Marketing Strategy Based on First Principles and Data

Marketing Strategy: Based on First Principles and Data Analytics Research Toolbox © Palmatier 1

Agenda n Introduction n Research Methods and Data n Marketing Engineering Analysis n Markstrat Simulation (Part II) n Marketing Strategy Project © Palmatier 2

Is Marketing an Art or a Science? © Palmatier 3

Marketing is Transitioning from an Art to a Science n Global 1, 000 companies spend over $1 trillion on marketing, but only about 16% use marketing analytics One unit increase in use of marketing analytics (7 point scale) led to a 8% increase in ROA (70 M$ in net income) for Fortune 500 firms Same change in competitive markets where customer needs are changing led to 21% increase in ROA (180 M$ in net income) Different study shows marketing productivity increases by 5 – 10% with minimal additional costs n “You can count on us to combine a strong quantitative and analytical culture with a willingness to make bold decisions. As we do so, we’ll start with the customer and work backwards. In our judgment, that is the best way to create shareholder value. ” (Jeff Bezos in letter to shareholders) © Palmatier (Germann, Lilien, and Rangaswamy 2013) 4

Data Insights for Action Analysis © Palmatier 5

Key Steps in Marketing Research Define question or decision (importance, timing) 1. E. g. , How best to acquire new customers? E. g. , Should we launch this new product? Decide on research “methods and models” 2. Qualitative (e. g. , interviews, focus groups) Quantitative (e. g. , experiments, models) Collect data 3. Sampling and source (e. g. , survey, CRM, interviews) Perform analyses 4. © Palmatier Decision type & variables (e. g. , go/no go, tradeoffs) Analysis/model (e. g. , cluster, choice, conjoint) 6

Agenda n Introduction n Research Methods and Data n Marketing Engineering Analysis n Markstrat Simulation (Part II) n Marketing Strategy Project © Palmatier 7

Two Major Research Methods: Qualitative and Quantitative n Qualitative: used to study highly complex and contingent phenomena where linkages and variables are mostly unknown (e. g. , what is the next new product) n Quantitative: used to study “natural” phenomena where linkages and variables are mostly known (e. g. , what is the impact of advertising on sales) n Often portrayed as an “either or” decision, but is actually more of a continuum and methods are very complementary (ideally use mixed methods) See Market Research: Listen and Learn © Palmatier 8

Qualitative Research Methods n Inductive approach where theories/ideas emerge n Softer less structured methods (sociology, anthropology), small sample sizes, more open to researcher induced biases n Best early in the process to help refine ideas; when you don’t even know what to study (brainstorming, defining pool of attributes) n Process 1. 2. 3. © Palmatier Pick method (case study, interview, focus group, etc. ) Pick sampling approach (deviant, snowballing, etc. ) Code and analyze data for themes, frameworks, and theories (visual, summarize, NVivo, Atlas. ti) 9

Qualitative Methods and Data n Methods Case study: single business situation, great for story telling, poor generalizability, very popular Interviews: structured/unstructured, flexible, time effective Observation: slow, finds the unexpected, undirected Historical/literature review: easy and cheap start, dated n Data or sampling approach Deviant case sampling: In Search of Excellence Intensity sampling: Good examples but not outliers Heterogeneous sampling: Diverse to cover domain Snowball sampling: ask person for referral when hard to find exemplars Opportunistic sampling: easy and quick © Palmatier 10

Qualitative Exercise n Research question: Why do people get an EMBA? n Process Pick method: Semi-structured interview Start broad and open ended n Why, when, what, where n Ask follow-up questions (follow the thread) Pick sampling approach: intensity sampling Code and analyze data: find themes and summarize n n In small groups: take 5 minutes to conduct interview to “answer” research question © Palmatier 11

