BBM016 Consumer Insights PDF

BBM016 Consumer Insights Janina Steinmetz Today Quantitative sampling Data analysis Recap: Connecting research objectives to primary research Example: One of Lizzie’s sub-objectives is to identify the best promotion channels for her vegan cake food truck For this objective, decide which method t...

BBM016 Consumer Insights Janina Steinmetz Today Quantitative sampling Data analysis Recap: Connecting research objectives to primary research Example: One of Lizzie’s sub-objectives is to identify the best promotion channels for her vegan cake food truck For this objective, decide which method to use (survey, interview, secondary research) If survey or interview Decide which variables to collect (e.g., social media use, interest in coupons/promotions, trend following, following food influences, etc.) Add questions that capture these variables Sampling What is sampling and why do we need it in marketing research? How do we select a sample? Introduce steps of the sampling process Understand the different sampling techniques and conclusions that one can derive from them Sampling gone wrong… Marriage survey in the 1970s Of the women who sent in their questionnaire, 98% were unhappy with their marriage and 75% had an extramarital affair Sampling gone wrong… However, in the sample, only 4% of those who had been contacted responded When surveying the whole population, actually only 7% had an affair Most women were happy, but the unhappy ones responded to the survey Sampling The process of using a small number of items or parts of the population to make conclusions about the whole population! Population of the study is a complete group of entities that share some common set of characteristics Example: all married women in the UK; all vegans; all college students, etc. Sample selection Define the population Select Sampling Technique(s) Execute the Sampling Process Determine the Sample Size Define the population Who do we want to study? Exclusions are part of the definition E.g., population = students in London; excludes students in other cities Always describe your population in terms of its characteristics = sampling criteria Simple sample definition increases the incidence (i.e., the larger the number of units in the sample) The more exclusions, the more difficult to sample Focus on psychographic variables! Example Population: all male skin care users in London who have purchased cruelty-free products in the last 12 months Where do you find these people? Surveying all of them would be very expensive, so we draw a sample Example Population: all male skin care users in London who have purchased cruelty-free products in the last 12 months What are their psychographic characteristics? Like to treat themselves Open to new products … Where do you find people who are like that? Two sampling techniques Non-probability Sampling: ▪ Probability Sampling: Subjective sampling ▪ An objective sampling techniques that relies on the procedure by which units are personal judgment of the drawn randomly researcher ▪ Representative samples People who are available (e.g., who walk by; who are on social media) Results are not representative Probability sample Requires list of addresses or phone numbers to randomly select people from Characteristics of probability sampling Respondents are selected through probability techniques Best way to establish representativeness Best for studies that: Are quantitative Require large sample size Have large budgets Have sampling frame Have heterogeneous populations Non-probability sample In the marriage survey, a non-probability sample was used That’s why not all women responded, but only the unhappy ones In fact, most samples today are non-probability samples because phone books or address lists are becoming less and less accurate and complete That doesn’t need to be a bad thing! Characteristics of non-probability sampling Researcher selects the respondents No representativeness! Best for: Qualitative studies Studies where population is fairly homogenous Low budget studies Non-probability sampling: Convenience Convenience sampling attempts to obtain a sample of convenient respondents. Often, respondents are selected because they happen to be in the right place at right time University students People on the street Questionnaires on website Non-probability sampling: snowball An initial group of respondents is selected After being interviewed, these respondents are asked to identify others who belong to the target population of interest. Subsequent respondents are selected based on the referrals (e.g., one marketing student will know another marketing student) Good for locating the desired characteristic in the population estimating characteristics that are rare in the population (e.g., vegan tattoo artists) Sample size We ask you to recruit 100 survey participants and 10 interview respondents These are based on industry-standard rules of thumb How to sample Lizzie has done her secondary research to identify an initial target segment: Adventurous Angela: Loves to try new food Is health- and eco-conscious Is willing to spend money on treats that are special Lizzie wonders where she can find Adventurous Angela’s for her survey and interviews? In your IMP Tell us how you found people from your target segment Document which channels you used to distribute your survey and recruit interviewees Convince us that these are probably people very similar to your target segment Bad example: If Adventurous Angela is your target segment, recruiting Lidl shoppers is probably not the best idea What would be better? Qualitative sampling Qualitative samples are not representative! That’s ok The important thing is that your qualitative sample comes from your target population (i.e., your target segment) Just because the sample