Latest / The Galaxy of Marketing / Coordinates: Marketing research

Transcript

Coordinates: Marketing research · The Galaxy of Marketing

• 0:00
  People are difficult creatures. We always want
• 0:02
  something and demand companies to know exactly
• 0:05
  what we want without even knowing it ourselves.
• 0:08
  And that mostly unpredictable and chaotic behavior
• 0:11
  is the pain in the neck of any company. But marketing
• 0:14
  has got a perfect instrument to offer a solution.
• 0:18
  Hello, marketing explorers! This is Captain K,
• 0:21
  marketing major and your personal marketing guide.
• 0:23
  It's been a long assessment time where I get
• 0:26
  the opportunity to conduct a research. So in
• 0:29
  this episode of the Galaxy of Marketing, we are
• 0:31
  going to delve into marketing research. An exact
• 0:34
  thing that makes marketing seem like a magic
• 0:36
  and evokes how -do -they -know reaction. We will
• 0:40
  cover why it is done, how to conduct it, implement
• 0:43
  it and analyze results. Ready for some brain
• 0:47
  work? Well, coordinates set, so prepare for the
• 0:50
  jump. By definition, marketing research is the
• 0:54
  process of gathering, analyzing, and interpreting
• 0:56
  data about consumers, competitors, and industry
• 0:59
  environment. It is often mixed up with market
• 1:02
  research, which is a separated analysis of the
• 1:05
  whole market, like PESTLE analysis. And marketing
• 1:08
  research always comes after the market research.
• 1:13
  To conduct effective marketing research, companies
• 1:16
  follow a six -stage framework that I call PP
• 1:19
  -DASA. Problem. Plan. data collection. analysis.
• 1:23
  strategy and action now let's talk in details
• 1:27
  about each stage. any research begins with a good
• 1:30
  research question in our context also known as
• 1:33
  the problem it should be very specific because
• 1:36
  it defines the very information that will be
• 1:39
  gathered most typical questions are why our sales
• 1:43
  are dropping or will consumers like this new
• 1:46
  product But there are also more complex themes
• 1:49
  that might be explored. For example, the use
• 1:52
  of AI in particular industry. For example, I'm
• 1:55
  now conducting my own research and the research
• 1:57
  questions evolve from perception of AI ads in
• 2:00
  Kazakhstan to the perception of AI -generated
• 2:04
  advertising of non -luxury clothing brands among
• 2:07
  Gen Z in Kazakhstan. The second version clearly
• 2:10
  defines the sample and the category of fat that
• 2:13
  will be analyzed, which is important when it
• 2:15
  comes to data interpretation. Because everything
• 2:18
  for everyone approach doesn't work here. What
• 2:21
  works for one may not work for another, and vice
• 2:24
  versa. It will get the researchers overwhelmed
• 2:26
  and take too much time to process. Next stage
• 2:32
  is developing the research plan. And this one
• 2:35
  is much more difficult than you think. Because
• 2:37
  before deciding what method to use, company needs
• 2:40
  to identify what type of data they need. Quantitative
• 2:43
  or qualitative? Or both? And whom are they going
• 2:47
  to ask for it? Have you ever seen these small
• 2:49
  surveys for buttons like rate these products
• 2:52
  from 1 to 5 or choose whether you agree or disagree?
• 2:55
  That is quantitative research. It deals with
• 2:58
  statistics and numbers to reveal patterns and
• 3:01
  predict trends. There are four main ways to conduct
• 3:04
  it. Survey, experiment, structured observation,
• 3:08
  which is basically observing people in their
• 3:10
  natural environment, and secondary data analysis.
• 3:14
  The huge benefit of quantitative research for
• 3:16
  marketing is that it is relatively easy and quick
• 3:18
  to gather, unless it is a worldwide research.
