Google Surveys is available to researchers in countries listed in the availability article, who can target our surveys to different countries across two different audience types. Google Surveys 360 is currently available in the United States and Canada.
Google Surveys offers two panels of respondents:
- Internet users reading content on a network of web publisher sites using Google Opinion Rewards for Publishers (available in all supported countries).
- App users opting into rewarded ad formats on Google Admob’s network of app publishers (available in all supported countries).
Within these panels, we offer stratified sampling and representative weighting for some countries, convenience sampling with representative weighting for others, and convenience sampling with no weighting for those countries that don’t have population data available. The availability of our networks, weighting, and languages by country is summarized in the table below. You can find a full description of the Google Surveys panels and methodology in this whitepaper or in the product overview. Note that the whitepapers do not include our newest network - app publishers on rewarded ad formats. Please refer to this article directly for more information about the network and the most up-to-date information about our targeting methodology by country.
Google Surveys doesn't support emailing surveys to a list of contacts, instead you can do this with Google Forms. This will enable you to send surveys with a link to your form and have the results automatically captured in a spreadsheet for review. All you need is a Google account to use Google Drive.
For more information please refer to the links below:
- Google Forms
- How to use Google Forms
- Google Drive Help Center
- System requirements and supported browsers
The following table outlines the language, weighting, and network availability for each country.
|Country||Language||Web-publishers available (Google Opinion Rewards for Publishers)||App-publishers available (AdMob)||Weighting available|
|United Arab Emirates||English||Yes||Yes||Yes|
Targeting and sample types: representative, weighted and non-representative
Targeting capabilities are dictated by the respondent network, while weighting is dependent on the availability of population data for a particular country. This section explains the nuances by network, whereas the previous section detailed the network and weighting availability by country.
Stratified sampling with representative weighting uses ground-truth data about a country’s population on three demographic dimensions (age, gender, and geography), and attempts to match the survey audience to reflect the same distribution of demographics. Google Surveys also uses weighting afterwards to more closely match the target population. We use the U.S. Census Bureau’s Current Population Survey (CPS) Internet Supplement as ground truth data in the US. In other countries, we rely on a combination of government data and internal Google data sources.
Convenience sampling with representative weighting offers no guarantees of what age, gender, or sub-region within a country users will come from, but we do weight the results after the survey is complete, provided we have ground truth data available for the country and the demographics of the survey sample don’t vary too far from our ground truth data set (learn more about our weighting methodology here). You can check whether a country has weighting available in the table below. We target these surveys to any users available who are located in the target country and speak the language your survey is created in (using their browser language settings). This creates a higher degree of sampling bias, but when weighting is possible, these surveys can still be useful for analyzing trends of cycles within a population.
Convenience sampling with no weighting offers no guarantees of what age, gender, or sub-region within a country users will come from, and we make no effort to reweight the sample due to lack of available population data. This creates a high degree of sampling bias, and results should be used with caution to analyze trends or make inferences about the broader population. All networks use this method when population data is not available (as indicated in the table above).
In the following table, you can see details for each country in which we can survey users and what sample type we use.
|Google Opinion Rewards for Publishers||Stratified sampling with representative weighting||Australia, Brazil, Canada, France, Germany, Japan, Mexico, Spain, United Kingdom, United States|
|Convenience sampling with representative weighting||All other countries with weighting available|
|App Publishers (AdMob)||Convenience sampling with representative weighting||All countries with weighting available|
General population or network selection targeting
You have the option to target either Android users or General Population for countries where both are available (reference country table above). If you choose General Population, we will automatically determine whether to source respondents either from mobile app publisher audiences (via AdMob rewarded ad formats) or web publisher audiences (via Google Opinion Rewards for Publishers), depending on several components of your survey design. The following criteria is used to determine whether our mobile or app publisher network is appropriate. We prioritize surveys on our app publisher network, if possible, due to its expanded reach with similar population representativity:
- Survey length: Surveys with five or less questions are eligible for our app publisher network, longer surveys will be fielded on our web publisher network.
- Survey questions: Our app publisher network supports multiple answer, single answer, and open ended text question formats (with and without images for the three formats), but we’re working on expanding question format coverage over time.
- Single language: Bilingual surveys are not supported on either our app publisher network or our web publisher network.
- Country level targeting: More granular geographic targeting (for example: state) is not currently supported by our app publisher network and will be sent on our web publisher network.
Standard website-satisfaction surveys are currently available only in English-language markets. You can write your own custom website-satisfaction survey in any of the languages where general-population and Android-smartphone-users targeting is available (includes English, Spanish, Portuguese, German, Italian, Dutch, and Japanese). Note that custom website-satisfaction surveys are not free.