Samenvatting
Marketing Metrics Exam Summary
- Instelling
- Universiteit Van Amsterdam (UvA)
Complete exam summary for the year 2023 at UvA MSc Business Administration
[Meer zien]Voorbeeld 4 van de 40 pagina's
Voorbeeld 4 van de 40 pagina's
In winkelwagengesponsord bericht van onze partner
Voorbeeld van de inhoud
Week 1 | Consumer preferences | Lecture: February 7, 2023(New) Product Design
● It all starts with understanding your customers goals, needs and preferences.
○ Psychological needs (e.g. social status associated with products)
○ Functional needs (e.g. does it fulfill its intended purpose?)
● How do they trade off these preferences for money? Would they rather keep their money in their pocket (or
buy an alternative product)? Will they pay for it?
● What we want to know: Given what people like, and need – How much value do products have for consumers?
How to measure customer value?
- Behavioral measures - only available for existing products
- Constrained measures to elicit consumer preferences are called constraint because consumers will not be able
to say everything is important. Consumers have to choose. Therefore, they cannot value everything anymore,
and thus are constrained in their preferences.
- What type of measure would you suggest to gain more information when developing a new product? A
perceptual measure. As you do not have behavioral data of a new product yet (since it is not being sold), you
can only use perceptual measures.
Thurstone’s idea to measure consumer preferences
● Thurstone presented an idea to measure consumer preferences in 1931 at a meeting of the Econometric
Society
○ “Perhaps the simplest experimental method that comes to mind is to ask a subject to fill in the blank
space in a series of choices of the following type: eight hats and eight pairs of shoes’ versus ‘six hats
and ___ pairs of shoes’ One of the combinations such as eight hats and eight pairs of shoes is chosen as
a standard and each of the other combinations is compared directly with it.” -> they say based on the
suggestions you cannot figure out what people are going to pay.
● In 1942 - Asking people to fill in the blank space in a series of choices. The academic crowd heavily opposed the
idea that non-market (non-actual) behavior could teach us something about consumers.
● In 1942, other academics summarized Thurstone’s “futile” attempt as:Thurstone’s fundamental shortcomings
probably cannot be overcome in any experiment involving economic stimuli in human beings
● Only from the 1960’s onwards academics really continued to try to extract valuable information from stated
preferences
Stated vs. Revealed Preferences
1
, - Revealed preferences – Preferences that are shown. Better indicated with your actual preferences.
- Unconstrained - no real constraint to the answers the respondent might give. You can give an answer to
everything. Answer even if you don’t care about them equally.
- Constrained - you cannot say everything is important
- downside of stated preferences: Consumers rarely follow their stated preferences. Consumers often
overestimate their likelihood of buying a product. You can ask for stated preferences for any product you want.
Intention vs. purchase
● For non-durable it’s close to 40% - for these people that were really convinced only 40% bought the product.
It’s worse for durable products. So it’s hard to rely on what consumers say.
● Left part of the graph - people who said will definitely not buy, some of these ended up buying the product.
● So there is a gap between intention and purchase.
● People overestimate their likelihood to purchase more for durable vs. non-durables
● This means that roughly 10% of the people that “Will definitely buy it” have actually purchased it.
● Especially for durable products it’s difficult for people to say whether or not they will want to buy it. Durable
line is also relatively flat.
Difference between “what people say and do”
● Sustainability x higher price → will not buy :) Keep their money in their pocket
● People care but they might not act on it if they actually have to put their money on it. People don’t just do as
they say.
● The goal is to figure out what they will do, not what they say they will do
● Social desirability bias - difference between what people say and do
Is there any value in asking people what they want an existing/a new product to look like?
● Consumers face different types of decisions on a daily basis: consumers are good at making trade-offs
(decisions) between benefits and costs. So it makes sense to make respondents make trade-offs
● Conjoint analysis: a survey based statistical technique designated to consumers to make trade-offs, forcing
them to reveal their true preferences. Produces a mathematical system of their preferences. Trying to get as
close to reality in terms of making a lot of trade-offs.
