marketing: survey and analysing qualitative data

why is it marketing research important?

it helps understand who buys your products, why, when and under which circumstances
understanding consumer insights, motivations, barriers and explicit or latent desires/needs
to pre or post test concepts, products and campaigns

what are the steps of market research?

1. survey design
   direct or indirect? quantitative or qualitative or experimental?

2. procedure
   questionnaire? guide? sample?

3. analysis
   how results are analysed

quantitative vs qualitative vs experimental

quantitative studies
describing trends, attitudes, behaviors
potential problem: representativeness
methods: online platforms and tools (Typeform, SurveyMonkey)

qualitative studies
understanding beliefs, cultures, implicit feelings
potential problem: biases
methods: focus groups, individual meetings, ethnography (observation), netnography (online)

experimental studies
understanding causal links, explaining phenomena
potential problems: enviroment and stimuli need to be controlled
methods: field experiments and lab experiments
results can be qualitative or quantitative

info about quantitative methods

aim to describe, quantify, predict

results are based on larger sample sizes → representative, macro view
data in form of numbers and statistics, charts, tables etc

as tools, quantitative questionnaires are usually used (closed questions)

which method should be chosen? (quantitative/quali/experimental)

ex: plus size fashion brand wants to identify characteristics of clothes to appeal to customers → mixed methods
- qualitative insights: understanding of customer needs, what drives behavior of consumers
- quantitative data: prevalence of characteristics on broader base to ensure product strategy is data-driven & representative

ex: study to understand path taken by Louvre visitors
→ observational research (ethnographic, qualitative)
using trained researchers to observe or heat mapping
identifying common patterns in visitor flow

ex: find purchase drivers of Birkin bag → mixed methods
- qualitative research, exploratory phase
interviews, ethnographic research, focus groups
Birkin bags are not just luxury but also status symbols so perception is important
- quantitative research, confirmatory phase
surveys can quantify importance of different purchase drivers after deep insights gained from qualitative phase

indirect methods

website reviews, social networking, forums

observation

usage, inferred data (quali/quanti)
from shopping assistants: virtual assistant, recommendations systems
from social media and chatbots
from behavioural data: purchase history, loyalty cards, cookies, delivery apps
user generated content: reviews etc

projective methods
builds meaning indirectly
construction techniques (movie in which the Orangina brand is the hero)
association techniques (first words that come to mind when I tell you Monoprix)
completion techniques (the people who watch at Arte are…)
expressive techniques (what if Ford was an animal/country/character?)

direct methods

text/SMS surveys
direct phone calls
surveys

direct vs indirect

direct: straightforward, conscious responses
indirect: digging deeper, unconscious opinions

limitations in research

limited cognitive abilities (verbalisation)

social desirability bias ‘
wanted to appear better, responding according to hypothesis
can be avoided with single blind design: subject doesn’t know purpose of experiment until after

the procedure for market analysis

qualitative methods: for interview guides follow logical order for the questions

quantitative methods: different scales to use, scales give you reliable data because it‘s all collected in the same way

different scales for quantitative methods

Likert Scale
1-5 with different levels of agreement, easy, statistical analysis

Semantic Differential
scale with adjectives (fun - frustrating), nuanced insights, brand positioning

Integer Scale
discrete numerical values like 1-10, precise, simple to read

Continuous Rating Scale
sliding scale, fine-grained data, detailed analysis

what to pay attention to in questionnaires

response options
shouldn’t be only agree, neither, disagree

clear questions
no vague statements
statements shouldn’t address two topics, only one at a time

open ended feedback
where respondents can elaborate for richer insights

what is a sample

a specific group chosen out of the general population to be studied

can a quantitative study be valid with a non-representative sample?

yes

when study wants to understand individual causes
when it compares groups under different conditions
in exploratory studies when goals is to get insights even if the findings might not apply to general populaion

can a quantitative study be invalid even with a representative sample?

yes

if it only focuses on having a large sample but ignores other factors
direct methods (survey etc) have limitations like leading questions or bias
if study doesn’t account for biases than can affect results and make findings unreliable

which sampling methods exist?

1. probability sampling
   everyone has equal chance of being selected
   random, stratified, systematic sampling
   statistically representative results → findings can be generalised

2. non probability sampling
   selection is based on convenience or judgment
   convenience, quota, snowball sampling
   quick insights
   for limited resources
   if generalization is not required

what is selection bias?

when the group chosen for a study doesn’t represent the whole population

makes results unreliable because the sample isn’t diverse enough to reflect broader audience

how to do qualitative data analysis

organise collected data
transcribe, translate, label

identify framework
the plan to structure, label and define data

sorting data into framework

descriptive analysis based on framework

second order analysis
identifying recurrent themes/patterns in data