Consumer Behaviour PDF

CONSUMER BEHAVIOUR ================== **Introduction to consumer behaviour** *What is marketing?* To critics marketing is: 1. Fluff 2. Manipulative 3. Wasteful - Not clear what contributes to profitability of a firm (advertising?) To our professor: 1. Marketing is not rocket science...

CONSUMER BEHAVIOUR ================== **Introduction to consumer behaviour** *What is marketing?* To critics marketing is: 1. Fluff 2. Manipulative 3. Wasteful - Not clear what contributes to profitability of a firm (advertising?) To our professor: 1. Marketing is not rocket science - It's much harder, it's a social science - You will not by accident launch a rocket that makes it to the moon and back - One-to-one relationship between quality of relationship between the quality of the process that you followed and the quality of the outcome 2. Marketing is behavioral science - No guarantee that good process will lead to a good outcome or that bad process will lead to a bad outcome (a lot of dumb luck) - Difficult to predict what consumers will do or how to influence them - Increase likelihood: valuable increase, not always recognized *Definition:* Marketing is the activity, set of institutions, and processes for creating,\ communicating, delivering and exchanging [offerings that have value] for customers,\ clients, partners, and society at large. - Marketing is about value creation for customers, stakeholders and firms - Value from engineering: good products - best possible products - In today's day and age, product quality is the most important (immediately observable: you just have to go online) - Value from marketing: everything around the product Not long ago, key pieces of information: - Price - Brand name - Salesman - No social influences Now: - Go online (f.e. Amazon) - Look at comments - Test safety (simulate the crash but with dummies) - Alternate the crash with different products - What's the effect of the product? *Correlation rating amazon and rating consumer reports* - Collecting data from different categories - There are some measurement instruments that allow you to measure the objective performance of products - Analysis: - High correlation → higher user ratings → higher quality - 0 correlation: no correlation at all - Negative correlation: higher rating on amazon have lower quality - Example: *Correlation (star, performance) = ?*  If you take two products and the difference in star ration is smaller than 0.20, the likely hood that the one with higher score is better: 50% (also if its smaller than 0.40) → only if its higher that 0.4: 70% What is effective in user ratings: - Perceptive consumers of head product World has changed: - Persuaded by pretty models - Now: you look up the product and see what people have to say about the product (comments are influenced by celebrity's information/choice) - User ratings are higher (the more you pay, the more satisfied you are), user ratings reflect promotions - How marketing variables influence quality perceptions (but not only by asking people to rate things, but also by experimenting with scans) - Asked to sample wine, while you are in a scanner - You taste 5 wine samples (A and B the same, C and D also the same and E is different), the order in which the wines are presented are different - Liking without price: - No difference in liking A and B - No difference in liking C and D - Liking with price: - Almost perfect correlation between price wine and liking - Other experiment: - What are you willing to pay? - A 10-ounce cup filled up with 8 ounces - Willing to pay \$1.66 - A 5-ounce cup filled up with 7 ounces - Willing to pay \$2.26 (willing to pay more) - Value is not so much determent by the ingredients but the packaging - Important: single evaluation context - Last experiment: - 48 new songs from new artists, make it available for consumers from different markets - Independent market A: 1000 consumers, record how much people listen to a certain song, no interaction with others - Independent market B: 1000 consumers, record how much people listen to a certain song, no interaction with others - Social market A: 1000 consumers, record how much people listen to a certain song, consumers can observe from each other how much a song is listened to - Social market B: 1000 consumers, record how much people listen to a certain song, consumers can observe from each other how much a song is listened to - Rank of each song in each market: looking at the correlation - What market has the highest correlation (consistency of popularity)? - Independent market: in the independent market you can only be influenced by the song (intrinsic quality is going to drive the popularity) - Social market: the random process can be different (market A: first song is 12, market B: first song is 36) → second listener is going to listen to the first song that first listener listened - In which market is the number of the listeners of the most popular song the largest? - Social market: more hoarding effects (increasing the already popular things) *Value = f(marketing mix)* - Product - Promotion - Price - Place *Value for consumers → leads to value for firm itself* Afbeelding met tekst, schermopname Automatisch gegenereerde beschrijving **\ ** **Managerial thinking traps** Big idea: value that consumers see depends on more than objective performance product itself (there is a correlation, but price is also correlated, names of brands) Drivers value: - Packaging - Society (social influences: preferences)  *Peter Drucker: "Because of the purpose of business is to create a customer, the business enterprise had two -- and only two -- basic functions: marketing and innovation"* - Influencal thinker in business management *David Packard: "Marketing is too important to be left to the marketing department."* - HD Afbeelding met tekst, cirkel, diagram, schermopname Automatisch gegenereerde beschrijving *Egocentrism: everywhere* - You think you are crystal clear but you communicational partner has no idea what you are talking about - You assume that your knowledge is the same as others (not concrete enough) - Harms communication with others - Can influence collaborations - You overestimate your input/share of the contribution - Undervalue other's contributions/overvalue yours *Company perspective ≠ customer perspective* - People who work in business are different than the people they are trying to sell to - Marketing people more online, think everyone uses it that much, put to much money in it - Ad people: advertising in organizations - If you want to be effective to influence the people you are trying to sell to: ask their expectations, things that they find important, their believes *Take perspective of your target customer* - 233 marketing managers watched two Rolex ads (sailing vs. golf) and rated personal preference for each ad (1-11 scale) - Split in to two groups: - Empathy group: Visualize target customer, imagined thoughts/reactions (how they would respond to the ad) - No-empathy group: No instructions given - All managers viewed market research on customer evaluations and then estimated target customers' attitudes toward each ad (1-11 scale) - Half: perspective taking instructions  Afbeelding met diagram, lijn, Perceel, Parallel Automatisch gegenereerde beschrijving *Mark Twain: "It ain't what you don't know that gets you in trouble. It's what you know for sure that just ain't so."* - He never said this  *Does negative feedback and searching contradictory evidence reduce over confidence?* - Negative feedback: telling someone they are bad - Searching contradictory evidence: if you think something is true, find evidence against it - Study: - 90 marketing students make 70 predictions, each time indicating their confidence - Monthly magazine advertising for smoking materials in September was 9% of all magazine advertising. In October, this percent will be: a. Smaller or the same b. Greater - What is the probability you will be correct? (50-100%) - Split into three feedback groups: - Favorable: You are in the top one-third - Neutral: You are in the middle one-third - Unfavorable: You are in the bottom one-third - Further split into two reasoning groups: - Contradictory evidence: Think about the major reasons why you may be wrong - No contradictory evidence: No instructions given (control group) - Students make 70 more binary predictions, each time indicating their confidence. - Criticize study: - Not making statements, trying to make a comparison between different groups - Sample size: 15 participants per condition (too small) - Something easy generalizable: not always possible - Major replication crisis because of the small sample size Afbeelding met tekst, lijn, schermopname, diagram Automatisch gegenereerde beschrijving Three control conditions without generate contradicting evidence: - Vertical axes: overconfidence - Confidence is not lower but accuracy is higher - Unfavorable: reduces overconfidence (they are more motivated to do well) - The accuracy becomes higher when getting unfavorable feedback With generate contradicting evidence: - Accuracy is higher - Thinking badly about yourself: good results - Thinking critically: you think a but why not A → make you smarter Important to know - Anything can have an influence on your behaviour - Behavioural science: try to understand how one variable influence your behaviour - You cant look at every variable Afbeelding met tekst, lijn, Perceel, diagram Automatisch gegenereerde beschrijving - CFOs are asked to make a prediction about the annual return of the s&p 100 - Does the actual performance fall within or out of their bounce? *! Projecting knowledge: giving precise numbers (precision persuades) !*  Afbeelding met tekst, schermopname, lijn, Perceel Automatisch gegenereerde beschrijving  known knowns: things you know and you know you know them ↓ if you find the answers to things you don't know, the part of known unknowns gets smaller ↓ and the part of known knowns gets bigger known unknows: things you don't know, but you know you don't know them ↓ if you ask yourself questions about a certain topic, the known unknowns part gets bigger ↓ and the unknown unknowns part gets smaller unknow unknows: things you are not aware of that you don't know - Education: recognizing the things you don't know *Mark Twain: "Education is the path from cocky ignorance to miserable uncertainty."* *Effectiveness online ads* *John Wanamaker: "Half of the money spend on advertising is wasted. The trouble is I don't know which half."