○ Conjoint analysis - adding hypothetical products - consumers to make trade-offs/decisions, which
forces them to reveal their true preferences. Produces a mathematical system of their preferences to
model their preferences to see which aspects they care. It shows that certain things are more
important that other things.
- What method gets you as close to revealed preferences, without actual purchase data: Trade-offs.
- You need actual purchase or choice data for behavioral measures. Absent that, trade-offs come
closest to revealed preferences.
Unconstrained vs. constrained (scanners)
- What does the traditional approach and the conjoint study tell us about the value of scan time with scanners?
Scan time can be valuable, but consumers value resolution, brand and price more. Respondents can find all
attributes valuable, but true (stronger) preferences come to light when they have to make trade-offs in their
choices.
- With conjoint analysis we’re able to have contraines results to see what consumers actual preferences are.
- Conjoint reveals what people really want. Rank based on conjoint. (Example of MBA’s and job attributes
ranking - revealing their true preferences) With traditional methods people give socially desirable answers.
Conjoint analysis helps us to see what really matters.
2
,Conjoint analysis → Conjoint analysis allows consumers to evaluate products. Consumers evaluate products, they have
to make trade-offs.
Conjoint world
- From a very functional perspective, a product is a collection of attributes.
- In conjoint, we treat all products as a bundle of attributes.
- Examples of attributes for a new phone: Size of phone, camera(s), cpu, price, color
- Examples of levels of attributes:
- Price: low (500), medium (850), high (1200)
- Camera: 12 megapixels, 32 megapixels, …
- Size of phone: 5.8 inch, 6.5 inch, 7.2 inch
- Iamsterdam city card options is also a collection of attributes, different types. Same with shoes, courses, airline
tickets.
Conjoint: why does it work?
- Respondents evaluate bundles
- Example: please indicate on a scale of 1-100 how you evaluate these products…
- Relate the rating or choice on the attribute levels.
- We decompose the “overall utility” in partworths.
- We uncover the relative importance of each attribute.
- We can include fictional products!! Since a conjoint does not have to be limited to existing products in the
market, any desired attributes or attribute levels can be used.
Conjoint: the process/designing the study
Attributes- what’s important when looking at a specific product.
Step 1: Attributes
How to decide on which attributes to include;
● Focus groups, interviews with NPD team, analysis of current market, pre-tests
● Not too many attributes (keep it simple/limited here)
○ Pick the most important ones
○ Consumers may get confused with too many
○ Pizza → Crust, toppings, price, cheese type, cheese amount
- For selecting attributes, you have an analysis of the current market available. However, there is a new attribute,
which you expect to be important, that was not measured. Do you include it? Yes, important attributes should
be included, even if they are not in the market yet.
- Important attributes should always be included. Untested attributes do not need to be confusing if
respondents are familiar with them, or can easily understand them.
Step 2: Attribute levels
- Need to cover the whole range of attribute levels (e.g. price ranging from low to high)
- Can be;
● Low/medium/high (e.g. price)
● Yes/No (e.g. is there pepperoni on the pizza?)
- Best to limit number of levels
- Best to keep number of levels roughly the same for different attributes (e.g. 3 different price levels, 3 different
toppings)
- Pizza → crust (pan, thin, thick), price (7,99 8,99 9,99)
● Ideally, how should our attribute levels look like: They should be realistic, and cover a wide (realistic) range.
Answers or choices are likely to be unreliable if they are not realistic: it will be hard for respondents to imagine
them. Attribute levels are also less valuable when they only cover a limited and unrepresentative range. You
ideally want to cover the range that is representative of reality.
Step 3: Develop profiles (profile – attributes we can put together)
- Factorial design: all possible combinations of all levels of alternatives
- In the case of pizza this means 3x3x3x4x3=324(!) theoretical bundles.
- Very costly to ask consumers to evaluate all 324 bundles.
3
, - Solution → fractional-factorial design → carefully select bundles to maximize information that you
get from consumers
- How to select?
