Customer Heterogeneity PDF

**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 -...

**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, diagram Automatisch gegenereerde beschrijving - Private labels: very good at providing good quality with reasonable prices (high quality over price ratio) -- hard to compete - 3 strategies: 1. Differentiate - Differentiation from laddering pyramid perspective - Functional benefits: softer, cleaner, 4-ply, 5-ply - Problem: run out of ideas, difficult to continue + if you come up with something: competitor copy (not sustainable) - Emotional/symbolic differentiation - Position brand around well-being, beauty and style, worked because of bald move (black toilet paper roll) - Various packaging of the black toilet paper rolls - Appeared in American shows → awareness - What can we learn from this? - What budgets do they have for marketing: not a lot → generate awareness for brand in low-cost way - Cannot be perfectly anticipated - Marketing research has limitations, in terms of extend how product will be preforming when you bring them up to the market (low-cost strategy) - Why did you buy product? Uninformative things, not out of the box suggestions - Identity changes: view of competition changes after introduction black roll 2. Reduce price - Tricky when margins are already slim, increase in volume needed to make profit is too big 3. Make private label - Lower to level of competitors *Renova -- Key learnings* - Recognize the limitations of traditional marketing - "Laddering": Think beyond functional attributes such as price and quality. Renova added experiential and symbolic benefits to toilet paper. Renova does not sell toilet paper. It sells wellbeing and decoration. - The power of brand awareness. Get yourself noticed! *Customer based brand equity model* - Strength of brand associations - Head on: apply to forehead and headache goes away (terrible ad) - Uniqueness of brand associations - Audi: ad reinforces information that consumers have about brand; does it change peoples' perspective on the brand - This ad works because you look at the whole ad - Do I think this ad will lead to behaviour change? *Which pen do you think writes better?* - Regular or regular + laser Regular one because it focusses on one aspect - Delusion effect: people find it more believable that one means serves one goal instead of multiple goals - Strength of individual associations: lower *Typical ad* - So many typical ads (easy to process, category recognition happens almost directly) - When you advertise for your brand, you also advertise for competition - building primary demand: category demand, how much they buy in the category, about overall sales - Secondary demand: given primary brand, how much do they buy from each brand (about market shares) Afbeelding met tekst, schermopname, Lettertype, logo Automatisch gegenereerde beschrijving *Original ad* - You don't immediately realize it is an ad, or what the brand is for - Top-down, endogenous attention (original ads) - You want to understand - Bottom-up, exogenous attention (typical ads), more desirable when limited time - Fluent to process category or brand - Originality increases attention and memory for the brand and pictorial - Especially when ads feel familiar *Dumb ways to die* - Safety around trains - Not ideal in conveying message it wants to convey - Brand information only comes at the end - Post Brand Entertainment and Pre-Brand Entertainment *"What information consumes is rather obvious: It consumes attention of its* *recipients. Hence a wealth of information creates a poverty of attention, and a need to allocate that attention efficiently among the overabundance of information sources* *that might consume it." -- Herbert Simon* *Common dimensions for customer segmentation* Afbeelding met tekst, schermopname, Lettertype, lijn Automatisch gegenereerde beschrijving - "Job-to-be-done": why did you buy it (fresh taste, not smoke a cigarette,...) - You can't really measure satisfaction *Pink tax* - Woman spend 200 000 euro more than man in the span of their life  - Identify similar products that are directed to men or woman: looking at what costs more Afbeelding met tekst, schermopname, nummer, Lettertype Automatisch gegenereerde beschrijving - Undesirable - What can we ask ourselves? - Are we comparing pink apples with blue apples or pink apples with blue oranges? Are products really the same? Is formulation of male and female shampoos really the same? More expensive to develop? - Philosophical point of view: woman experience more benefits from a shower (feeling more confident) - More expensive to market product to woman than to man? - Have the products been sherry picked? Sampling process needs to be unbiased - Non-selected set of products(categories): no evidence, 3/5 products for woman are cheaper *Underpaying men* - What can we ask ourselves? - Who did the study? How did they study it? - Comparison between two categories: you highlight one dimension (are they the same on other dimensions?) - What might have happened? - 1 through 9: talented - Guys apply when they feel higher than 4, females only apply when they feel higher than 7 - Should be payed more money because they are more qualified *Are police officers racist?* - New York Times: officers in Houston were about 20 percent less likely to shoot if the suspects were black - What can we ask ourselves? - Probability that an officer shoots, given as a person is a suspect How do you become a suspect? - You cannot assume that two groups are the same on all dimensions *Pain of paying* - Example: Following is a scenario describing the behaviour of two shoppers. After reading about each shopper, please answer the question that follows. - **Mr. A** is accompanying a good friend who is on a shopping spree at a local mall. When they enter a large department store, Mr. A sees that the store has a \"one-day-only-sale\" where everything is priced 10-60% off. He realizes he doesn't need anything, yet can't resist and ends up spending almost \$100 on stuff. - **Mr. B** is accompanying a good friend who is on a shopping spree at a local mall. When they enter a large department store, Mr. B sees that the store has a \"one-day-only-sale\" where everything is priced 10-60% off. He figures he can get great deals on many items that he needs, yet the thought of spending the money keeps him from buying the stuff. - In terms of your own behavior, who are you more similar to, Mr. A or Mr. B? - Answer: - Emotion you experience when you need to pay for something - Rational: if I am spending this money, I cannot pay this money for future utility - Consequentialist models of decision-making take into account expected emotions, but people also experience immediate emotions at the moment of choice - For instance, people may experience an immediate "pain of paying". Some people ("tightwads") experience this more than others ("spendthrifts"). Afbeelding met tekst, schermopname, diagram, lijn Automatisch gegenereerde beschrijving *Study: ground truth (300 men)* - Psychographics - Openness to experience - Agreeableness - Extraversion - Clustering algorithm (k) - Demographic - Age - Marital status - Number of children - Customer preferences - Which slogan? - Nike Running: - Nike Running: - Which shoes? - Traditional - Innovative - Which car? - Which travel package? - Which investment? 1. Receive summary information 2. Determine segment names 3. Create personas 4. Predict preferences - Prediction accuracy - For each decision, we computed, - The predicted difference between segments: P~Diff~ = P~A~ -- P~B~ - The true difference between segments: T~Diff~ = T~A~ -- T~B~ When then computed the prediction error of the differences: \|P~Diff~ -- T~Diff~ \| - Condition 1: Info + Name + Persona - Condition 2: Info, then 4 (then 2, then 3) Afbeelding met diagram, Perceel, lijn, schermopname Automatisch gegenereerde beschrijving *Product specific observable thing: RFM segmentation* - Recency: when was the last time you came to the store