* - Merger in the 1900 - Nowadays: rise of digital advertising changed us, we can look more into things *Common advertising metrics* - Cost per thousand impressions (CPM): how much you are paying for 1000 exposures of an ad - Click-through-rate (CTR): proportion of consumers that click through when seeing your ad - Cost per click (CPC): how much are you paying per click - Conversion rate (CR): proportion consumers who make a purchase when seeing your ad - Customer Acquisition Cost (CAC): price paid by advertiser to require 1 more customer - Return on Advertising spend (ROAS): sales, how many dollars consumers are spending that were exposed to an ad - Advertising Cost of Sales (ACOS): how much did you spend on advertising/sales by people who saw the ads Because of ability to measure this (good view of effectiveness advertising) *Article: offline impact of online ads* - Display ad: next to an article \>\< Search ad: you look something up yourself - Wrong time: you buy because of the holiday season (correlation may not indicate causation) - Comparing multiple groups (4) - Search ad: most effective (bar higher) - Display ad - Both - None - Magnitude: too large - Nothing wrong with bars, but words - If you invest a dollar, you are getting three dollars back: marketing people would be rich - Search ad: you want to make a purchase in this category - Consumer was going to make a purchase anyway - Not necessarily because of the ad - Display ad: company has collected data about you, predict what your interests are → send it to you - Conclusion incorrect 'boost', 'effective': wrong - Lift probably not looking at causal effect and only correlations *'LIFT'* - R: revenues - LIFT = (R~exposed~ -- R~not\ exposed~) /R~not\ exposed~ For instance: - If R~exposed~ = 200 and R~not\ exposed~ = 50 - CR: conversion rate (probability of buying) - LIFT = (CR~exposed~ -- CR~not\ exposed~) /CR~not\ exposed~ For instance: - If CR~exposed~ = 3,3% and CR~not\ exposed~ = 1,1% *Correlation is not causation* - Correlation between ice cream sales and shark attacks, but buying ice cream is not the cause of shark attacks - 3^rd^ variable: wheater (warm → ice cream → beach → swimming → sharks) - Inconsistence with mental model of how the world works - Correlation: when X changes, Y changes also - Causation: when X changes, Y changes because X changed Afbeelding met diagram, tekst, lijn, Lettertype Automatisch gegenereerde beschrijving - Consumers can see a display ad: yes or no - Consumers can see a search ad: yes or no - Consumers can buy a product: yes or no - Causal variable? *ROAS = Non-causal metric* - An associative metric: you are dividing two things, you are dividing sales by a metric - Revising sales by advertising - Campaign does not necessary cost the revenues - You don't know if there is a causal impact if the number of ROAS is high *The purchase funnel* Afbeelding met tekst, schermopname, Lettertype, lijn Automatisch gegenereerde beschrijving *Do consumers visit the eBay website less if ebay stops brand keyword search ads?*  *Article: about measuring sales impact* - You get customer id's, twitter says if they viewed the tweet, Ben and Jerry ads how much they spend on Ben and Jerry - Problem: Twitter shows tweets to people likely to buy ice cream *Example: Privacy policy Apple* - Serve personalized ads: most say no - Facebook: problem if people say no, make campaigns against Apple - Without personalized ads, small companies will die - Without personalized ads, Facebook data shows that the average small business advertiser stands to see a cut of over 60% in their sales for every dollar they spend - What is the data that Facebook really used: are these the same people? - Logic Facebook: analyzed thousands of campaigns across many industries and regions. Facebook split campaigns into two groups: 1. One group was optimized for link clicks (generic advertising) 2. The other group was optimized for purchases (personalized ads) - No causal effect RANDOM ASSIGNMENT IS KEY!!! *Vaccines* - Randomize control trials (key square test) - Pfizer: was found effective to be more than 90% effective in preventing COVID-19 - There was a RCT conducted with about 40 000 participants - There were 94 confirmed cases - 8 in the vaccinated group (out of 20 000) - 86 in the placebo group (out of 20 000) - Vaccine efficacy = 1 -- (8/86) = 0,906977 - Moderna: vaccine efficacy of 94,5% - Indeed more effective - You want evidence from randomized control trial, if you don't have it, you ask yourself what is missing, what variable is missing and what is the correlation with the other two variables - Data from Israel: - 515 patients in the hospital with severe cold, 58% of them fully vaccinated → vaccines not effective - How many people are vaccinated? - 5 643 643 people are fully vaccinated - 1 302 912 people are unvaccinated - Probability to be in the hospital when vaccinated and when not vaccinated: - 0,0053% of people who are fully vaccinated are in the hospital - 0,0164% of people who are unvaccinated are in the hospital - Vaccine efficacy: 1 - (0,0053/0,0164) = 67,5% - Who are the vaccinated people? *Simpson Paradox*  **\ ** **Why consumers (don't) buy** Specific pricing in stores: intentional? → more weight on the first digit you see (first digit bias) Round prices: easier to process → cheaper *Nobel prize* - Effects of colonization on wealth and nations (change in the system to look at the general effect) - They examined political and economic systems introduced by European colonizers. When Europeans colonized large parts of the world, the existing institutions sometimes changed dramatically, but not in the same way everywhere. - One factor that influenced institutional differences was the population density of the colonized area - Denser indigenous populations (larger) → more resistance → fewer settlers → extractive institutions (when they did move in) → prosperous elite - Sparser indigenous populations (less) → less resistance → more settlers → inclusive institutions → prosperous population - Unique reversal of relative prosperity - Before colonization: denser areas were doing better - After colonization: sparser areas were doing better - Doesn't happen a lot - Another factor that contributed to institutional differences was the severity of the disease in the colonized area - Regions closer to equator → more numerous and dangerous diseases → higher settler mortality → extractive institutions → prosperous elite - Regions further form equator → fewer and less dangerous diseases → lower settler mortality → inclusive institutions → prosperous population - Persuade other people of causal effect - Not randomized control trial!!! - To understand causal effect: collect historical data, give indication of causation but it's not perfect (what variables am I missing? How does conclusion changes) - Truly interested in effect of institutions and economic prosperity - Increase Nobel prize winners: - Government recommends eating more chocolate - High correlation between amount consumed chocolate and Nobel prize winners (could be a causal effect, but not necessary) → ingredient in chocolate that has a positive influence on the brain working - Hypothesis 2: Smart people know the effect of chocolate and then start eating it - Additional set of data, describe this also in our model → the coefficient that we estimate for chocolate consumption is going to show the partial correlation between chocolate intake and Nobel prize winners controlling for all these variables (marketing-mix-model) - Variable: sales - "affect": causal word - Observations: - The richest 20% of countries are 30 times richer than the poorest 20% (prediction that this will stay like this) - Because of differences in political and economic institutions - Empirical challenge: correlation does not necessarily imply causation! - Institutions may influence prosperity - Prosperity may influence institutions - Rich and poor countries differ in many ways (may have influence on type of political and economic institutions or prosperity) → access to research, climate, geographic location (multitude of variables) *Experiment, A/B test, RCT* - Difficult to do: randomized control trial (long study) - Study: - Students making an exam - During the exam they can take a chocolate from a bowl at the front of the class - 20 students took one: mean score = 85 - 30 students didn't take one: mean score = 68 - You cannot conclude if it has a causal effect - Different types of students (did not study that much → need a lot of time → no time for chocolate) - Type of reasoning relevant to understand experiments in the world *A/B Test (randomly assign ads to consumers)* - Compare two versions of an ad strategy by changing variables - Each version is shown to a segment of your audience and ensure that nobody sees both - Determine which version performs best - Good if you are trying to measure changes to your advertising or quickly compare two strategies - Not good if you test informally (overlapping audience) - Interrupt campaign and look at something else - Algorithm based on interests: targeting same type of people? No, they converged on different users - You cannot draw causal conclusions  - Blue ad versus pink ad → you cannot conclude that pink ad is more effective - Man: only on blue ads or woman: only click on pink ads - 1000 impressions bought buy men (blue) → 10% of men buy K - 2000 impressions bought buy woman (pink) → 20% of woman buy K *RCT* "Without personalized ads, Facebook data show that the average small business advertiser stand to see a cut of over 60% in their sales for every dollar they spend" Afbeelding met tekst, Lettertype, schermopname, logo Automatisch gegenereerde beschrijving  *Algorithmic targeting* - Online world mediated by algorithms (algorithms decide what you see) - Different categories of people that can see the ad (based on interests) - People interested in the outdoors and the environment - People interested in flowers, pets and gardens - People interested in crafts, decor and conservative - People that are less relevant - Algorithmic targeting: not a random selection, you look at people' interests and give them ads based on their likings *Conversion Lift* - You have an audience, algorithm that targets specific users based on background characteristics - Doesn't serve ads to all the users the algorithm would target - Algorithm converges on a subset of the audience, it serves ads to a subset of that subset (at the last moment: it decides not to serve the ad to some users → random) - Correlation → causal - How large or causal effects?  *Divergent algorithmic targeting in "Lift tests"* Afbeelding met tekst, schermopname, diagram, Lettertype Automatisch gegenereerde beschrijving - A versus B case: different subset of users (exposed and unexposed group of users) - Causal effect on different groups: not a comparation of causal effects on same group of people - You cannot control targeting algorithm, you have to accept that there is mediation (these tests give you good idea if you are better off doing ad A or ad B) → Where is largest incremental effect/largest changes? - Is this useful when looking at: Which type of psychological mechanism is driving purchases? When presented an ad: look at it or look away? - What is the variable that leads to this decision? - No, is confounded, you look if they are looking at you or not in a different part of the population - Comparison across ads always confound (invalid) *Is it possible to approximate true causal effects with non-experimental approaches?