• 3:22
  However, it lacks emotional depth. For example,
• 3:25
  if we use standard scale strongly agree or strongly
• 3:28
  disagree, we might see that mostly people agree
• 3:31
  with something, but we won't know why they agree
• 3:35
  with it. To answer that, we have qualitative
• 3:38
  research that deals with feelings, opinions,
• 3:41
  and worldviews. It's typically conducted for
• 3:43
  three reasons. First, explore poorly understood
• 3:46
  topics and discover new variables. Second, develop
• 3:50
  new theories and frameworks based on collected
• 3:52
  data or also known as grounded data. And the
• 3:56
  third is understand the perspective of participants
• 3:59
  on the object of the research. Most common method
• 4:02
  is a long detailed interview, but you also can
• 4:05
  use narratives and live descriptions. Also, you
• 4:08
  can conduct case studies and ethnographical researches,
• 4:11
  which is also some kind of structured observation.
• 4:14
  The problem with qualitative research, though,
• 4:16
  is that it is very difficult to conduct, because
• 4:19
  you need individuals to speak with you for about
• 4:21
  an hour, and trust me, it's not so easy to find
• 4:24
  such people. There's one more important thing
• 4:27
  that we do before collecting the data, is checking
• 4:29
  whether questions or actions are ethical and
• 4:32
  unbiased. To do so, researchers should write
• 4:35
  down their own biases, check that they ask participants
• 4:38
  exactly what they signed up for, and ensure that
• 4:41
  there is a warming structure. It means that researcher
• 4:44
  is starting with the most simple and generalized
• 4:46
  questions like Do you like AI tools? to sensitive
• 4:50
  ones Why do you like or dislike AI influencers?
• 4:54
  Now we finally get to the data collection part
• 4:57
  where large amounts of raw data are gathered,
• 5:00
  transcribed if necessary, and organized It relatively
• 5:03
  quickly transitions into the first phase where
• 5:06
  our ship is gradually falling into the giant
• 5:09
  scary black hole called Microsoft Excel Scary,
• 5:13
  right? It breaks down data and spits patterns,
• 5:16
  correlations and statistical data at us, provided
• 5:20
  that we know which buttons to press, of course.
• 5:22
  The momentum is gained and our ship is thrown
• 5:25
  at the strategy stage, because when we have all
• 5:27
  the data we need, we can finally develop strategy
• 5:30
  to solve the problem. In why our sales are dropping
• 5:33
  example, our genius minds discovered that people
• 5:36
  consider that company products are badly designed
• 5:39
  and are hard to find because people quit large
• 5:42
  malls and prefer small shops. So company needs
• 5:45
  to find new sales channels in customer preferred
• 5:48
  places and make packaging bright and more sophisticated.
• 5:53
  On the action stage, this all comes to life and
• 5:56
  if everything goes right, customers become satisfied
• 5:58
  and sales increase. Well, and if not, the marketing
• 6:03
  research begins again. If you think that we are
• 6:06
  finished, no, we didn't. These were the core
• 6:09
  processes, and since I am very meticulous when
• 6:11
  it comes to show you the full picture, we are
• 6:14
  moving to nuances, starting from qualitative
• 6:16
  research. Nuance 1. How to choose what customers
• 6:21
  to research. It is also called sampling. To choose
• 6:24
  the right sample, you need to understand what
• 6:26
  exact question you are trying to answer. Let
• 6:29
  us get back to my example. I conducted research
• 6:31
  about perception of AI ads for non -luxury clothing
• 6:35
  brands. What method could I choose? It could
• 6:38
  be random sampling. I don't need to explain what
• 6:41
  is random, right? Also convenient sampling. It
• 6:45
  is the choice of any available for interview
• 6:47
  customer, but it is risky because the data obtained
• 6:50
  this way may be biased. Another method is a snowball
• 6:54
  sampling. It's basically asking a person who
• 6:56
  already participated in the interview to recommend
• 6:59
  someone else to participate in it. Last but not
• 7:02
  least is cluster sampling. You can divide people
• 7:05
  into groups based on certain features, like age,
• 7:08
  gender, purchasing behavior, etc. Which clusters
• 7:12
  are to be analyzed is defined by the research
• 7:15
  question, but usually it is just one. However,
• 7:18
  if you need to question two or more, mind the
• 7:21
  ratio gap. To make it simple, imagine that you
• 7:24
  analyze university students. 60 % of them are
• 7:27
  studying accounting, 20 % IT, and remaining 20
• 7:30
  % are designers. For your data to be representative,
• 7:34
  you need to keep that ratio among participants
• 7:36
  the same, like 60 students from accounting, 20
• 7:39
  from IT, and 20 from design. Now, what was the
• 7:43
  name of that method? Yeah, short attention span.