Quantitative Research Methods n Deductive approach where theories/ideas are tested n Data and analysis focused, larger sample sizes (psychology, science, engineering, economics) n Best later in the process to select among proposals and fine tune or test ideas found during qualitative stage n Process (depends on question) See Practical Regression: Convincing Empirical Research in 10 Steps 1. Pick model/analysis (e. g. , experiment, multivariate, choice model) 2. Pick sampling approach and data source (e. g. , random, stratified, survey, CRM) 3. Analysis © Palmatier n Check and cleanup data (make sense, outliers, problems, etc. ) n Run models and interpret n Sensitivity, robustness, holdout 12

Quantitative Analyses/Models n Data reduction (simplify) Factor analysis (find common factors in data, group variables) Cluster analysis, latent class (find common segments in data, group people) n Link variables to outcomes (identify cause) Experiment (link treatment to outcome, strong causality) Multiple regression (link many predictors to continuous outcome) Choice models (link predictors to an individual’s discrete choice) n Find tradeoffs among variables (optimize) Conjoint analysis (determine an individual’s unit-less tradeoff among unlike features; size, price, status) Response models (find level of inputs for optimum output) Customer lifetime value analysis (accounts for customer, time, and dynamic differences to make tradeoffs) © Palmatier 13

Data: A Good Sample is Key to Valid Results n Random sample n Law of large numbers n Removes alternative explanations, averages out noise Random-stratified sample (when everyone is not equal) Power and statistically significance Probability > 95% not due to chance Qualifying respondents Knowledgeable Population, potential customers, lost customers n Robustness and sensitivity (hold outs, retests) n Sources (survey, lists, customers, intercept, CRM database, secondary sources, etc. ) n Survey Multi-items, 7 point scale, avoid yes/no, center of scale (use adjectives to move mean) Simple, single idea questions (no and/or/if/buts) © Palmatier 14

Summary of 4 Stages of Marketing Research Stage I Planning and Design Internal Assessment & Planning • Objective(s) of segmentation • Resources • Constraints Database Review • Primary data already available • Secondary data • … Prototype Implementation Exercises Stage II Qualitative Assessment Qualitative Research • Interview Materials Development · · • Qualitative Data Collection · · • “Deep needs” Identification · · • Decision-Making Process Assessment · · Stage IV Analysis and Implementation Stage III Quantitative Measurement Quantitative Survey • Sample Design · · · Analysis • Cluster Analysis • Portfolio Analysis • Questionnaire Development · · · • Data Collection · · · • Conjoint Analysis Implementation • Call Center • Web Classification Tool Development • Discriminant function • Sales call patterns • Promotion • …. • Binary tree • … • What ifs? • Relevant groups involved? • …. . © Palmatier See Note on Market Research for process for startups 15

Agenda n Introduction n Research Methods and Data n Marketing Engineering Analysis n Markstrat Simulation (Part II) n Marketing Strategy Project © Palmatier 16

Marketing Engineering (ME) n Software that is an “add in” for Excel, which does many of the most powerful marketing analyses n Analyses and models n Factor and cluster analyses Discriminant and classification analyses Choice and conjoint models Bass diffusion models Response models, GE matrix, CLV You will have three assignments using ME Factor and cluster analyses, conjoint analysis, and choice models Tutorial under each menu item (software) Technical note on ME website (optional) Help session by TA (optional) © Palmatier 17

Agenda n Introduction n Research Methods and Data n Marketing Engineering Analysis n Markstrat Simulation (Part II) n Marketing Strategy Project © Palmatier 18

Success Depends on Using Strategic Processes and Research Tools n Segmentation, targeting, positioning (STP process) n Building SCAs Innovation of new products Modify existing products Build brands and relationships n Balancing portfolio of products and spending priorities n Market research data and tools Positioning maps Sales and advertising experiments Conjoint analysis Many different survey results (awareness > perception > intentions) You can now reposition using perceptual objectives and launch new products © Palmatier 19

STP is One Key Process for Success Strategy STP Process 20

Consumer Decisions Are Based on Brand Perceptions Your perception of brands Brand A Brand B Brand C Brand D Technical quality Their perception of brands ? Brand C Brand B Brand D ? Brand A 21

Perceptual Maps Are a Valuable tool for Brand Positioning Need 2 Segment ideal point Need 1 Perceived position of SOLO 22