* Afbeelding met tekst, diagram, lijn, Perceel Automatisch gegenereerde beschrijving Afbeelding met tekst, diagram, lijn, Perceel Automatisch gegenereerde beschrijving  *Multi-Attribute Utility Theory* - 17^th^ century - Probability theory - 18^th^ century - Utility theory *Expected Monetary Value* Which alternative has the highest Expected Monetary Value? A. €1000 B. 50% chance at €2000 and 50% chance at €0 C. 60% chance at €6000 and 40% chance at -€2000 (correct) D. It depends - In history: if you aspect A or B: chances are (A+B)/2 - EMV = p~1~x~1~ + p~2~x~2~ +... + p~n~x~n~ Which alternative should you choose? A. €1000 (correct) B. 50% chance at €2000 and 50% chance at €0 C. It depends - Risk preferences Imagine your wealth is €0: A. U(€1000) = (€1000)^1/2^ = 31,62 → wealth €100 000 = 317,80497 B. 0,5\*U(€2000) + 0,5\*U(€0) = 0,5\*(€2000)^1/2^ + 0,5\*(€0)^1/2^ = 22,36 → wealth €100 000 = 317,80107 Which alternative has the highest expected utility? A. €1000 B. 50% chance at €2200 and 50% chance at €0 C. It depends (correct) Imagine your wealth is €0: A. U(€1000) = (€1000)^1/2^ = 31,62 → wealth €100 000 = 317,80497 B. 0,5\*U(€2000) + 0,5\*U(€0) = 0,5\*(€2000)^1/2^ + 0,5\*(€0)^1/2^ = 23,45 → wealth €100 000 = 318,7462 *Diminishing Marginal Utility* - Relationship between monetary wealth and the value of monetary wealth - First dollar you have is more valuable than the 1000 dollar - Formula: EU = p~1~u(x~1~) + p~2~u(x~2~) +... + p~n~u(x~n~)  *Why should the rich get richer, and the poor poorer?* - Wealthier people will accumulate wealthier decisions (higher expected monetary value) *Multi-attribute utility* 1. Identify alternatives 2. Identify a set of dimensions/attributes (consequences) 3. Assign a weight to each dimensions/attribute (consequence) 4. Ranking each alternative in each dimension/attribute based on utility (consequence) 5. Integrating all information in to one utility per dimension using a weighted additive rule  Afbeelding met tekst, schermopname, Lettertype, nummer Automatisch gegenereerde beschrijving *Multi-attribute view* - Decomposing in components (utility is higher with one of the two choses) - Fuel efficiency - Space in the back Afbeelding met tekst, schermopname, Lettertype, nummer Automatisch gegenereerde beschrijving Afbeelding met tekst, schermopname, auto Automatisch gegenereerde beschrijving *Rating-based conjoint analysis* - How likely are you to purchase this airline ticket for your holiday to Spain?  - Estimate the weight of the utility that people put on something - Invite people to take a survey, imagine you are presented with something, how likely are you to buy? Variate variables → correlation → coefficient says something about weight - Example: Courtyard by Marriott - Landscape - Building size - Prize - Statistical model *Choice-based conjoint analysis* - Which airline ticket would you purchase for your holiday to Spain? Afbeelding met tekst, Lettertype, schermopname, nummer Automatisch gegenereerde beschrijving - Not individual alternative, but multiple - Very similar data: attribute level per alternative - Most people say this works better *Multi-attribute utility*  - Decisions according to this → more alternatives are better because you have a 3th column, the utility of most preferenced item will only go up *And ordinary supermarket...*... contains 285 varieties of cookies, including 21 chocolate chip options alone; 20 different types of Goldfish crackers; 13 "sports drinks," 65 "box drinks," and 85 flavors of juice; 12 varieties of Pringles potato chips; 80 pain relievers; 40 lipstick shades; 16 varieties of instant mashed potatoes, 75 different instant gravies, and 120 different pasta sauces; 175 different salad dressings; and a whopping 275 types of cereal. - If you would go through model → to much - More alternatives not always better → overwhelmed feeling - Study jams: - In a store: jam tasting booth - Manipulating the amount of jams that are offered - 6 - 24 - How likely are people to stop? - 24: a lot of people stop - How likely are people to buy? - 6: more likely to buy - 24: a lot of people don't buy - No replications - Study jobs:  - Participate in pension plan → tax benefit → employer will match your savings - Likelihood that people invest in pension plan based on the amount of alternatives of investments they have - Decrease with number of alternatives - Study grocery store: Afbeelding met tekst, schermopname, diagram, lijn Automatisch gegenereerde beschrijving - Different alternatives in each category → decides to eliminate some - Observe effect - Vertical axis: effect on sales - Horizontal axis: how much they eliminate *Independence of alternatives* - Alternatives are independent of each other - Utility only depends on the alternative which you calculate it for  - You are about to buy an iced tea: which size do you choose? a. 236 ml for €3 b. 354 ml for €4 (preference) c. 473 ml for €5 - They don't know the utility/value or disutility → want more iced tea but want to pay less - You are about to buy an iced tea: which size do you choose? a. 354 ml for €4 b. 473 ml for €5 (preference) c. 591 ml for €6 - Irrational *Preferences* - Study subscriptions 1: - Online subscription for \$59.00 (57%) - Paper subscription for \$125.00 (43%) - Study subscriptions 2: - Online subscription for \$59.00 (28%) - Paper subscription for \$125.00 (0%) - Paper subscription and free web for \$125.00 (72%) - Weights are different → difficult to predict *Asymmetric dominance effect: adding dominated alternative 3 increases preference for dominating alternative 2* **\ ** **Goals and reference points** *Video about illusions*: - The same thing can be looked at in different ways based on reference points - Perception is not reality (can be very different) - Context dependent *Should our professor buy travel cancellation insurance?* - Flight to Barcelona for €350 - 1% chance that he'll have to cancel - Travel cancellation insurance: €30 → reimburse full price - Total wealth: €100 000, after buying ticket, before buying insurance Expected monetary value/wealth: - 1% of 350: €3,5 - Insurance: €30 Afbeelding met tekst, schermopname, Lettertype, nummer Automatisch gegenereerde beschrijving - If you don't buy: - Total wealth: €100 000 (flight: 0,99) - Total wealth: €100 000 (no flight: 0,01) - If you buy: - Total wealth: €99 970 (flight: 0,99) - Total wealth: €100 320 (no flight: 0,01) - Expected monetary wealth without insurance: €100 000 (= 0.99 \* 100 000 + 0.01 \* 100 000) - Expected monetary wealth with insurance: €99 973,5 (= 0.99 \* 99 970 + 0.01 \* 100 320) Assume your professor's utility of monetary wealth follows a square root function of wealth. If he wants to maximize expected utility, should he buy the insurance?  - Risk averse: lower risk option - You don't care about expected monetary but about utility - Bernoulli: without insurance *Daniel Kahneman (-- Thinking fast and slow)* - Founding father behavioral economics - Descriptive model of choice (how. You perceive information) \>\< normative models of choice (ration: how you should make decisions, didn't think if it was... plausible) - How do people think about probabilities? - Subjective utility transformation about outcome - No transformation of outcome - Do people really perceive a 1% chance as a 1% chance or do they perceive it as 10%? *Prospect theory's probability weighting function* Afbeelding met diagram, lijn, Perceel, tekst Automatisch gegenereerde beschrijving *Availability heuristic*  - Consider all the information you know about that event - Subset: influenced through several forces (companies, social media,...) - Overestimate - Underestimate *Age distribution for people who died with insurance \>\< age distribution of people who died* Afbeelding met tekst, schermopname, Lettertype, nummer Automatisch gegenereerde beschrijving Afbeelding met tekst, schermopname, Lettertype, nummer Automatisch gegenereerde beschrijving *Opinions* Get a book from a bookstore located 10 minutes away because it's €5 less (€15 → €10) - Depends on value of your time - Many say yes Get a book from a bookstore located 10 minutes away because it's €5 less (€215 → €210) - Proportional value: irrational - Few say yes *Illusions and reference points*  Suppose that your current net wealth is worth €10 000. You win a prize and can choose one of the following options. What would you choose? (1) - Receive €10 000 for sure (may say this) - Flip a coin: €20 000 if heads, or nothing with tails Suppose that your current net wealth is worth €30 000. You must choose one of the following options. What would you choose? (2) - Lose €10 000 for sure - Flip a coin: lose €20 000 if heads, or nothing if tails (many say this) - You should be risk averse *Diminishing Marginal Utility* Afbeelding met tekst, ontvangst, lijn, Lettertype Automatisch gegenereerde beschrijving Utility of monetary wealth is concave function of monetary wealth - Certain option: higher utility *Prospect theory* Value function: - Shape cannot explain but describe how people make choices  - People don't evaluate their absolute outcomes but compare outcomes to reference point (not in linear way but proportional way) - Diminishing sensitivity to gains: U(win 10k) \> 0,5 \* U(win 20k) - Diminishing sensitivity to losses: U(lose 10k) \< 0,5 \* U(lose 20k) *Questions* Imagine you invested €10 000 in a company's stocks exactly one year ago. Today, you check your investment and find that its value had dropped to €5 000. What would you do? - Hold on to the stocks - Sell the stocks → What do you expect to happen to €5 000? Reference point is €5 000 Imagine cat jumps on your computer: sells stock → reference point is €5 000 instead of €10 000 - Why do people do this? - Losing €5 000 doesn't seem different from €6 000 - Hold on to losers, selling winners Imagine that a large car manufacturer has recently been hit with a number of economic difficulties that may affect 6 000 employees. The vice President of production has been exploring alternative ways to avoid this crisis. Which plan do you prefer? - Plan A will save 2 000 jobs - Risk-averse - Plan B has one-third probability of saving all 6 000 jobs, two-third probability of losing all jobs Version 2: - Plan A will result in the loss of 4 000 jobs - Plan B has one-third probability of losing no jobs, and has two-thirds probability of losing all 6 000 jobs - Risk-seeking What do you prefer? - Flip a coin; if heads, you win €250. If tails, you lose €100 - I'd rather not flip a coin. - Half says A, half says B (risk aversion) *Diminishing marginal utility* - Losses: aversive - Losses feel heavier than gain (losses loom larger than gains, gains have a less effect than losses) *Marshmallow experiment* - 1 marshmallow now or 2 marshmallows on a later time - Age is an important factor: the younger, the faster they eat (replicatesp - See how well these children are doing → grades, university, salary, jailtime? - Different coping strategies → looking away vs staring at it - ROI needs to be huge to not immediately consume something (2 