Identify and survey your audience, from asking screening questions to estimating your sample size and incidence rate.
Let’s say you want to start an online subscription-based tea delivery service. You’ve got lots of questions, like:
Is there a market for my tea business?
Who would be most likely to buy?
How much are they willing to pay?
Where are they primarily located?
The first step to reach your target market is to send out a survey to a balanced sample of the general population. When you send a survey to our Audience panel, the following demographic segmentation information is automatically included in your survey results in the US:
Tip: When surveying a representative sample of the general population, make sure Age and Gender Census balancing is turned on!
Also, be sure to ask any additional demographic questions or behavioral questions that are relevant for your category. To reach the target market for your tea business, you’ll probably need to know answers to questions like:
How often do you drink hot or iced tea?
How often do you shop online?
How interested would you be in a tea delivery service?
Once you have your results, apply a filter to see which segment of the general population is most likely to be a good fit for your subscription tea business (e.g., people who frequently drink tea at home and shop online).
From here, you can get a great sense of who your target market is: the gender split, how old they are, where they shop, and the size of the market compared to the general population. Now you’re ready to develop your business idea even further!
After you identify your audience, you need to determine how many people are in your population. Let’s say you discover through a general population survey, the people most interested in a tea-delivery service in the U.S. are women between the ages of 30 and 45 who drink tea weekly.
Based on the 2010 U.S. Census data, that’s approximately 30 million people in your target market. Even though that seems like a lot, there’s an easy way to survey far fewer people for a representation of your target market.
Use our sample size calculator to figure out the number of people you should survey to get an accurate understanding of your target market.
For example, if you want a representative sample of your 30 million people, with a 95% confidence level (the probability that your sample accurately represents your population) and +/- 5% margin of error (the range around your survey results the actual results of your population would most likely fall into), you’d have to send your survey to ~400 women between the ages of 30 and 45. The more people you sample, the closer your results will mimic the actual population.
Sample size | Margin of error |
50 | +/- 14% |
100 | +/- 10% |
150 | +/- 8% |
250 | +/- 6% |
400 | +/- 5% |
600 | +/- 4% |
1,100 | +/- 3% |
2,400 | +/- 2% |
Margin of error calculations above are based on a population size of 100,000+ and a 95% confidence level.
How you will be using your results may also impact the same size you need. For example, let’s say you need to segment your results by region or income level or age groups. Once you start looking at cuts of your results at a granular level, the margin of error will increase for those smaller groups. So, you’ll need to survey more people in total.
Tip: If you're trying to publish results, most reputable publications use a rule of thumb of collecting at least 1,000 responses or a +/- 3% margin of error.
Of course, it isn't always possible to get 1,000 people for your survey. You might be up against budget or time constraints—or it might be difficult to find the exact people you need in an online panel.
As a professional in the real world, you’ll need to strike a balance between what’s best in theory and what’s practical. And sometimes it's more practical to survey 150 people for the business decision you need to make today.
There are two ways to reach the people you need to take your survey:
When people join our panel, we ask them questions to learn more about them and build out their profile. We ask about basic demographics like gender, age, location, and income level as well as behavioral questions in categories like product ownership/usage and shopping habits.
You can target people for your survey by selecting from these pre-profiled targeting options. Check out our Audience preview tool to see what we have available.
If you don’t see the targeting option you need, it’s possible we haven't already asked our panelists that question. In this case, you can ask your own screening question and disqualify respondents who don’t meet your targeting criteria.
Tip: You can use multiple screening questions in the same survey to target a specific group of people who meet multiple criteria.
You should use screening questions when you need to reach a specific group of people you can’t find using just our available targeting options.
Carrying on with our tea delivery service example, you’ve determined that the target market is women between the ages of 30 and 45 who drink tea at least once a week.
In this case, you can use targeting options to find women between the ages of 30 and 45. But you’ll also need to use a custom screening question to make sure only women who drink tea weekly (or more frequently) take your survey.
At the beginning of your survey, ask a question like:
How often do you drink either hot or iced tea?
Next, you’ll want to apply disqualification logic to the “Monthly,” “Yearly or less,” and “Never” answer options.
Once you launch your survey, any respondent that chooses “Monthly,” “Yearly or less,” or “Never” will skip over the survey and land on a page that thanks them for taking your survey—or will be routed to another survey they qualify for.
There’s a whole lot more to writing screening questions in order to get the panel of respondents you need. Get our latest tips and tricks in, “A guide to using screening questions in your survey."
The incidence rate is the percentage of respondents that will qualify for your survey after answering your screening questions.
When your survey contains a screening question, you’ll need to provide an estimated incidence rate so we can manage the number of survey invitations to send our panelists.
Here are some techniques to help you estimate the incidence rate for your survey:
It’s possible that your company has relevant data or has done some research on your topic already. For instance, if you have an established brick and mortar tea shop, you might already have information about your current customers to inform the research for the new online subscription business. Try asking your colleagues if they know who’d have access to data like this.
If you're surveying people in a popular or well-researched category, there may be useful industry evidence documented online.
For example, you could search “How many people drink tea in the U.S.?” and find a ton of articles citing research that’s already been done.
Similar to the survey we used to initially identify your target market, you can send a quick survey to the general population to determine the incidence rate of your target market. Here’s how:
It’s never been easier to reach your target market—all you have to do is ask.
Get real-time feedback from around the world in minutes.