Data Collection Methodology: Visitor Survey
Ideally, Visitor Surveys should be conducted on the same days as Visitor Counts. This makes it easier to know if the market collected a representative number of responses based on the attendance that day. If it’s not possible to complete both the Visitor Count and Visitor Surveys on the same day, then surveying the week following a Visitor Count is the next best approach. This will allow for a closer relationship between the attendance and the number of surveys. The method used in FM Metrics is to collect survey data for at least 2 days but preferably over 4 days in one season. A market season is usually around 20-25 weeks. If you want to collect data for a year round market, you would conduct data collection on two sets of days in the summer/fall season and then in the winter/spring season, for a total of 4-8 data collection days over the year.
Step 1: Identify Sample Size, Selection Interval and Number of Data Collectors
The term sample size means the number of surveys needed to provide a reasonably representative amount of responses. The sample sizes listed in the table below will provide a 90% confidence level, with 5% margin of error which is a level that most researchers would find adequate. Try to reach your minimum sample size over the course of multiple market days, so you reach the widest variety of visitors but be careful to not choose special days or go past the season that you are measuring.
Visitors who volunteer to take the survey, but aren’t in the selection interval should be allowed to take the survey, to support the desire for participation. The volunteered surveys can be separated out when entering the data (be sure to choose “volunteer” on these surveys). If multiple people in one group want to take the survey, allow if they are each from a separate household, but step a few feet away to then survey them one by one.
To account for personal biases, data collectors should follow a systematic approach to selecting people to be interviewed. Depending on the sample size required at your market, a corresponding selection interval is provided. For example, if the visitor survey selection interval is 4, the data collector will survey every 4th adult that crosses the survey line. The table below gives recommended interval period and survey counts for a season.
Step 2: Map the Survey Line
Identify locations in the market where you can administer the survey. Near the market info booth, or a main entrance typically works well. Establish (or chalk) a line near the survey area. The data collectors will count the adults that walk past the line and ask every Nth adult to take the survey. If possible, have a table for shoppers to put down their items, or an umbrella for cover if the day is warm or misty.
This is an example market map, identifying primary entrances and counting zones.
Step 3: Prepare Materials
If you’re using the online survey, ensure that all collectors have the URL up on their phone or tablet and that you have access to an internet connection.
If you’re using the paper format, customize the Visitor Survey template with your market name, the date of the survey, the target number of surveys to be completed and the selection interval that data collectors should use. Print enough surveys to fulfill your goal, as well as plenty of extras. Gather enough clipboards for your data collectors, pens and pencils.
Tips for a Successful Survey
Conduct a few role plays with collectors before they begin so they can be confident from the start in their surveying. Positive energy makes surveying people at markets a fun job and increases their effectiveness.
Whatever the target number of surveys is, spread the number evenly across the hours the market is open. If all surveys are collected in the first hour, the answers will skew to a specific demographic and may not represent the entire spectrum of shoppers that the market attracts.
Whether the surveyors approach people exiting or entering the market is really based on the type of questions being asked. If the questions are about their shopping that day (as in how much they spent), try to catch people as they leave. If the market is asking only questions about their mode of transportation or zip codes, then it doesn’t matter at what point the survey is taken. However, market shoppers are often nervous about their favorite products being sold out before they arrive, so more people may brush the request off on their way in.
Show the market map to your data collectors. Identify where they will administer the survey, how they will execute the selection interval (i.e., count adults that cross the ‘survey line’).
If possible, ask one of the data collectors to manage interval selection; they’ll approach every Nth market visitor to cross the survey line, ask that visitor if they’re willing to take a short survey, and direct them to a data collector with the surveys. This is especially important for larger markets.
Provide the data collectors with their materials, and go over the following instructions:
- Ask every Nth adult visitor who crosses the survey line to take a survey. A good opener is “Do you have a minute to help the market?”
- Ask the survey questions verbally, and record the answers on the survey form.
- If a visitor outside of the interval count volunteers to take the survey, circle ‘volunteer,’ on the survey. If a visitor does not want to complete the survey, asking them to participate does not count towards your minimum sample size. Identify the next respondent through the selection interval. In other words, don’t just ask the next adult to complete the survey.
Follow up with data collectors throughout the day to answer questions and hear about their experiences.
At the end of the market day, gather your data collectors together to review your success.
- What worked and what didn’t?
- Did shoppers respond better to one request over another?
- Were there areas that you felt visitors were not getting surveyed?
Discussing these questions while they’re fresh in your mind will better prepare you for your next visitor count.
After the market day is over, if you’ve collected paper-format surveys, be sure to input those into the FM Metrics website.
- It is most markets’ experience that visitors are usually very willing to stop and answer a few questions for the market. That is why we encourage surveyors to say ”do you have a minute to help the market?” Or “can you take a minute to share your opinion with the market?”
- Expect that people will often start a response by saying “I don’t know’ as a response to some of the questions. Be patient and let them think for a minute or press them by saying ‘Take your best guess” rather than marking it as DK or suggesting answers (usually a no-no in survey work!)
- Writing responses as ranges is also a no- no for FMM; if someone offers one, surveyors should press for a specific number within that range. The point is that sooner or later someone will need to pick an actual number and it is better that it is the respondent rather than the surveyor or the data entry person!
- Some surveyors will hit the ground running, while others will work more slowly. If a surveyor is doing either fewer or more surveys than others by a significant amount, observe them for a few minutes to see why that is the case. In the case of moving lightning fast, it may be that the surveyor is ignoring the interval process or moving too quickly through the survey. If they are lagging behind, they may need to pair up for a few minutes with another person that is doing well or even have a few role plays run through again or be moved to a different location.
- Some hours are often slower than others, so hitting the goals for the number of surveys completed may take longer. If the market is setting a goal for each survey or each hour, set a range of “okay” to “great” in terms of how many they collect. If a surveyor hits their “great” goal quickly, have them find the person who is managing the collection day to decide what to do next. The solution may be to move them another place in the market (helping someone who is struggling). move to another task (such as helping with the visitor count) or to take a break, until the start of the next hour. The point is to not end the day with most of the surveys all gathered in just a short span of the market day, but instead to spread it out in order to get a good cross-section of responses.