marshmallows) *Selfcontrol* What do you want for dessert? - Apple - Now: few say apple - Later: many say apple - Tiramisu What do you prefer? - Receive €1 000 twelve months from now - Receive €1 100 thirteen months from now - Many wil say €1 100 Version 2: - Receive €1 000 now - Receive €1 100 in a month - Many say €1 000 - Irrational: return between now and a month is not 10% - Irrationality lies in combination of the two - Discount rate should be the same everywhere *Exponential discounting* What is the present value of €1 000 you receive now vs €1 100 you receive next month? What is the present value of €1 000 you receive in twelve vs €1 100 you receive in thirteen months? - Monthly interest rate of 5% Formula: PV = FV/(1+i)^t^ - PV = present value - FV = future value - I = interest rate - T = time - PV delayed/larger amount is larger than PV sooner/smaller amount What is the present value of €1 000 you receive now vs €1 100 you receive next month? What is the present value of €1 000 you receive in twelve vs €1 100 you receive in thirteen months? - Monthly interest rate of 15% - Sooner smaller reward is - Behavioral pattern: we apply different discount rate in near future compared to far future - Present biased (here and now very important for us) *Save more tomorrow -- Richard Thaler* → people are undersaving for retirement (extensive social system in Belgium), 3 simple principles - Present bias → ask to commit now to saving more for the future - Do you want to start saving in a year? More chances of saying yes - Now: they need the money - Loss aversion → link increases in savings to future pay raises - Feels like a loss, you lose the money now - Taking part of a raise to put it in a savings plan (losing part of a gain, feels better) - Minimal effort → once enrolled, you remain in the program unless you opt out - People are going to keep on doing it because stopping takes effort *Two models of human behavior* - Homo economicus = rational - Expected utility theory: diminishing marginal utility - Decision-makers should be risk averse - Forward looking and patient - The present is just another moment in time - Homo sapiens = predictably irrational - Prospect theory: diminishing sensitivity to gains and losses relative to a reference point - Loss aversion - Decision-makers can be risk averse or risk seeking - Myopic: the present is especially important and valuable *Meta-goals for choice processing* - We choose between attributes because they give some goal - Fill a need/goal 1. Maximize decision accuracy - Alternative that best helps you accomplish your goal 2. Minimize the cognitive effort needed to reach a decision - Painfull, it takes a lot of time 3. Minimize the negative emotion experienced while making the decision - Alternative has an effect on different times of goals (not making progress in another goal) - Incompatibility 4. Maximize the ease with which a decision can be justified - Towards yourself or others - How easy is it to explain your choice *Weighted additive rule* Afbeelding met tekst, schermopname, nummer, Lettertype Automatisch gegenereerde beschrijving *Equal weight heuristic*  *Compensatory vs. non-compensatory* Compensatory - The weighted additive rule and the equal weight heuristic consider all attributes and all alternatives. They are compensatory strategies. This means that stronger performance on one attribute compensates for weaker performance on another. - Consumers will often not consider all attributes or not all alternatives. They use non-compensatory strategies. This means that stronger performance on one attribute may not offset weaker performance on another. Non-compensatory strategies - Lexicographic heuristic - Compare on most important dimension - If tie on most important dimension, compare on 2nd most important dimension - Elimination-by-aspects heuristic - Eliminate alternatives that do not meet cutoff on most important dimension - Eliminate alternatives that do not meet cutoff on 2nd most important dimension - Continue until 1 alternative remains - Satisficing heuristic - Consider alternatives one at a time, in the order they occur - Choose first alternative that meets cutoff on all dimensions *Figure* **\ ** **Goals and reference points (cont'd)** - American elections - Not deterministic *Figure 1* Different decision processes  *Normative theory vs descriptive theory* - Normative theory: how should people decide? - Descriptive theory: how do people decide? - Distinction between the two *Homo economicus vs homo sapiens* - Homo economicus - How rational people look at behavior - Homo sapiens - Overweight small probabilities, underweight probabilities - Diminishing sensitivity: the impact of the win/loss is not times 2 if it you get double as much - Loss aversion: more sensitive to negative things, the weight of a loss is heavier than that of a win (animals are also like this) *Why do consumers often reject new products?* - Electric cars - Online grocery shopping - why slow to adopt them? - Not certain about benefits innovations: hesitant to use it (never 100% sure → 90% is seen as 50%) - Almost always an innovation contains some gains but also some losses - Innovators & early adopters: focus on gains because they already adopted 3 possible scenario's: Afbeelding met diagram, lijn, wit, Lettertype Automatisch gegenereerde beschrijving - Same benefits at lower cost (certainly going to buy) - More benefits at same cost (certainly going to buy) - Benefits and cost increase - Benefits don't occur immediately (grow overtime) - Losing now to win in the future ! in absolute terms: all the scenarios are the same ! *Barriers to adoption* - Reference point - Often the status quo, not your product! - Loss aversion - Losses \>\> gains - Subjective probabilities - 100% = "100%", but 90% = "50%" - Present bias - Losses are often immediate, while gains accrue over time - But gains and losses now \>\> gains and losses later *The default effect*  *Free trial* - Netflix: offers free trial to watch movies - You already put in you bank account information - Effort & poor memory - They offer a gain: loss is \$7,99, after a month: - Loss has become a gain - Gain has become a loss *Goals as reference points* - 2 people, normally doing both 25 push-ups - 1^st^ person: setting goal of 31 → completing 35 (feels happier) - 2^nd^ person: setting goal of 39 → completing 35 - Goals, like reference points, divide outcomes into regions of good or bad, success and failure (satisfaction with behavior) - Trish experiences more emotion: negative stronger than happy emotions Imagine this: - They can both do 1 more: who is going to work harder for the extra one? Trish (more motivated the closer you get to the goal) Afbeelding met tekst, diagram, lijn, Perceel Automatisch gegenereerde beschrijving Other examples: - Marathon finishing times  - Kickstarter Afbeelding met tekst, diagram, schermopname, ontwerp Automatisch gegenereerde beschrijving - Likelihood to fund is higher when they are far away from goal + magnitude is also going up  *Purchase acceleration* Afbeelding met tekst Automatisch gegenereerde beschrijving - The closer people are to goal: how much more they purchase - Consumer speeds up purchases, the closer they get to a goal *Post reward deceleration* Afbeelding met lijn, diagram, tekst, ontvangst Automatisch gegenereerde beschrijving - Back to initial interpurchase times - Then they accelerate again - Stampcard: - First two are prestamped - Prospect theory has reference points: ask yourself questions within these reference points  **Obesity crisis** Afbeelding met tekst, diagram, kaart Automatisch gegenereerde beschrijvingAfbeelding met tekst, kaart Automatisch gegenereerde beschrijvingAfbeelding met tekst, diagram, kaart Automatisch gegenereerde beschrijving *What is the most important cause of the obesity crisis?* - Calories consumed: people eat more - You can consume calories faster than you can burn calories - More efficient - Behavioral change to do this is very small - Calories burned: people move less - Online functions make it possible to do less yourself (ex. Livestream, online shopping, doordash,...) - Less manual labour, more intellectual - The same: you don't gain weight Intake: increases with 250 calories in first period Afbeelding met tekst, schermopname, Kleurrijkheid Automatisch gegenereerde beschrijving - Experiment - You and your clean eat the same, but one eats a twix bar every day and the other doesn't - Divide calories by 10 - 250 calories a day: 25 pounds more each year (= 11,25kg) *Why do people consumer more calories?* - Caloric density of food had risen - Foods with high caloric density are cheaper - Whatever you bring in your house: you eat (innovations: a lot of new products) - Larger sizes of f.e. Coca Cola - People eat until things are gone, they don't have an internal meter of how much calories they've eaten - Portions have grown bigger (eat all: eat more)  - Estimate how many calories you eat - If you don't eat much calories: a good estimate - If you eat more calories: gap between actual calories and estimate is big *Key lessons* - "details" matter: - Even small behavioral changes can lead to substantive effects - Stimulate brand sales: a few bites more make a difference - Cognitive and physical limitations matter: - Consumers are to minimize effort - If its easy they do it more - Increasing availability: do it more - Consumers often make biased inferences about the consequences of consumption - Easier to think about: do it more *Cognitive reflection test (CRT)* - A bat and a ball cost €1.10 in total. The bat costs 1 more than the ball. How\ much does the ball cost? - If it takes 5 machines 5 minutes to make 5 widgets, how long would it make 100\ machines to make 100 widgets? - In a lake there is a patch of lily pads. Every day the patch doubles in size. If it\ takes 48 days for the patch to cover the entire lake, how long would it take for\ the patch to cover half of the lake? - People believe their intuition is right: to much effort to check - If you were to give them money: they would think longer *When is accuracy high?* Afbeelding met tekst, schermopname, Lettertype, lijn Automatisch gegenereerde beschrijving - Imagine a family that needs two cars and owns two cars\ One model gets 10 MPG (Miles Per Gallon) and the other gets 20 MPG. Both\ cars travel 10,000 miles a year.\ The family has enough money to replace one model with a more-fuel-efficient\ vehicle to reduce costs and the family's carbon footprint. What is your recommendation? - Replace the 10 MPG vehicle with a 20 MPG vehicle - Replace the 20 MPG vehicle with a 50 MPG vehicle - Answer: 10 MPG to 20 MPG - Majority says B - They want to know gallons per mile (metric per gallon is really important) - How much are you willing to pay to increase your internet connection  - Domino's pizza: different sizes - You eat the more pizza when you eat two regulars instead of XXL - 1 large pizza is slightly larger than 2 regulars - You consumer more calories *Weighted additive rule* - You get information about the attributes - We have a hard time mapping out attributes to consequences - Attribute information: deliberate decisions that companies make *How would you rate ground beef that is 25% fat/ 75% lean?