520 likes | 626 Views
PREDICTING AND EXPLAINING THE CHOICES OF BUYERS. (EVALUATION OF ALTERNATIVES). by Dr. Ted Mitchell. EVALUATION OF OUTCOMES. EVALUATION OF ALTERNATIVE PRODUCTS IN A USAGE SITUATION BY CONSUMERS EVALUATION OF MIDDLEMEN BY PRODUCERS
E N D
PREDICTING AND EXPLAININGTHE CHOICES OF BUYERS (EVALUATION OF ALTERNATIVES) by Dr. Ted Mitchell
EVALUATION OF OUTCOMES • EVALUATION OF ALTERNATIVE PRODUCTS IN A USAGE SITUATION BY CONSUMERS • EVALUATION OF MIDDLEMEN BY PRODUCERS • EVALUATION OF NEW PRODUCT IDEAS BY NEW PRODUCT SCREENING COMMITTEES • EVALUATION OF SALES PERFORMANCE BY SALES MANAGERS
TWO POINTS OF VIEW • NORMATIVE: EVALUATION MODELS REPRESENT THE PROCESS THAT A DECISION MAKER OUGHT TO EVALUATE ALTERNATIVE OUTCOMES, PRODUCTS, ETC. • POSITIVE: EVALUATION MODELS ARE GOOD DESCRIPTIONS OF HOW DECISION MAKERS OUGHT TO MAKE DECISIONS BETWEEN ALTERNATIVE OUTCOMES, PRODUCTS, ETC.
TWO BASIC TYPES OF EVALUATION MODELS • COMPENSATORY MODELS • NON-COMPENSATORY MODELS
MARKETING MANAGERS ARE MOST INTERESTED IN FINDING THE EVALUATION MODEL THAT ALLOWS THEM TO PREDICT WHICH PRODUCT A PARTICULAR TARGET MARKET WILL CHOOSE
STYLE 1 2 3 4 5 6 7 NOT IMPORTANT EXTREMELY IMPORTANT POWER 1 2 3 4 5 6 7 NOT IMPORTANT EXTREMELY IMPORTANT PRICE 1 2 3 4 5 6 7 NOT IMPORTANT EXTREMELY IMPORTANT COMFORT 1 2 3 4 5 6 7 NOT IMPORTANT EXTREMELY IMPORTANT WHEN YOU ARE THINKING ABOUT THE PURCHASE OF A TRUCK, HOW IMPORTANT ARE THE FOLLOWING WHEN YOU ARE EVALUATING DIFFERENT TRUCKS
STYLE 1 2 3 4 5 6 7 NOT IMPORTANT EXTREMELY IMPORTANT POWER 1 2 3 4 5 6 7 NOT IMPORTANT EXTREMELY IMPORTANT PRICE 1 2 3 4 5 6 7 NOT IMPORTANT EXTREMELY IMPORTANT COMFORT 1 2 3 4 5 6 7 NOT IMPORTANT EXTREMELY IMPORTANT WHEN YOU ARE THINKING ABOUT THE PURCHASE OF A TRUCK, HOW IMPORTANT ARE THE FOLLOWING WHEN YOU ARE EVALUATING DIFFERENT TRUCKS RESULTS FROM RESPONDENT 2135
WHAT PERCENTAGE OF THE TOTAL WEIGHT TO PUT ON EACH OF THE ATTRIBUTES? STYLE 6 4 3 7 POWER PRICE COMFORT TOTAL SCORE 20
WHAT PERCENTAGE OF THE TOTAL WEIGHT TO PUT ON EACH OF THE ATTRIBUTES? STYLE 6 4 3 7 POWER PRICE COMFORT TOTAL SCORE 20
STYLE 1 2 3 4 5 6 7 VERY POOR VERY GOOD POWER 1 2 3 4 5 6 7 VERY LOW VERY HIGH PRICE 1 2 3 4 5 6 7 VERY HIGH VERY LOW COMFORT 1 2 3 4 5 6 7 VERY POOR VERY GOOD WHEN YOU ARE EVALUATING BRAND A TRUCKSHOW MANY POINTS WOULD YOU GIVE BRAND A FOR EACH OF THE FOLLOWING COMPARED TO OTHER TRUCKS RESULTS FROM RESPONDENT 2135