- Design needs to be balanced and orthogonal: every level (i) occurs in roughly equal proportions, (ii)
across the levels of other attributes)
- Balanced: Seeing the low price as many times as the high price.
- Orthogonal: A low price should not be always accompanied by pepperoni. Seeing each level
often equally and there is no correlation between pepperoni and price.
- Use computer program (e.g. SPSS/SAS/R/etc.) to select best design (more complex but efficient
designs exist)
- Include at most 25, but even better, 16 designs (for this scenario)
● Orthogonal design - a designed experiment is orthogonal if the effects of any factor balance out (sum to zero)
across the effects of the other factors. Orthogonality guarantees that the effect of one factor or interaction
can be estimated separately from the effect of any other factor or interaction in the model.
● the main downside of a factorial design: There are often too many combinations to test. With just a few of
attributes and attribute levels, the number of possible combinations are far too many to test. Unrealistic
combinations are not ideal, but their data could always be excluded later.
● Why do we want a fractional-factorial design to be balanced and orthogonal? Because we can obtain the most
information with diverse trade-offs. You obtain most information with diverse trade-offs. You might still want a
balanced and orthogonal design if you expect certain attributes/levels are more important, since combinations
of different designs might be hard to predict.
Step 4: Design data-collection procedure
How to present profiles?
● Verbal - waiter explaining you the pizza
● Pictorial
● Physical (prototypes) – Ideally you want to present options as clearly as possible since we want to learn about
their preferences. (e.g. physical/prototypes as presentation of pizza)
How to evaluate?
● Choice-based conjoint (or pairwise evaluations)
○ Mimics the reality most, you just pick one in reality - more realistic
○
○ Early days - Pairwise evaluation: Do you prefer Pizza 1 or Pizza 2? This allows us to compare different
products.
■ Thurstone wasn’t wrong after all…
○ Current practice: via choice in a supermarket
● Rank-based conjoint
○ Rank from worst (e.g. 16) to best (e.g. 1)
○ You get more comparative information, usefulness is not guaranteed
○ Downside: respondents usually do not care about “middle” , only about top & bottom. You can really
know what someone wants and definitely doesn’t want. But then there's the gray middle. Products
people really don’t care about.
● Ratings-based conjoint
○ What their preference is on a specific scale.
○ Good: you evaluate 1 at a time
4
Voordelen van het kopen van samenvattingen bij Stuvia op een rij:
Verzekerd van kwaliteit door reviews
Stuvia-klanten hebben meer dan 700.000 samenvattingen beoordeeld. Zo weet je zeker dat je de beste documenten koopt!
Snel en makkelijk kopen
Je betaalt supersnel en eenmalig met iDeal, creditcard of Stuvia-tegoed voor de samenvatting. Zonder lidmaatschap.
Focus op de essentie
Samenvattingen worden geschreven voor en door anderen. Daarom zijn de samenvattingen altijd betrouwbaar en actueel. Zo kom je snel tot de kern!
Veelgestelde vragen
Wat krijg ik als ik dit document koop?
Je krijgt een PDF, die direct beschikbaar is na je aankoop. Het gekochte document is altijd, overal en oneindig toegankelijk via je profiel.
Tevredenheidsgarantie: hoe werkt dat?
Onze tevredenheidsgarantie zorgt ervoor dat je altijd een studiedocument vindt dat goed bij je past. Je vult een formulier in en onze klantenservice regelt de rest.
Van wie koop ik deze samenvatting?
Stuvia is een marktplaats, je koop dit document dus niet van ons, maar van verkoper idilugurlu. Stuvia faciliteert de betaling aan de verkoper.
Zit ik meteen vast aan een abonnement?
Nee, je koopt alleen deze samenvatting voor €12,79. Je zit daarna nergens aan vast.
Is Stuvia te vertrouwen?
4,6 sterren op Google & Trustpilot (+1000 reviews)
Afgelopen 30 dagen zijn er 73918 samenvattingen verkocht
Opgericht in 2010, al 14 jaar dé plek om samenvattingen te kopen