* Afbeelding met tekst, lijn, ontvangst, Lettertype Automatisch gegenereerde beschrijving - Beef that is 25% fat is seen as worse *Which product would you be more likely to buy?* - Premium bandages - Bandages have bubbles in the padding - Waterproof - 25 per pack - €8,00 - How would you communicate the extra benefit? 1. You just say: "Bandages have bubbles in the padding." (**no mechanism**) 2. You say: "Bandages have bubbles in the padding. The bubbles increase air circulation around the wound, thereby killing bacteria. His causes cuts to heal faster." (**shallow mechanism**) 3. You say: "Bandages have bubbles in the padding. The bubbles decrease contact with the wound allowing air to circulate. Oxygen in the air kills many bacteria causing the wound to heal faster." **(intermediate mechanism**) 4. You say: "Bandages have bubbles in the padding. The bubbles push the padding away from the wound, allowing air to circulate. Oxygen in the air interferes with the metabolic processes of many bacteria, killing them and allowing the wound to heal faster." (**detailed mechanism**)  - Generic bandages - Waterproof - 25 per pack - €4,00 *S&P 500 company decisions* - How well do you understand what a company does? - How risky would you rate company's stocks? - Correlation/relationship between sense of understanding and perceived vs. objective risk: Afbeelding met lijn, diagram Automatisch gegenereerde beschrijving - Negative correlation: when people feel they understand what the company does, they believe it's less risky to invest in the company - Read companies description: you have to predict what will happen to stocks in the next year, look at: - Standard deviation - Much lower when sense of understanding is high - Average - Higher when they understand what company does - Relationship between sense of understanding and mean of (predicted) returns - Relationship between sense of understanding and mean of (predicted) returns *Question* - Genetically modified foods are generally... to eat - Safe - Unsafe - I don't know enough to say - Data Afbeelding met tekst, schermopname, nummer, software Automatisch gegenereerde beschrijving - Where does this come from - Mnm's: partially created with genetical engineering - People who more strongly oppose genetically modified foods feel more knowledgeable  - People who less strongly oppose genetically modified foods are more knowledgeable (seen in scientific literacy) Afbeelding met tekst, diagram, lijn, Perceel Automatisch gegenereerde beschrijving - Sense of understanding drives decisions *How to reduce polarization?* - Ask people to explain causal mechanism of certain things - Much less certain *Explanation reduces extremity*  - How strong do people oppose against something before and after they give an explanation **Customer-based brand equity** *4^th^ meta-goal of choice processing* - Products: bundles of attributes (motivation: what are best bundles of attributes, the higher motivation and ability to process information → the more consumers are going to be able to complete their goal) 1. Maximize decision accuracy 2. Minimize the cognitive effort needed to reach a decision 3. Minimize the negative emotion experienced while making the decision - Minimize uncomfortable feeling when choosing 4. Maximize the ease with which a decision can be justified - How easy is it to justify your decision (to yourself and others) *Which vacation spot do you prefer?* Limit amount of information: which one do you want to book if you are planning a week vacation in a warm spot over spring break? - Spot A - Average weather - Average beaches - Medium-quality hotel - Medium-temperature weather - Average nightlife - Spot B - Lots of sunshine - Gorgeous beaches and coral reefs - Ultra-modern hotel - Very cold water - Very strong winds - No nightlife - 67% says B, study in influencing paper Limit amount of information: which one do you want to cancel? - 48% says B - Why: spot B is easier to justify - 3 arguments/reasons pro B - 3 arguments/reasons against/con B - Weight people put on attributes depends on whether the attributes help them justify what they want to find *Do you continue searching?* First choice set: - \$290 a month, 25 minutes from campus - \$350 a month, 7 minutes from campus - If you keep looking apartments might not be still available - Difficult to compare the two apartments, not 1 choice is dominated - 64% continues search Second choice set: - \$290 a month, 25 minutes from campus - \$330 a month, 25 minutes from campus - More should continue their search, but only 40% continues search - Justification: they stop because the first option dominates the second option *Referendum* - Reason survey: execute will of population - Version 1: do you support or oppose reducing the voting age from 18 to 16? - 37% supports - 56% opposes - 7% no opinion - Young people don't get to vote - You make a change on something (feels like a loss, deviation from status quo) - Learn what people really want (inferring from behavior what people want) - Version 2: Do you support or oppose giving 16- and 17-year-olds the right to vote? - 52% supports - 41% opposes - 7% no opinion - You give a reason (yes they should be able to vote indeed) - Feels like a win: you give them a right - Illicit behavior in different ways: same behavior but preference varies just because the question is asked differently (context changes) Afbeelding met tekst, Lettertype, schermopname, lijn Automatisch gegenereerde beschrijving - Fundamental attribution error - Kurt Lewin: I can ride down an equation to model human behavior - Behavior is function of latent/unobservable wants/desires/believes in a person and context - MARKETING: observing behavior, don't forget the context (it's not always what the true latent preferences are) → but there are no true latent preferences in people because there is always a context (they are constructed o, the spot) - Market research: says something about people but also the instruments that are being used - Conjoint analysis: people are presented with multiple bundles of attributes, they have to choose which they want - Regression Model: regressing people's choices on bundles of attributes they saw (context dependent: you invite people) → different bundles, attributes, weights gives a different attribute weight *Survey* - How many hours a week do you spend reading or listening to the news? - Up to 1 hour - 1 -- 2 hours - 2 -- 3 hours - More than 3 hours - Few say \> 3 hours - How many hours a week do you spend reading or listening to the news? - Less than 3 hours - 3 -- 4 hours - 4 -- 5 hours - More than 5 hours - Many say \> 3 hours ! careful to interpret datapoints in absolute way instead of relative way ! - Context shapes behavior *Video* - Disturbing situation: dragged off plain because flight is overbooked - If you are already on the flight: nobody wants to get off → removing people from the plain with security - Study: - Have you heard about United recently? - Have you not heard about United recently? - Same question to everyone: \$204 united flight for nonstop or \$204 American airlines flight nonstop - Not heard: 50/50 - Heard: 21 united/79 American - Same question to everyone: \$204 united flight for nonstop or \$270 American airlines flight 1 stop - Not heard: 86 united/14 American - Heard: 56 united/44 American - Event might be activated (availability bias): large influence on answer + social component (do you care about what happened to those people?) + you don't actually have to do it (easy to say → may be different if you do have to pay) + people forget *Modern family* - Trojan horse technique/opposite from foot in the door technique/ door-in-the face technique: you ask for something big, so you certainly get the smaller thing you want ***Door-in-the-face technique*** - Version 1: will you volunteer to chaperone juvenile detention center inmates on day trip to zoo? → Few say yes - Version 2: will you volunteer to serve as unpaid juvenile detention center counselor for two hours a week for the next two years? → everybody says no If you cant do that*...* will you volunteer to chaperone juvenile detention center inmates on day trip to zoo? → many say yes - Setting high versus no reference point (less bad) - Element of reciprocity: people want to reciprocate - Social dimension: somebody did something for you, you want to do something for them aswell ***Foot-in-the-door technique*** - Version 1: will you place a large sign in your front yard reading "Drive carefully"? → few say yey - Version 2: Will you place a small sign in your front yard reading "Be a Safe Driver?" → Almost everybody says yes Two weeks later: Will you place a large sign in your front yard reading "Drive Carefully"? - Many say yes - Social element: if you don't reciprocate relationship gets stuck - Element of consistency: show consistent or inconsistent behaviour *Kahneman* - Followed over period of time: sometimes they get asked a question - Where are you? - Who are you with? *Would you complete this survey? It's 40 questions long and will take you 15 minutes.*  - Response rate 8%: no incentive - Red: if you return survey we will pay you \$1 (non-transactual relationship) - Non-contingency incentives: you get money before hand *Groupon* - Attractive deals: if you buy it, you can exchange it for a deal 1. **Scarcity**: reason to find it attractive (other people already have it, you want it to) - Limited time left to buy 2. Follow others (**social proof**) - Information of others: gives indication of value it has (risk reduction) - Element of conformity: I need to do it too to belong to the group - Risk reduction and conformity 3. **Consistency**: Want more deals like this: like the heart below → email with new deals (consistency, more and more consistent deals) - Consistent in behavior: inconsistency not rewarded 4. **Liking**: Refer friends: get \$10, your friend Bart recommends you looking into this (if friends ask you to do something: more likely) - Easiest way to get people to like you: do something they like (copy behavior → more likely in hierarchical situation) 5. **Reciprocity**: you return favor if somebody does something for you 6. **Authority**: more likely to do something if somebody with authority tells you to do so - Principes powerfull to chield yourself/have larger influence *Tax compliance* - Financial arguments for paying your taxes/ filling taxes on time/ accrual declaring your income on tax forms: - File jacked: fine, jail,... - Probability of getting caught: low - Expected value: higher - Social reason: - Contribute - Reasons for not filling: - Busy → forgetting - Every governance struggles with people not paying taxes - Easy to implement letter - Reminder letters send to population with tax residence - Short letter (good) - Which is most effective: country norm + minority norm - Country norm: response rate 33% - Country norm + minority norm: - Other lines to add: *Homo economicus vs Homo Sapiens* +-----------------------------------+-----------------------------------+ | Homo Economicus = rational | Homo Sapiens = predictably | | | irrational | +===================================+===================================+ | - Archeology metaphor | - Architecture metaphor | | | | | - Well-defined pre-existing | - Preferences are constructed | | preferences are uncovered | based on the decision context | | | | | - Choices are based on the same | - Choices are based on a wide | | invariant process | variety of heuristics (mental | | | shortcuts) | +-----------------------------------+-----------------------------------+ *What is the most valuable brand in the world?