STYLE 1 2 3 4 5 6 7 VERY POOR VERY GOOD POWER 1 2 3 4 5 6 7 VERY LOW VERY HIGH PRICE 1 2 3 4 5 6 7 VERY HIGH VERY LOW COMFORT 1 2 3 4 5 6 7 VERY POOR VERY GOOD WHEN YOU ARE EVALUATING BRAND B TRUCKSHOW MANY POINTS WOULD YOU GIVE BRAND BFOR EACH OF THE FOLLOWING COMPARED TO OTHER TRUCKS RESULTS FROM RESPONDENT 2135
STYLE 1 2 3 4 5 6 7 VERY POOR VERY GOOD POWER 1 2 3 4 5 6 7 VERY LOW VERY HIGH PRICE 1 2 3 4 5 6 7 VERY HIGH VERY LOW COMFORT 1 2 3 4 5 6 7 VERY POOR VERY GOOD WHEN YOU ARE EVALUATING BRAND C TRUCKSHOW MANY POINTS WOULD YOU GIVE BRAND CFOR EACH OF THE FOLLOWING COMPARED TO OTHER TRUCKS RESULTS FROM RESPONDENT 2135
STYLE 1 2 3 4 5 6 7 VERY POOR VERY GOOD POWER 1 2 3 4 5 6 7 VERY LOW VERY HIGH PRICE 1 2 3 4 5 6 7 VERY HIGH VERY LOW COMFORT 1 2 3 4 5 6 7 VERY POOR VERY GOOD WHEN YOU ARE EVALUATING BRAND D TRUCKSHOW MANY POINTS WOULD YOU GIVE BRAND DFOR EACH OF THE FOLLOWING COMPARED TO OTHER TRUCKS RESULTS FROM RESPONDENT 2135
BASIC ASSUMPTIONS • PEOPLE SEE PRODUCTS AS BEING MADE UP OF VARIOUS ATTRIBUTES • PEOPLE PUT DIFFERENT AMOUNTS OF IMPORTANCE WEIGHT ON DIFFERENT ATTRIBUTES • PEOPLE HAVE BELIEFS ABOUT THE AMOUNT OF EACH ATTRIBUTE A PRODUCT CONTAINS
THE BEST PRODUCT IS THE ONE YOU LIKE BEST FOR YOUR ANTICIPATED USAGE SITUATIONTHE RATIONAL CHOICE IS TO CHOOSE THE PRODUCT YOU LIKE THE BEST
Different People Evaluate Products in Different Ways The evaluation they give to products and the manner in which they evaluate products Is a primary dimension for segmenting markets and choosing target markets.
RESULTS FROM RESPONDENT 2135 BRAND A BRAND B BRAND C BRAND D STYLE WEIGHT ? ? ? ? ? POWER WEIGHT ? ? ? ? ? PRICE WEIGHT ? ? ? ? ? COMFORT WEIGHT ? ? ? ? ?
RESULTS FROM RESPONDENT 2135 BRAND A BRAND B BRAND C BRAND D STYLE 30% ? ? ? ? POWER 20% ? ? ? ? PRICE 15% ? ? ? ? COMFORT 35% ? ? ? ?
Approach to Prediction • Compensatory Models Seek an overall score for prediction of choice • Non-compensatory Models assume a decision in the process of evaluation
All Models Start With The Premise That if a product is best on all attributes by a customer then it will be purchased by that customer That is to say: The Dominance Model Of Choice is Universal
Brand X is the dominant choice because it is best on all attributes. BRAND A BRAND B BRAND C BRAND X STYLE 30% 1 2 5 6 POWER 20% 6 3 4 7 PRICE 15% 3 4 2 5 COMFORT 35% 5 6 4 7
When there is no dominant brand, then Compensatory explanations
Compensatory Looks For BRAND A BRAND B BRAND C BRAND D STYLE 30% 1 2 5 6 POWER 20% 6 3 4 5 PRICE 15% 3 4 2 5 COMFORT 35% 5 6 4 2 OVERALL SCORE ? ? ? ?