Two Ways to Re-position Brands Perceived Performance Advertising R&D Perceived Economy 23

Positioning with R&D Involves Matching Products to Customer Needs CUSTOMER NEEDS Battery Life No. of Features Convenience Design Display Size Price SOLO Performance Processing Power Economy BRAND CHARACTERISTICS 24

Positioning With Advertising Involves Setting "Perceptual Objectives" Perceived Performance 20 Sh 11 10 5 Perceived Economy 0 -10 -20 -10 0 5 10 12 20 25

Innovation Is A Second Key Process for Success STP Process Strategy Innovation Process 26

R&D Success Requires Ability to Identify Evolving Opportunities Hi Sh New target segment Changing segment needs Pressure on margins Pr Competitor entry Sa 27

R&D Cost is a Function of Several Factors Physical characteristics R&D Experience Requested base cost To find out more: Budget required for completion Accuracy Feasibility study vs. Speed Online query 28

R&D Project Can Modify Existing Brands or Introduce New Brands Performance Ex Pr Sh Hi Economy New Brand MOKA Sa New MOVE higher performance Old MOVE 29

New Decisions in this Round R&D • Initiating new projects • Continuing or shelving uncompleted projects Brand Portfolio • Launching new brands • Upgrading or withdrawing existing brands Marketing Mix • Production planning, Pricing, Advertising & segmentation strategy • Setting Perceptual Objectives Commercial Team • Size of commercial team • Allocation across distribution channels and brands Market Research • Ordering industry-wide market studies • Ordering market-specific market studies 30

Initiating New Projects Click here to modify, delete, or shelve an existing project Launch a new brand targeted at Highs Click here to initiate a new project, and follow the assistant 31

Launching New Brands Click here to launch a new brand Choose a name … … enter the role of the new brand … … and choose the base R&D project 32

Modify Existing Brands You keep the same name Choose the new R&D project Select the Modify option 33

Set Perceptual Objectives for Each Brand 1 Do you want to communicate on the semantic scales or on the multidimensional scales? 2 On which dimensions do you want to communicate? 1 3 Input the coordinates of the point that you want to reach on the map 2 3 34

Each Team Can Get One Loan n Limit is 15, 000 M$ Interest is 5% Duration is 10 years (beyond end of class) Cannot be split up into multiple periods n Need to send me an email requesting the loan by the deadline of the decision round You will submit decision with an error of overspending your budget (I will give loan before I run simulation) n Key way to “catch-up” with new product(s) 35

Agenda n Introduction n Research Methods and Data n Marketing Engineering Analysis n Markstrat Simulation (Part II) n Marketing Strategy Project © Palmatier 36

Marketing Strategy Project n Each team will do a marketing strategy project for this class n Project feedback is recommended but optional n Last day of class hand in 5 page report, and Make a 20 minute presentation Email to me or meet at lunch/after class and I will give some feedback Business problem, decision, method(s), data, and analyses Must be a marketing strategy problem © Palmatier Use your firm or business idea Must do some quantitative analysis 37

Marketing Strategy Project Ideas n STP for an existing product or service n STP for new product n Plan to launch new or existing product or service n Test a marketing strategy using an experiment n Analyze “CRM or existing data” to find out what has worked in the past n Design a new or test an old loyalty program n How to best acquire, expand, or retain customers (AER strategies) n How to best use brands, offering, or relationships to achieve business goals (BOR strategies) n Evaluate marketing strategies based on CLV analysis © Palmatier 38

Exercise: Talk Among Your Team Members About Ideas for Your Project n Interesting and important problem n Ease of data collection n Pick one of your firms n “Clean” project n Ask me for project ideas (UW programs, local firms) n Will ask you about your idea in 5 to 10 minutes © Palmatier 39

Readings n Market research readings Market Research: Listen and Learn (overview of qualitative and quantitative research) Practical Regression: Convincing Empirical Research in Ten Steps (excellent discussion of models and analysis) Note on Marketing Research (good process on market research process for startups) n Data Analysis Techniques (DATs) in book across most all chapters © Palmatier 40