How to get the most of the Google consumer survey tool
What makes the Google Consumer Survey tool so appealing? Like all Google products, it gives you an impressive amount of data. What makes the Google Consumer Survey tool so scary? It is very easy to spend a lot of money on a survey that provides no useful data. Let’s take a look at the basics of how to make a Google Survey that provides value.
1. Coupons and Campaign Costs: They don’t let you experiment for free.
You will receive a $75 coupon to create your first Google survey if you have an active Google Analytics account. There is no red tape to get your coupon, so don’t be afraid to apply for the credit when you are ready to get started.
Like most people, I was excited when I received my first $75 credit for the Google Consumer Surveys. I assumed it was similar to the $100 AdWords credits I have used many times in the past, but this credit has some limitations.
First, you can only apply one $75 credit to each Google survey campaign. This would not be a big issue in an AdWords account, because I could stop my campaigns before they cost more than my coupon covered. Sadly, you are not allowed to pause campaigns in the Google Consumer Survey tool. You pay upfront the total cost of running each campaign. The cheapest campaign I could run was $100.
I understand the reason Google does not want us to stop surveys early. It makes sense to pass a threshold of responses before making an accurate decision. I also understand that Google does not want to give its new service away for free, yet I see a big problem: It does not allow you to stop a survey that is not running as expected.
Because you can’t stop your surveys when you see a problem, you need to make sure your survey has no flaws when it is launched. A single question survey with no audience segmentation costs $100. Many of the survey templates cost $800 or more. Here at Integrity, we have created surveys that cost more than $1,200 to run.
2. Test, test, test. Do your own tests before you accidentally throw your money away on useless data.
Just like you QA test all the links and forms on a webpage before launching, make sure to QA test your questions.
The first step is to make sure everyone knows what you are talking about. This is sometimes called the Homer Simpson rule. You have to create questions you think Homer would have no problem understanding. If anyone in your office is even slightly confused by a question, change it.
It is best to run a test on a smaller, cheaper scale. One of the best ways to do this is to run a controlled test on Survey Monkey. This site has free and low-cost options for running surveys. Send this survey to your coworkers, your kids and your grandmother. They can help you work out the kinks before you spend hundreds of dollars with Google. Survey Monkey also has options to survey random Internet users, too.
Don’t forget to explore all the different survey templates and questions before you even create the goals of your survey. The templates include basic questions for branding, brand usage, brand attitudes and perceptions, purchase behavior, ad effectiveness, ad testing, and consumer satisfaction. The further your goals stray from the templates, the more testing you need to do. It’s possible to expand the usefulness of this tool, but proceeded with caution.
3. Beware the Internet population.
It is tempting to purchase the 10 cent response survey option in Google Consumer Surveys. These 10 cent survey responses don’t allow you to segment your audience in any way, but they give you the cheapest options for campaign budgets, starting at just $100 a campaign. Yet, this is the quickest way to get useless information.
These responses will go out to anyone, so you must have questions and answers that can be answered by every possible person. You can’t ask the question “Do you ride a bike or walk to work?” without allowing a “Neither” response.
If you had created a custom audience using the screening question “Do you drive or ride a bike to your job?,” you would not pay for any responses from users that said no. In the 10 cent per response survey, you will pay for users that said no. Many users will say no, even if that is not true. You will still get good data from the 10 cent per response results, but we’ve found the segmented audience responses are much better.
4. Bad questions equal bad responses.
Don’t be fooled into thinking this new tool from Google will make consumer research experts obsolete. If anything, it will make you appreciate the people in your office, or the agency you work with, even more. It can be hard to write questions that are effective, and it is even harder to analyze the data and create an effective action plan from that data.
There is no way I could teach you how to create actionable data from the Google Consumer Survey tool in this simple post, but I can direct you to some great resources to get you started. To get the most of our your surveys, you can start with the following resources:
The latest from Integrity
What is Omnichannel Marketing?
Corporations often ask us about omnichannel, integrated and multichannel marketing. What do the terms mean, and what do they mea...