Non-compensatory: Lexicographic Identify Most Important attribute, then BRAND A BRAND B BRAND C BRAND D STYLE 30% 1 2 5 6 POWER 20% 6 3 4 5 PRICE 15% 3 4 2 5 COMFORT 35% 5 6 4 2
Non-compensatory: Lexicographic Identify Most Important attribute, then BRAND A BRAND B BRAND C BRAND D STYLE 30% 1 2 5 6 POWER 20% 6 3 4 5 PRICE 15% 3 4 2 5 COMFORT 35% 5 6 4 2 Choose the brand with the best score on that attribute.
Types of Non-compensatory Models • Conjunctive • Must meet a minimum level on all attributes or drop • Disjunctive • Be greater than a specific level on a few of the important attributes • Lexicographic • Highest on most important attribute, go to second most important attribute only to break a tie.
Compensatory: Expectancy Value Model BRAND A BRAND B BRAND C BRAND D STYLE 30% 1 2 5 6 POWER 20% 6 3 4 5 PRICE 15% 3 4 2 5 COMFORT 35% 5 6 4 2 OVERALL SCORE ? ? ? ?
RESULTS FROM RESPONDENT 2135 BRAND A BRAND B BRAND C BRAND D STYLE 30% .3 (1) =.3 .3(2) = 0.6 .3(5) = 1.5 .3(6) = 1.8 POWER 20% .2(6) =1.2 .2(3) = 0.6 .2(4) = 0.8 .2(5) = 1.0 PRICE 15% .15(3) = .45 .15(4) = 0.6 .15(2) = 0.3 .15(5) =0.75 COMFORT 35% .35(5) = 1.75 .35(6) = 2..1 .35(4) = 1.4 .35(2) = 0.7 OVERALL SCORE ? ? ? ?
RESULTS FROM RESPONDENT 2135 BRAND A BRAND B BRAND C BRAND D STYLE 30% 0.3 0.6 1.5 1.8 POWER 20% 1.2 0.6 0.8 1.0 PRICE 15% .45 0.6 0.3 0.75 COMFORT 35% 1.75 2..1 1.4 0.7 OVERALL SCORE ? ? ? ?
RESULTS FROM RESPONDENT 2135 BRAND A BRAND B BRAND C BRAND D STYLE 30% 0.3 0.6 1.5 1.8 POWER 20% 1.2 0.6 0.8 1.0 PRICE 15% .45 0.6 0.3 0.75 COMFORT 35% 1.75 2..1 1.4 0.7 OVERALL SCORE 0.3 + 1.2 +.45 + 1.75 = 3.75 ? ? ?
RESULTS FROM RESPONDENT 2135 BRAND A BRAND B BRAND C BRAND D STYLE 30% 0.3 0.6 1.5 1.8 POWER 20% 1.2 0.6 0.8 1.0 PRICE 15% .45 0.6 0.3 0.75 COMFORT 35% 1.75 2..1 1.4 0.7 OVERALL SCORE 3.75 3.9 4.0 4.25
BRAND D STYLE 30% 1.8 RESULTS FROM RESPONDENT 2135 IN A SIMPLE EXPECTANCY VALUE MODEL PREDICT HE WILL PURCHASE BRAND D BRAND A BRAND B 0.3 0.6 POWER 20% 1.0 1.2 0.6 PRICE 15% 0.75 .45 0.6 COMFORT 35% 0.7 1.75 2..1 OVERALL SCORE 4.25
RESULTS FROM RESPONDENT 2135 BRAND A BRAND B BRAND C BRAND D STYLE 30% 1 2 5 6 POWER 20% 6 3 4 5 PRICE 15% 3 4 2 5 COMFORT 35% 5 6 4 2 OVERALL SCORE 3.75 3.9 4.0 4.25
Why Is this Basic Model so popular? • forces us to be explicitvertising/product decisions • It allows us to think about probabilities of purchase and using them as a measure of intention strength • Use the probabilities)
Basic Weakness of The Compensatory Model • Weights are constant, beliefs are stable • Attributes are constant across products and across useage situations. • Attributes are not necessarily independent • safety & power, quality and price, • “Bigger is always better” • Evaluation and Articulation are not the same. • New products can cause new dimensions
Not Weaknesses of ModelWeaknesses • There is nothing statistical about the model per se. (no statistical weakness.... sample size etc. • Failure to include an attribute is not a weakness of the model (zero weight) • “You don’t know if the model has the right attributes.” • Using wrong model is not a weakness of the model (wrongly used lexicographic) • Subjectivity is not a weakness! Lack of consistency is a potential problem.