* - ≠ not company - Maybe Apple, Tesla, Unilever owned brands,... *Interbrand:* 1. Apple (brand value: 482 215 000), 18% increase 2. Microsoft (278 288 000), 33% increase 3. Amazon (274 819 000), 10% increase - Value comes from knowledge that resides in everyone's minds - Putting a number on value: difficult - Ask experts - Consumer studies - Elaborate methodology - Numbers precisely wrong *Kantar brandz:* 1. Apple (947 062 000), 55% increase 2. Google (819 573 000), 79% increase 3. Amazon (705 646 000), 3% increase 4. Microsoft (611 460 000), 49% increase - Numbers all over the place (in the billions, yes) *Brand finance: global 500 2022* 1. Apple 2. Amazon 3. Google 4. Microsoft - Not helping us understand much about business - Why different? - Different equation/formula - Estimate variables is different *Drake equation* Afbeelding met tekst, handschrift, Lettertype Automatisch gegenereerde beschrijving *"If this company were to split up I would give you the property, plant, and equipment and I would take the brands and the trademarks and I would fare better than you." - John Stuart* - What makes brand valuable: definitely knowledge *A brand is said to have positive brand equity if consumers react more favorably to the product, price, promotion, or distribution of the brand than they do to the same marketing mix element when it is attributed to a fictitiously named or unnamed version of the product or service.* - Brand more preferable: if 4P's are good - Differences in responses - Because of logo, they are willing to pay more for a brand they know/like - Logo reduces price sensitivity  - Brains lives in consumer minds - Brand: real-estate in consumer's mind (it will decay after a long time if you don't invest in it) - Brand knowledge: - Brand awareness - Brand recall - Brand recognition - Brand image *Associative memory* - Information nodes are stored in memory and are connected by links (long term memory associative in nature) - Eclectic network of information - Emotions,... - Activation spreads through the network based on the strengths of links - King connected to queen - Ex. frozen entrees by Colgate: fail - Failed because people associate Colgate with a taste of toothpaste → doesn't make you hungry (failed because of brand image, incompatible with food consumption) - Why did they do it? think more about Colgate/brushing your teeth also when they eat (brand awareness) - Information nodes with highest activation are most likely to influence behavior - Brand try to find a desirable place: different form others, activated a lot, activates other things that have desirable \... in behavior *Limited attention* - Consumer's attention is limited and selective - Break though attentional filter to get to long term mind - Colgate: teeth most important, first thing you see - Don't see phantom hand or missing ear - Blocking out everything else - Two different pictures: - Big engine appears and disappears - Shadow of the plain - The box on the boat - Either you see it or you don't *Brand recall* - Think of chocolate: - Lindt, kitkat, milka, snickers, cote d'ôr, kinder, tony chocolony, Leonidas - Who & when? - Recall depth: who? - The percentage of people who recall the brand - Recall width: when? - The cues that lead to brand recall - Break → kitkat/coke - Subcategories - Places - People - Time *Brand recognition* - Do you remember having seen this brand of chocolate? - Recognizing something is easier than recall (recognizing is generalizing) *Brands compete with ingredients in your mind to explain product performance* - Ex. Headache → dafalgan/paracetamol (identical) - Paracetamol inside Dafalgan - Paracetamol is the thing that makes the headache go away *Non-product-related attributes compete with product-related attributes*  - The higher the income: the less likely they buy store brands - The more knowledge people have about how things work: the more likely they are going to buy the generic brand - Medical knowledge jobs: buying generic brand **Gast lecture -- Tom Meere (why five)** - Qualitative research for over 25 years - Founder and managing partner - Why five: strategic questions of big brands - How can you cut through what really drives consumers, and how you can use it to build constructive relationship *Why five* - Frustration: information hidden in jars and never used again - Interdisciplinary team: different teams/people - Work globally: with scientist,... - Business and marketing for number of brands (brussels airport, vrt, telenet, Delhaize,...) - Question: how can & should a premium chocolate brand behave in NYC - First understand what consumers are looking for - Streamz: what content should we now make/buy to win? - Business challenge, not research question - Trains: how to convince people to take the train more often? - Understand what people's needs are - HLN: how can the biggest news brand become also the most loved one? - Try to understand perspective of consumer *Choosing between 1,2,3 and 4* - People don't pick at random - People tend to choose something in between - Not much you can learn from this (first lesson of qualitative research: do not start counting when it doesn't make any sense) *"People do not say what they do, and they do not do what they say" -- E. Dichter* - How much do you brush your teeth? 2 times → consumed toothpaste is lower - Why did you buy a BMW? Will be about status, but is only what you find under first layer (what's under need first layer of information → not easy) *What motivates people to act?* - Case - Shows how you can use a collage technic to start understanding what people are looking for - Range of specialty beers without cannibalizing with main brand - Focus groups: when do you drink them? → mostly same answers - Italy - Men, 26 -- 35 - Monthly+ drinkers of specialty beers - SEC AB - Collage technic: express how you feel about specialty beers without using words, give a collage - Which dimensions do you see? - What atmosphere does this project? - First collage: oppervlakkige antwoorden eerst (luxury, relaxation,...) - Second collage: very different answers (own needs projected into images, you see what specialty beers means to them) - Third collages: similar behavior → variety of needs because of different people - Different ways of engaging with specialty beers (anchor points marketing campagnes) - Different types of beer - Fundamental understanding of consumers needs - How cultures look at different brands (case: make a collage about dairy beverages with fruit, it shows swimsuits, flowers, sea,...) *So what?* Challenge: deepdiving To do so, we need 2 things: 1. Ways to ask the right questions (methodical element) 2. Tools, models, methodes to help structuring and understanding responses (ways to make sense of what people are saying) *Why is PET (plastic) such a challenge in beer?* - It feels/seems cheap - Rationally: no reason why not to use it (better for beer than glass, only sustainability is a factor) - Consumer point-of-view: 2 questions hidden in question - What does PET mean to people? - What does beer mean to people? - Cold/refreshing, masculine aspect/image, social aspect - PET = opposite (heads up very fast, - Solutions: overcompensating *Markets today* - Increasingly cluttered - More products and brands competing for same pie - Differentiation: everything looks/is the same - Increasingly similar products: - Functionality - Quality - Performance - Even more and more design similarity - Information overload for consumers *The why is important because...* - It ties consumer knowledge - It makes positions coherent and relevant to consumers - It makes otherwise descriptive information actionable - More than 90% of people's decisions are emotionally based (at least partially) - Tackle the why do people do something - Reason to drink exotic fruit juices: bored with the rest, driving after,... → you don't know which flavor/packaging/price without the why - If you don't understand the why: you start AB testing but you cant reproduce success because you don't know the why *Always take a step back*  - If you really want to understand what Ikea can do to you, you have to understand what furniture does to people, what home means,... - Dettol: kills stuff (fear of contamination) - Launched shampoo: but what is the meaning of hair (contributes to identity, lifeforce) → not something you want to kill *Motivation bridges the gap between need and satisfaction* Afbeelding met Lettertype, ontwerp Automatisch gegenereerde beschrijving - Try to solve tension that you are confronted with in external world (trying to build a bridge) - Hungry: look for food - Tensions that you are confronted with in the outside world: find solutions = overcoming the bridge - Marketing = building a bridge (what do they need, is there a match with existing brand or do we need to create a new product) - Mobile phone: created need? (underlying need has always been there, need to be updated, relate to other people,... → need itself is very universal/stable) - Segment about status: big pick-up trucks, lot of chrome, boosting ego - In Europe: underlying need manifestation of ego: expression was different (much more understated, more modest) - Bridge gap of what people need and what is available *Brand have satisfaction and relational potential* - Relational potential: play role in people's social contexts *Iceberg*  - Under iceberg: decisions made & long term needs - Need to find a way of how decisions are made there *Behavior: adaptive energy investment which serves to reduce the tension between internal and external world* - Every behavior that we pose as humans is an energy investment to reach goal - Tension: if you are hungry, you eat *Birth* Afbeelding met tekst, schermopname, Lettertype Automatisch gegenereerde beschrijving - Before you are born: everything happens automatically (perfect) - Born: fun is over/starts, source of needs is separated from source of solutions - Cry or smile to get what you want - Older: do something to get what you want, set off strategies are different (complex strategies: need for comfort/recognition) - The more primary solution is as marketeer: the more successful it will be - Ex. Prepacked soup: variants closer to home → more successful *Model* 2 dimensions - Vertical: individual experiences of people, how they look at life - Horizontal: social, want to stand out or fit in/belong - Apposes very different attitudes - Giving kid banana: release and control: just give it or cut it up - Many different contexts: - Top left: energy, progress, exploring things - Top right: togetherness, sharing stuff - Right bottom: fear driven, safety and authority - Left bottom: differentiation - Redefine in function of context you are working in *Dairy in the model* - Clearly: group of people or occasions where function of milk is really important, but also getting energy of it, products that empower you - Different meanings of diary: platforms to start building brands → 2 brands here (caring aspect and functional aspect) - Opportunities: shift brand from one to another (communication to make brand more relevant) - Model: is tool, not purpose of study (helps you think about situation in the market) - You can quantify it (size segments → strategic management tool) - Yogurt: - Pro fuel: energy - Yo crunch: childish, more playful - Natural: rest - Premium: sophisticated *Color model* - Color variations also linkable to underlying needs and motivations *Commercial Volvo* - Building it on safety - Capitalize on overall feeling/need of people to feel safe - Did it in the best way: choice version of a song that talks about threats in the world but still explore in a safe way (implicit way of signaling fear and dangerous environment) *Linking occasions to motivations* - Linking motivations to indulgence products - Rows: occasions that people use products - Column: consumer needs - Dominant motivations: indicated by colors (give team of client toolbox for activation: choose if they want it in the morning → come up with right activation mechanism to deal with this) *Quantitative segmentation attitudes towards online shopping* - Start measuring incidents of motivations and adjust brands to it *Why a model-based approach?* - Differentiating anecdotical vs structural insights - Try to make difference between anecdotical/incidental information and structural information - Benefits: focus group with six people, chat about coffee → 2 was in traffic jam, are unhappy → affect how they behave in situation - Challenge subjectivity - Tunnel visions - Challenge completeness of analysis - Facilitate collaboration - Facilitate communication *Influencing behavior* - Others: doesn't help to influence people's behavior in certain way - Influencing behavior in direct way - Start understanding how we can change sometimes very stubborn and anchored behavior - Have to understand why, descriptive information is not enough (changing habit short term → probably not enough) *Four fields* - Marketeers only look at problem, promise whole new world (forget about the fact that people are comfortable with routine/what they already do) Understand this: - Comforts that keep people from changing - Fears of new behavior *Alula* - Try to put it on the map (place in Saudi Arabia) - Change people's behavior when start looking at new destinations - People expect from travels that they come back as a new person (a lot of value in traveling) - Different points-of-view in different countries - Not only paying attention to pains and gains, also comforts and fears - No clue what they are - Concerns as reason not to go: will I be able to have some alcohol? - More preformat: wasn't enough to know what people want in the experience, also looking at fears and take them away *Qualitative diagnostic research: approach & techniques* - Offline - 1 on 1 interviews - Interactive group sessions/focus groups - 1 on 2 interviews - Immersion participating observations - Online - Live conversations - Non-verbal information - Asynchronous conversations - Observations - In common: way we look at respondence ≠ cognitive beings - Power of projective techniques: bit less strong, more power in offline setting *Problems to deal with* - People are defensive - People are verbally limited - People are unaware or only partly aware of several influences/reasons Direct, straightforward questioning will not bring the solution You will get an answer Whether it is a truthful and relevant answer is uncertain - Example: Make food and health campagnes more effective: what words, what mechanisms, what pictograms make sense to them - You can not always express what fears and underlying needs are - Be aware of fact that there is only so much that people want to share or can share *The role of the interviewer is key* - Moderator is key - Authentic human contact works best - Goal: inviting people to share what is really important to them (only do it in safe space → attitude interviewer is instrumental, you cannot fake it) *The interviewer's attitude* - Different ways in different countries to moderate good - Europe: getting rid of table, living room setting (sharing) - Other countries: big table with one at the end of table - Make people feel safe, to make respondents feel confident, motivate respondents, obtain required depth *SEMI-structured conversations* - Indirect questioning, no closed questioning - If you ask why: people always answer (polite) → rephrasing questions - Doesn't always results in underlying needs - Why will trigger justification if you are not carefull *Techniques* Afbeelding met tekst, ontvangst, algebra Automatisch gegenereerde beschrijving - 'Feeling of security'- supporting technique → goal: transferring trust - Photosorts: collage of photo's (that are balanced on what they convey) - Select person that deals with category the same way you do - Pick picture: bring picture to life, talk about what they do and who they are, relation with product → create a context in which people feel free to talk about themselves - Deep techniques: fantasy technique, self-image projection  - Mistake: warming up exercise already giving all the depth you need (think what technique you are going touse for what purpose) *Unconscious vs spontaneous* Afbeelding met tekst, schermopname, menu Automatisch gegenereerde beschrijving - Personify brands - Different brands on table: these are families, describe them - Matter of interpretation: technique doesn't say the wright answer - Quality depends on quality of interviewer and analysis **Consumer learning about brands** *About guest lecture:*  - Brand recall: all the connections that come to mine - Brand recognition: do you know this brand? - Non-product related attributes and product related attributes: how do they compete in consumers minds to explain the benefits of a product - Competition against ingredient - The more active ingredients people know: the more they are going to buy generic brands - Need or desire of specific product have changed over time (means to accomplish goals differ) - Extend to which goals are activated in people's mind (intensity of goal) changes also → influenced by marketing (need to be connected is higher now) - Influence, not creation *Cloth diapers* - Pains: not convenient, got to wash it (effort), smells, requires skill - what do you show in first commercial? - Split screen (struggling parents with cloth diapers \>\< non-struggling parents) → not super effective (not persuasive for parents) - Are you sure it is going to work - High order emotional, don't take the easiest way (are you a good parent if you make easy choice) → benefit of child important - Lower price → emphasize it *Pyramid of value* Afbeelding met tekst, schermopname, Computerpictogram, software Automatisch gegenereerde beschrijving *Nike* - Lot of emphasis on technology and product and some benefits (lower level pyramid) - More recent commercials: not such a strong emphasis on this, more emphasis on higher levels of pyramid (is about identity, who you want to be as a person) → not a single product in it *Affiliation/belonging* - Aspirational group: a group you aspire to be - F.e. athletes (reclama: buy getorade, be like Mike) - Dissociative group: a group we reject and avoid being like - Used in advertising (immediate functional benefits that you don't like, as who do you see yourself) - Associative group: a group we currently belong to *Renova* - Portuguese toilet paper brand - Market share worldwide: basically 0% - Dominated by companies such as P&G, Kimberly-Clark and private labels (store brands) - Competition: large international companies and private labels *What can Renova do?*  - What do you do against private labels? - Difference more in mind than irl 1. Make a private label - Trader Joe's in US, produced by national brands - Private label value share in countries: - Switzerland: 52% - Western Europe: 36% - Private labels: in Europe more important, they care less about the brand 2. Compete with private labels, compete with them directly by lowering prices 3. Differentiating: what are different elements in what we can differentiate *Why weaker in emerging markets?* - Supply side factory - Stronger marketing support for national brands - Many small, independent stores - Retailers not organized/consolidated as much as in Europe - Europe: 60% - US: 30 -- 40% - National brand manufacturers are primary source of private labels in emerging markets - No companies that specialize in products that are brought into the market as private labels (or fewer than here) - Here many companies that focus on efficiency, high quality with low price - Demand side factors - Price-quality heuristic stronger in emerging markets - Information, education about ingredients that compose products, more transparency here in markets than there - Risk reduction function of brands more important - Social impression of brands more important *Why weaker in North America?* - Supply side factors - Stronger marketing support for national brands - The largest retailers account for smaller proportion of overall sales - Lager quality gap between national brands and private labels - Objective quality of private labels here higher than in US - Less consistent quality of private labels across countries and across product categories - Demand side factors - Perceived quality of private labels is lower - Brands play a more important role in the lives of American consumers *Only trademark that recurs throughout the store* Afbeelding met tekst, schermopname, Lettertype, lijn Automatisch gegenereerde beschrijving *Why private labels so strong?* - Control over point of sale - Private label: competitor but also collaborator (engage with competition) **\ ** **Consumer learning about brands (cont'd)** *Benefits* - Functional benefit: f.e. toothpaste: teeth whitening - Emotional benefit: f.e. toothpaste: crocodile on package (more fun to use) - Self-expressing benefit of toothpaste: healthy person *Why would you not buy store brand?* - Because labels attract you (familiar: seems better) → Price-quality heuristic - Symbolic or social function - Symbolic: you have always used it at home - Social: what will other people think - Psychological reasons *Renova: Portuguese producer toilet paper* - Competing with private labels and international companies (very good at nurturing brands) - Strategic direction 1. Increase collaboration with retailers: produce private labels 2. Compete with private labels: lower price 3. Accelerate differentiation (focus on innovation) - Element of value higher in the pyramid - Experiencal benefits *Why are private brands so strong?