• 7:46
  Oh, oops, I meant cluster sampling. Moving on.
• 7:50
  Nuance 2. How to make it ethical? First, do the
• 7:55
  preparation. We are not the exam sheet to speed
• 7:57
  questions on people without preparing them first.
• 8:00
  As I said before, people need to understand clearly
• 8:03
  what they are going to be asked, without any
• 8:05
  surprises. So the researcher should only collect
• 8:08
  the data that is genuinely necessary, especially
• 8:11
  when it comes to private information. And if
• 8:14
  it's a face -to -face interview, do some warm
• 8:16
  -up small talk to show the good attitude. Or
• 8:18
  at least pretend to have one. Check for biases.
• 8:22
  Questionary absolutely must not contain leading
• 8:25
  and manipulative questions that make people more
• 8:28
  likely to choose one option over another or simply
• 8:31
  do not represent the whole sets of opinions.
• 8:33
  The drastic example is something like how harmful
• 8:36
  the AI is, how negative AI influencers are. The
• 8:41
  better wording would be like this. What do you
• 8:43
  think about the use of AI? What is your opinion
• 8:46
  on the existence of AI influencers? Neutrality
• 8:50
  is your best friend. Besides, you need to leave
• 8:53
  some room for unexpected answers. And the third
• 8:56
  thing is avoid disturbing questions or make them
• 9:00
  easier to answer. Like if there is no need in
• 9:03
  asking what is the person income level or household
• 9:05
  type, don't ask it. But if you really need to,
• 9:08
  then make sure there are some categories, like
• 9:11
  from 200 ,000 Tenge to 300 ,000 Tenge. that is
• 9:15
  the category and this way person will not feel
• 9:18
  that their personal sacred data is collected
• 9:20
  nuance three context matters tone atmosphere
• 9:26
  emotions voice even the body language all of
• 9:29
  that matters when you interpret data because
• 9:32
  two people may give you a single sentence like
• 9:34
  i agree but it could be yeah i agree it's super
• 9:38
  cool and the other could be well yeah i agree
• 9:42
  same text completely different perception and
• 9:45
  interpretation. That is why doing interviews,
• 9:48
  experts make actual notes of the participants'
• 9:51
  behavior and their own reactions and thoughts.
• 9:54
  Now, done with feelings, we are changing the
• 9:56
  course to statistics. Quantitative research is
• 9:59
  relatively easier to do, but it bears two huge
• 10:02
  nuances. First, how do we measure it? One of
• 10:07
  my research questions was, are you confident
• 10:09
  in the result that you get from AI? Not that
• 10:12
  easy to answer, huh? But that is where record
• 10:15
  scale comes to help. Very confident, confident,
• 10:18
  slightly confident and not confident are perfect
• 10:20
  to group people and see which ratio is bigger.
• 10:23
  But the second nuance is much more complicated.
• 10:27
  It is how do we understand the difference between
• 10:30
  these gaps? Because again, it could be, yeah,
• 10:33
  I'm confident, AI is a good thing, it is great.
• 10:36
  And it can also be, well, yeah, generally I agree,
• 10:39
  but there are some issues. That is why companies
• 10:42
  prefer a mixed approach, where they first see
• 10:44
  the general dynamics with a quantitative research
• 10:47
  and then go in -depth with qualitative research.
• 10:50
  Our ship is finishing this trip. Thank you for
• 10:53
  choosing the Galaxy of Marketing. We will hear
• 10:55
  again in the next episode. Have a great morning,
• 10:58
  day, evening or night and don't forget to follow,
• 11:01
  because more marketing facts are waiting for
• 11:03
  you.