where interviewers are briefed and who supervises them.
2. Ask who does the Tamil and Telugu work. In-house interviewers or a subcontractor? Who translated the questionnaire, and who checked the translation against the English master?
3. Ask for the field map. A vendor that cannot tell you whether Tambaram or Sholinganallur is in its Chennai sample has not planned your study yet.
4. Ask how area is verified. Show the vendor the residence screener above and ask how it codes write-in localities and what it does with respondents outside the frame.

Chennai briefs often grow into state studies. For an official way to rank the state's cities, borrow the central house rent allowance (HRA) classes. Chennai (UA) is Tamil Nadu's only "X" city, while the Coimbatore, Madurai, Tiruchirappalli, Salem, Tiruppur and Erode urban agglomerations are "Y" (Finance Ministry HRA memorandum, July 2017). Every other town is "Z".

If the decision is state-wide, sample those cities by name and keep their results apart from Chennai's. Sampling a named list of non-metro cities is covered step by step in this guide to Tier 2 and Tier 3 research.

Start with the frame question. The methodology page says targeting goes down to city level, alongside NCCS and age (panel methodology); it does not say whether Tambaram or Sholinganallur can be picked apart from Chennai. If they cannot, the residence screener does the job inside the survey.

A 600-response survey on Pro costs 600 × ₹120 = ₹72,000 at published prices, and a 100-response test on Plus costs ₹15,000 (pricing). Screen-outs from answer 6 are a feasibility question, so ask how many eligible members sit in each frame before you set quotas: no Chennai panel count is published.

Respondents answer in an app, so households without smartphones fall outside every frame in the table. Surveys run in as many Indian languages as the study needs; get Tamil and Telugu confirmed in writing.