* - Competitor and collaborator - If you compete to strong you will become a collaborator - Everywhere in the store (in every aisle) - Private labels: increasingly sophisticated brand architectures - Now: higher quality and higher prices *How to grow your brand?* - Line extension: you have an existing brand and introduce new products in same category f.e. Coke and Diet Coke, Michelin tires → winter tires - Extend brand in same category - Category extension: generalizing same brand into new category f.e. Nutella making cookies (spread → cookies) - Drivers category extension: - Product feature similarity: consistency in product features (shoes and tires: soul rubber, same material) - Brand concept: - Extend brand in new category *Brand associations* - Rolex: (brand image: expensive/luxury/prestige) - Rolex Calculator - Feature similarity: precision (functional association) - Brand concept: maybe prestigious - Rolex Bracelet - Feature similarity: yes - Brand concept: prestigious bracelet possible - Rolex Flashlight - Feature similarity: lower than others - Brand concept: not prestigious - Rolex Cologne fragrant - Feature similarity: lower than others - Brand concept: - Timex: (brand image: functionality) - Calculator fits better than bracelet - Flashlight better than cologne - To what extend is there fit between two categories in terms of brand concept/core association *New brand active in same category* - House of brands: company that owns multiple brands, they can compete in same category, but can also be in different categories - Exist in isolation in consumers' minds - Allows you to build brands with precise position - Ex. Procter and Gamble, Unilever *New brand active in new category* - Branded house: same brand in many categories - Ex. Virgin (Cola, Music, Airlines,...) *Brands* - Sub brands - Ex. Diet Coke: important driver to buy it is Coca Cola, but brand is different - Endorsed brands - Nestle: house of brands, not as (they hold the 'Nes') = endorsed brands - Buy it mostly because it is 'Nescafé', buy it a little bit because it is 'Nestle' - Carrefour: - Carrefour Discount - Discount is displayed very small - Cheapest - Carrefour Bio - Bio is displayed very big - Differentiate - Selection carrefour: - Selection displayed very fancy - Carrefour: very small - Higher quality, higher price **\ ** **The art and science of effective advertising -- Sorin Patillnet (Marketing Engineer) -- Mars** *Why don't we grow while getting so many awards?* - Different category *What is effectiveness in advertising?* - Ability of a communicating message to make a change in behavior consumer to drive a purchase - On anything that's quantifiable and drives business *Creativity is number one driver of sales impact* - Power of creative idea is what makes or breaks campaign - Half of opportunity is lost if this isn't optimal *Effective creative (initial spark)* - Goals - Impact memories - Drive sales - Start at stimuli and end in mind (build brand in mind, so that when you are purchasing you choose this brand) - Smoothing path from stimuli to brain 1. Attention - Physical mechanism that allows us to concentrate on one mechanism at expense of everything around you - No attention: chain breaks 2. Emotional advertising - Has higher impact - Fuel that makes process go faster 3. Memory (easy of memory) - Easier for brain to process *Measure effectiveness* - Facial expressions for attention - Link: watching ad - Camera that analysis what the eyes do, how you react - Neuro testing for emotions - Measure brain activity - Happens in lab, very complex - AI eye tracking for memory - To simplify, where eyes are going to look in an ad *What works* - Attention: - High attention is a market for high ad impact - Capture consumers' attention at the start - Capture attention in first second, hold it during duration ad - Avoid build tension, it's not a movie - What happens at second 26 - Emotions: - Generate positive emotions early - Increase emotions throughout the ad - Solve negative emotions fast - Memory (mindful how we function as human beings): - Make it easy for the viewer to follow - Avoid multiple storylines - Use your brand imagery very often integrate it - They need to know the content (ex. Pepsi: brand color) *How can we improve?* - Get platform basics in place: different apps with different behaviors, where your finger is on the phone - Make ads fit for the platform you plan to use - Use subtitles always on to maximize audio comprehension - Growth in subtitles, helps people with hearing problems to follow the message - The first seconds matter the most, grab attention early *Overemphasizing branding, make them feel something* - Integrate branding in everything we do: Coca Cola, people happy in form of bottle - Move me, entertain me, make me feel something - Participating is winning (Nike) - Advertising a bit like standup comedy (M&M's) - Solves a negative emotion (Dove) - Advertising is a mix of science and art - AI will help you create a creative idea but will not do it for you **Gast lecture -- Steven Van Steen -- Tony Chocolonely** *Information* - The Fortune Seller -- Sales Manager Belux - 6.8 billion kg of chocolate is eaten each year - Present in over 60 countries - Farmers in Africa are not being paid fair (no living income: \$0,17) - Ghana - Côte d'Ivoire - Farmed by 2.5 million farmers - 1.56 million children - Official numbers - Child labor - The industry's average: 46.7% - Tony's Open Chain: 4.4% - Actively looking for these examples and minimizing them - Deforestation (to grow cocoa, 1/3 because of cocoa) - 0% at Tony's - Imbalance (inequality) *Teun van de Keuken* - Journalist -- founder of Tony Chocolonely - Harkin Engel Protocol: non-binding agreement to agree that there would be no child slavery - Couldn't change industry so made a company itself - Chocolonely: started fight by himself - Red package because situation is alarming *Roadmap (together, we'll end exploitation in cocoa) -- Open chain model* - Create awareness - Bar: marketing tool to tell story - Not one single piece is the same: not equally divided - Inside: tells story - West-Africa is hidden in bar - Once you know, you can't unknow - Campaign: pay farmers not lawyers (campaign: not purple, obviously) - Lead by example (5 principles) - A higher price - Paying living income price - Factors: how many children do you have, average income household, cost farming - Traceable cocoa beans - From bean to bar: shipping - All cocoa bean traceable - Strong farmers - Stronger voice, to speak up - Improved quality and productivity - To become better farmers - Higher turnover per square meter - The long-term - 5-year contract (time to grow cocoa) - So that they can invest in a brand - Inspire to act - They broke \#1 rule in business: don't share your USP (unique selling position) *Tony's Open Chain* - Invite big cocoa company's - Examples: Aldi, Colruyt, Ben&Jerry's, Albert Hein, Hema,... *Chocolate in Belgium* - We eat 8.5 kg per capita in Belgium - 74% of bars - A lot of bonbons, pralines - Brussels Airport: most sold - Tony's 4the place, 1ste place is Côte d'or - Brand awareness: ¾ Belgians know brand Tony's Chocolonely - Reasons why people buy chocolate? - Great taste - Brand I trust - To spoil myself - Moments? - Me time - Satisfy craving - Daily tea/coffee complement - Pretty conservative with flavors: praline or regulars - Brand awareness - Find associates - Distribution - If you can't find us, you can't buy us - Impulse buying - TV campaigns - Consumer day: selling 523 068 bars (you got them for free), almost 5% of Belgian population - Increase in brand awareness - Types of brand awareness - Aided: do you know the brand if I show you the package? - Unaided: do you know Tony's Chocolonely? *What would you do to change consumer behaviour in Belgium?* - Do survey: put Belgian Flag on package to promote Belgian Chocolate - Chocolate milk: pitch on Christmas market - Downplay others by presenting children that are picking up the other brands *What would you do to increase brand awareness for Tony's?* - Spread awareness by showing the kids - Putting up murals (with children and farmers) so people can take pictures with it - Giving information while people are tasting your chocolate - The chocolate you get in coffeeshops (Madmum, Starbucks) **Customer heterogeneity** *Sub brand and endorsed brand* = continuum (extend to which modern brand is driver of purchase decision, how important is modern brand) -- house of brands, sub brands, endorsed brands, branded house - Endorsed brand: modern brand less dominant than in sub branding - House of brands: modern brand is strong driver *Alternative based choice processes and attribute-based choice processes, which heuristic to use* Brand 1 Brand 2 Brand 3 Brand 4 ------------- --------------- --------- --------- --------- Attribute 1 Utility score......... Attribute 2............ Attribute 3............ Attribute 4............ - Alternative: look at brand 1 and look at all the attributes (first: column, second: row) - Attribute: look at attribute and look at all the brands - Lexicographic: what brand scores the highest on the most important attribute - Weighted additive rule: you take into accounts either all the brands or all the attributes (attributes do not have the same weight, some more important) - Overall weighted utility: some weigh more - Equal weight heuristic: all attributes have same weight *Compensatory and non-compensatory choices* - Compensatory: poor choices on one attribute can be offset by better choice on other attribute - Weighted additive rule - Equal weight heuristic *Exam questions* - Kahneman (Nobel price), Tversky (prospect theory) are names we need to know 1. Value is relative to reference point, value function S-shaped (probability value function: inverse S-shaped) and value function is steeper for losses than gains (loss-aversion) → A 2. If you build a dataset of products across product categories and would look at how they would perform in standardized test: correlation is quite low (0.15) → A - Reduced correlation is not the case (has been 0.3 for a long time) 3. Laddering because they try to persuade people to be a good father/mother (higher order benefits) - Laddering: the idea that you ladder up products to higher order emotional, social, symbolic benefits that consumers might care about - Predictably irrational decision-making: complexities of human psychology - Rational decision making: the ideal choice - Social proof: harnessing signs of persuasion (1 million people have bought this book, the idea that many other people value this product) - A typical ad: inform, persuade and build brand awareness *What can renova do?* ![Afbeelding met tekst, schermopname, lijn, diagra