What You Need to Know About Google Consumer Surveys
From the company that made web analytics free comes a novel service that combines consumer research with monetization of Internet content.
On March 30, Google announced Google Consumer Surveys, a self-serve market research tool. This is an insanely clever idea that kills two birds with one stone: how to leverage the Internet for cheap, accurate market research, and how to let content publishers finally make some money from the valuable information they’ve been forced to offer for free. The premise is very simple: Google Consumer Surveys lets brands who want to do consumer research create a quick 1 – 2 question micro-survey. Content publishers host the survey on their sites. When a consumer tries to access information on the content publisher’s site, a “surveywall” pops up and asks the consumer to take the survey before they can access content.
- The brand gets back from Google some aggregated results in nice, presentation-friendly formats.
- The consumer gets to read valuable content without having to pay actual money.
- The content publisher makes some money for hosting the survey.
- Google makes some money for being so darn clever.
The pricing is simple as well: Google does everything on a per-response basis. It costs the brand only $0.10 per question-response to target a representative sample of the U.S. Internet population or $0.50 per question-response for demographic targeting using inferred demographic data such as gender, age group, and US geographic region.
What’s Good About Google Consumer Surveys
Google is using their reach to leverage a crowd-sourcing approach to market research. Rather than select a small focused audience and ask them to answer 10 pages worth of questions, the approach here is to ask a couple of simple questions and put those out to a huge audience, then run the results through some algorithms for statistical weighting.
Good response rate
During trials of Consumer Surveys, Google achieved an average response rate of 16.75% compared to the latest industry response rates of less than 1% for most Internet intercept surveys; 7-14% for telephone surveys; and 15% for Internet panels. This was due to the brief nature of the surveys, where one or two easy clicks complete the survey, and to Google’s ability to infer demographic data, making it unnecessary to add those questions to the survey.
Google’s secret sauce is their statistical expertise. They take the raw response data and run it through weighting and analysis algorithms to deliver results that turned out to be more accurate than results from the Internet panels used during trials. Google reported the average absolute error for non-Google samples was 5.29% across all benchmarks, while the Google Consumer Survey samples averaged a lower 3.76%. This suggests that brands can use Google Consumer Surveys instead of traditional Internet-based panels and achieve the same, or higher, accuracy.
It only takes a few minutes to set up a survey, and Google takes care of the rest: putting it out to their network of publishers, data collection, and analysis. If your brand needs to be aware of trends and shifts in the consumer mindset, it’s tough to think of an easier, faster way to take the pulse of your audience.
The Google recommendation is to go for at least 1,500 responses in order to achieve a statistically significant sample. At $0.10 each, that’s $150; at $0.50 each, that’s a $750 spend on a small survey that could deliver timely and extremely valuable insights to your business decisions.
A better business model for publishers.
It’s a more viable business model over the long term for content publishers than putting up a hard “paywall” for access to content. According to David Cohn, founder of Spot.us, a journalism crowdfunding site that tried a variation of this pay-for-access model, only 1% of users were willing to pay money for content, but this number rose to 10% when they could respond instead to a survey. Furthermore, the survey takers were willing to continue taking surveys, making their long-term contribution greater than those who made one-time cash donations.
What to Keep in Mind About Google Consumer Surveys
It’s good for micro-surveys
This service is ideal for A/B testing, binary questions, or questions with a few multiple choice responses. Do you prefer this name or that name? This logo or that one? This is not the place to conduct longitudinal surveys. Think of it more as polling.
Don’t expect analysis of relationships between survey questions
The nature of Google Consumer Surveys allows only one-question or screening two-question surveys. Google provides examples of multi-question surveys, where each question is set up as a separate micro-survey but this is not the same. As long as your survey does not depend on having the same 1,500 people respond to all the micro-surveys, this should not be a problem.
Visuals make it better.
While it’s easy to set up text-based surveys, companies with attractive visuals make a better impression and get higher response rates. In some cases, if there is a design choice involved, a good visual is a must.
It’s only available for the US population.
At this moment, Google has launched Consumer Surveys for US demographics but knowing them, other markets can’t be far behind.
Sample population is not ‘representative’.
Google notes that their ‘representative US population sample’ is actually representative of the US Internet-using population, who tend to be younger, better-educated, and with higher incomes. At the moment, your options for slicing demographics are limited to age, gender, and geographic region because Google infers this data from IP addresses and DoubleClick cookies. So you need to decide whether this is an issue for the type of market research you want to run.
Survey allocation depends on content publisher inventory
Also, while Google does its best to optimize how surveys get allocated to content sites, there can be limitations due to publisher inventory and survey deadlines. Again, this is where their secret sauce comes into play, algorithms that compensate for these factors. Google’s content publisher network is not as large as their display ad network, but over time as more publishers sign up, it’s reasonable to expect that these issues will go away. Google may even be able to target surveys more closely to a required demographic as a broader variety of content publishers come on board: technology, science, lifestyle, sports, arts, travel, are just a few of the verticals that come to mind. Keep in mind also that this is Google’s first iteration of Consumer Surveys. For them, it’s a proof of concept, and if the concept gains traction, expect richer functionality for surveys because the logical end game for Google is to deliver full survey capabilities.