Implications for Changing Strategy • Modify the product (real positioning) • Alter beliefs about amount of attributes • Alter beliefs about competitor’s amount • Alter importance weights • Call attention to neglected attributes • Shift the buyers’ ideals
Assume your are the manager of Brand B BRAND A BRAND B BRAND C BRAND D STYLE 30% 1 2 5 6 POWER 20% 6 3 4 5 PRICE 15% 3 4 2 5 COMFORT 35% 5 6 4 2 OVERALL SCORE 3.75 3.9 4.0 4.25
Use promotion to change beliefs about amount of attribute in your brand. BRAND A BRAND B BRAND C BRAND D STYLE 30% 1 2 5 6 POWER 20% 6 3 4 5 PRICE 15% 3 4 2 5 COMFORT 35% 5 6 4 2 OVERALL SCORE 3.75 3.9 4.0 4.25
Use promotion to change beliefs about amount of attribute in your Competitor’s brand. BRAND A BRAND B BRAND C BRAND D STYLE 30% 1 2 5 6 POWER 20% 6 3 4 5 PRICE 15% 3 4 2 5 COMFORT 35% 5 6 4 2 OVERALL SCORE 3.75 3.9 4.0 4.25
Change the importance Weights BRAND A BRAND B BRAND C BRAND D STYLE 30% 1 2 5 6 POWER 20% 6 3 4 5 PRICE 15% 3 4 2 5 COMFORT 35% 5 6 4 2 OVERALL SCORE 3.75 3.9 4.0 4.25
Talk About Neglected Attributes BRAND A BRAND B BRAND C BRAND D SMOOTH RIDE 1 9 1 1 STYLE 30% 1 2 5 6 POWER 20% 6 3 4 5 PRICE 15% 3 4 2 5 COMFORT 35% 5 6 4 2 OVERALL SCORE 3.75 3.9 4.0 4.25
Probabilities of Purchase • Highest value takes all the market? • A value reflects a high probability of Purchase is more realistic • A probability of purchase on an individual basis implies a market share percentage over the total market
RESULTS FROM RESPONDENT 2135 BRAND A BRAND B BRAND C BRAND D STYLE 30% 1 2 5 6 POWER 20% 6 3 4 5 PRICE 15% 3 4 2 5 COMFORT 35% 5 6 4 2 OVERALL SCORE 3.75 3.9 4.0 4.25
RESULTS FROM RESPONDENT 2135 BRAND A BRAND B BRAND C BRAND D OVERALL SCORE 3.75 3.9 4.0 4.25
RESULTS FROM RESPONDENT 2135 BRAND A BRAND B BRAND C BRAND D OVERALL SCORE 3.75 3.9 4.0 4.25 Probability of A
Repeat for each brand BRAND A BRAND B BRAND C BRAND D OVERALL SCORE 3.75 3.9 4.0 4.25 Probability of Purchase 23.6% 24.5% 25.2% 26.7% Probability of Purchase implies Market share
Implications for Changing Strategy • Modify the product (real positioning) • alter beliefs about amount of attributes • alter beliefs about competitor’s amount • alter importance weights • call attention to neglected attributes • What degree of shift in beliefs about power is necessary to give us a 1% increase in market share? • What will it cost to gain this increase in beliefs? • Will the increase in Market share pay for the effort?
Disadvantages Of Simple Expectancy Value Model • All the attributes are considered independent with no interaction • safety & speed, price & quality, size & comfort • All beliefs are stable and consistent • “Bigger is always better”
The Concept of An Ideal Brandis at the basis of many product positioning ideas. Comfort B C A Ideal Style D