Google Consumer Surveys: even bigger data?

Google has recently launched Consumer Surveys in the US and the news has spread throughout the research community quickly. Is this a new paradigm in market research and is it a threat or an opportunity?

Google has been forthright in its argument that traditional market research can be expensive and time-consuming. Google Consumer Surveys aims to change all of that. You can watch the promotional film below …

…but the basic premise is as follows.

• The survey will be distributed via the web, using ‘survey walls’. You’ll encounter a survey wall if you try to access premium content (e.g. news articles or films) and can choose to take the survey rather than paying directly. Google and the content owner will share the fee (currently $0.10 – $0.50 per response depending on targeting).
• You can only ask 1-2 questions per response, although you could use more questions across multiple surveys. This means that you won’t necessarily be able to draw analysis across questions in the way you would in a typical survey. This does make GCS ideally placed for mobile surveys however, where brevity is king.
• Demographic questions are not allowed and Google will instead infer demographic information for you to use for analysis through IP addresses and the presence of a cookie belonging to its ad network DoubleClick. This reduces the burden on respondents, but does mean there is a small margin for error.
• It’s the ultimate self serve tool. Create the question, set your target and wait for the responses to roll in which GCS will analyse for you (in the form of cross tabs and basic charts).

Google and GCS fans have been quick to point out that GCS provides a more representative sample than opt in online research panels. Arguably this is true as the respondent does not need to have opted in (thus restricting your sampling pool) and they don’t habitually take part in lots of online surveys. As someone who uses the internet on a daily basis however and rarely, if ever, pays for content, I have to wonder exactly how representative it will be. Is the majority of the internet population paying for content on a regular basis, and if not, what proportion is likely to encounter the survey wall?

GCS is also based on standard, text-based questions at the moment. A lot of discussion in the research industry over the last few years has been around the value of text-based questions and the extent to which they reveal a respondent’s true behaviour or intentions. New techniques such as neuroscience, analysis of facial expression, ethnography and semiotics have been touted as the solutions to this issue, with practitioners claiming their enhanced ability to reveal deeper and more accurate analysis of the human condition. Google has been the largest proponent of behavioural data online to date however, through AdWords, and the two combined make a pretty potent combination.

So what is the likely impact on the market research industry?

• GCS will make large scale research available to small companies who previously couldn’t afford it. If it’s done well, this can only be a good thing.
• GCS won’t replace the full service research agency’s services as Google is not providing any skills or expertise in research design or analysis (as yet, anyway).
• It will bolster the researcher’s toolkit in terms of providing a new, and in some cases better, tool to access consumers online. Minimising the burden on respondents will be a healthy challenge for those used to serving long and boring surveys.
• It won’t necessarily provide a convenient route to target niche or B2B audiences, although this could be something that is introduced as the service develops.