won’t be representative doesn’t mean you can interview your grandma (unless your grandma is in your target segment) Connecting sampling to data https://www.youtube.com/watch?v=sxYrzzy3cq8 Quantitative data analysis Step 1: Download your data In your Qualtrics survey, go to Data & Analysis Make sure to export as numeric values Step 2: Look at your data Step 3: Validating data Some informants will respond so quickly (< half the average time) or left all questions blank that you might think they responded carelessly You can then exclude their answers, but describe this is your report Only exclude someone when you really think their answers make no sense to include Be very conservative! Don’t delete them: you can create a new variable that is called exclude, and define 1 for those you want to keep and 0 for those you want to exclude Then select exclude = 1 for the analyses Step 4: Coding quantitative data Assigning values to answers (should have been done by Qualtrics) E.g., 1 = female, 2 = male, 3 = other For ordinal scales, the numbers should have meaning (should have been done by Qualtrics; double-check) 1 = no education, 2 = some education, etc. If you have questions with multiple responses, you need dummy coding E.g., which restaurant have you visited this month? Eat, Pod, Itsu, Pret a manger? Dummy for each, 1 = Eat visited, 0 = Eat not visited Step 5: Coding quantitative data For Likert scales, graphical rating scales, and semantic differentials, Qualtrics will give you values Be careful with how values are coded! Step 6: Descriptive statistics Descriptive statistics provide simple summaries about the sample and about the observations that have been made. Include the numbers, tables, charts, and graphs used to describe, organize, summarize, and present raw data. Frequency distribution: Measures of central tendency Mean/average Median Mode Measures of dispersion Range Standard Deviation Frequencies Charts Bar chart Pie chart Central Tendency: Mean Central Tendency: Median Central Tendency: Mode Dispersion: Range Dispersion: Standard deviation Analysis in Excel Cross-Tabulation Cross-Tabulations are tables that reflect the joint distribution of two (or more) variables with a limited number of categories or distinct values. help to understand how one variable (e.g., brand loyalty) relates to another variable (e.g., gender) a cross-tabulation table contains a cell for every combination of categories of two (or more) variables Examples: How many brand-loyal users are males? Is product use (heavy users, medium users, light users, and nonusers) related to outdoor activities (high, medium and low)? Is familiarity with a new product related to age and education levels? Is product ownership related to income (height, medium, and low)? Cross-Tabulations Cross-Tabulations Recap: 5 people: 1, 2, 2, 5, 100 cups of coffee a day What is the mean? The median? The mode? For your IMP Frequency tables/charts, descriptive statistics and cross tabulations are simple but effective methods to observe basic characteristics of the sample and to explore associations between two or more variables You do not need to report test statistics (e.g., p-values, etc.) Means, standard deviations, frequencies, and ranges are sufficient! Interview guide: getting started https://www.youtube.com/watch?v=8z8XV1S7548 Open-ended questions Open-ended: What is important for you when you buy cosmetics? Close-ended: Do you consider price important when buying cosmetics? Yes versus no An interview is not a conversation! “Normal” conversation Interview (purposeful conversation) Both parties have equal time to talk Interviewee does most of the talking Topic shifts are common Topic stays focused on one thing at a time Both parties control the direction of Interviewer controls the direction of the talk the talk Both parties share their feelings Interviewer need to suppress feelings What do you want? Descriptions of situations, perceptions and definitions of terms,.. Opinions, reactions to situations, feelings, evaluations, judgments, market pain Relations of these to opinions of others Predictions of the future Preparing and using the Guide DO DON’T Develop a guide based on ▪ Read from the guide! variables; cover the central topics! ▪ Ploddingly adhere to the guide. The less you know in advance, ▪ Turn the interview into an open- the more detailed the interview ended survey questionnaire! guide! Concentrate on informant, not the guide. Use it as a checklist at the end of the interview. Doing the Interview 1. Choose a setting with the least distraction. 2. Give the informant the participant information sheet and wait for them to read. 3. Allow interviewee to clarify any doubts about the interview. 4. Prepare a method for recording data, e.g., phone 5. Record their consent. 6. Start. During the Interview Occasionally verify the phone is still recording Ask one question at a time. Attempt to remain as neutral as possible. Encourage responses. Be careful about the appearance when note taking. Provide transition between major topics. Don’t lose control of the interview. Manage transitions. Don’t shape the answers or talk about yourself! Evaluating the Interview YOU PROBABLY HAD A GOOD INTERVIEW IF ✓It felt like time went by quickly. ✓Covered most or all the topics in your guide. ✓You were able to follow up on interesting points made by informant. ✓Informant talked for several minutes several times. ✓You felt that you learned a lot from the interview. In your IMP Start with your initial target segment Document which channels you used to distribute your survey and recruit interviewees Where can these people be found, with a focus on the psychographic characteristics? End of Session 5 Prepare for next session Take the Session 5 Quiz on Moodle! Draft your survey and think about your sample