Method of measuring and evaluating consumer response for the development of consumer products

• US 5,041,972 A
• Filed: 04/15/1988
• Issued: 08/20/1991
• Est. Priority Date: 04/15/1988
• Status: Expired due to Fees

• Current Assignee: Maritz Incorporated
• Original Assignee: Frost W. Alan
• Inventors: Frost, W. Alan
• Primary Examiner(s): Smith, Jerry
• Assistant Examiner(s): Hayes, Gail O.
• Application Number:
  US07/181,784
• Time in Patent Office: 1,222 Days
• Field of Search: 364/401, 358/84, 455/2
• US Class Current: 705/7.32
• CPC Class Codes: G06F 17/18 for evaluating statistical data , e.g. average values, frequency distributions, probability functions, regression analysis forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0203 Market surveys; Market polls

Abstract

A method for evaluating consumer response comprising conducting interviews to obtain emotional, rational and personality descriptors of functionally related items, eliciting from consumers evaluations of the extent to which the non-eliminated descriptors are attributable to each item, creating a discrimination index of the evaluated descriptors whereby descriptors, which provide the most discrimination between items and which account for the greatest amount of behavioral variance over 70% among consumers, are identified as attributes, eliciting from consumers evaluations of the extent to which attributes are attributable to reference items and to given items, eliciting from consumers evaluations of a degree of preference for each given item, performing an independence factor analysis of the attributes in order to form clusters of related attributes, creating a multi-dimensional matrix for each use associated with given items wherein points representing reference items and given items are plotted based upon the attribute evaluations associated with each item, and determining the effect of a given change in an attribute evaluation for all consumers for a given item by measuring the relationships between the Euclidean distances between the points representing the items and reference items on the matrix and the degrees of preferences for the given items.

Claims

1. A method for evaluating consumer response comprising:

a. conducting interviews of consumers whereby rational, stereotype and personality descriptors of related items are elicited and descriptors, which the interviews indicate are least effective as bases for the consumers to distinguish between the items, are eliminated;

b. eliciting from consumers evaluations of the extent to which the non-eliminated descriptors are attributable to items;

c. calculating a discrimination index of the evaluated descriptors whereby the least number of descriptors, which provide the most discrimination between items and which systematically account for the greatest amount of behavioral variance over 70% among the interviewed consumers, are identified as attributes;

d. eliciting from consumers evaluation of the extent to which attributes are attributable to reference items and to given items for uses associated with the items;

e. eliciting from consumers evaluations of a degree of preference for given items for uses associated with the given items;

f. performing an independence factor analysis of the attributes in order to form clusters of related attributes;

g. creating a multi-dimensional matrix of factors for uses associated with the given items wherein points representing the reference items and the given items are plotted based upon the attribute evaluations associated with the respective item;

h. determining the effect of a given change in an attribute evaluation for a plurality of consumers for a given item by measuring the relationships among the Euclidean distances between the points representing the items and reference items on the matrix and the degrees of preferences for the given items.

2. A method according to claim 1 which further comprises computing for each consumer a purchase elasticity curve determined by points defined by (i) the relative proportion which the likelihood of purchasing each item identified as a potential purchase bears to the sum of all of such likelihoods for all such items and (ii) the Euclidean distances between the points on the squeezed matrix representing the items identified as potential purchases and the point representing the ideal item.

3. A method according to claim 2 which further comprises determining for each consumer the propensity of a given change in an attribute evaluation to effect a change in the likelihood of purchasing each item relative to the concomitant change in the likelihood of purchasing each other item whereby a propensity weight is assigned to the attribute evaluations for each consumer which weight corresponds to the slope of the portion of the purchase elasticity curve computed for said consumer falling between the points which represent the items for which the propensity to change is being determined.

4. A method according to claim 1 wherein a squeeze analysis of the attributes is performed whereby (i) a matrix is created wherein points representing the Euclidean distances between each item and the ideal item are plotted based upon the attribute evaluations associated with each item and (ii) an importance weight is assigned to each attribute so that the Euclidean distances between the points on the matrix representing each item and the point representing the ideal item are re-ranked into the same order as the likelihoods of purchasing each item.

5. A method according to claim 1 wherein the effect of a given change in an attribute evaluation for all consumers for a given item is further determined by measuring the correlation between attributes and calculating concomitant adjustments in other attribute evaluations in accordance with such correlations.

6. A method according to claim 1 which further comprises creating a perceptual map wherein points representing attributes are plotted with respect to the ability of the respective attributes to provide a basis for consumers to distinguish one reference item from other reference items and points representing reference items are plotted with respect to the attribute evaluations associated with each respective reference item, eliciting from consumers evaluations of the degree of preference for a reference item as associated with a given use, plotting points on the perceptual map representing the given items which points are plotted with respect to the attribute evaluations associated with each given item respectively.

7. A method according to claim 6 which further comprises measuring the change in the relative preference for each given item which results from a given change for all consumers in an attribute evaluation for a given item which change is reflected by the resultant relocation of the point representing the given item on the perceptual map and is measured by the Euclidean distances of that point from the points representing the reference fragrances on the perceptual map.

8. A method according to claim 1 wherein the set of attributes chosen achieve a level of behaviorial variance of about 90% or greater.

9. A method for evaluating consumer response comprising:

a. eliciting descriptors of related items from consumers including rational, stereotype and personality descriptors;

b. submitting the descriptors to consumers in qualitative interviews in order to reduce the number thereof by eliminating descriptors which the interviews indicate are insufficient as bases for the consumers to distinguish between the items;

c. eliciting from consumers evaluations of the extent to which the non-eliminated descriptors are attributable to items;

d. calculating a discrimination index of the evaluated descriptors whereby descriptors, which provide the most discrimination between items and which systematically account for the greatest amount of behavioral variance over 70% among the interviewed consumers, are identified as attributes;

e. eliciting from consumers identification of uses which consumers associate with items;

f. eliciting from consumers evaluations of the extent to which attributes are attributable to items for uses associated with the items;

g. eliciting from consumers evaluations of the extent to which attributes ideally should be possessed by an item for uses associated with the items;

h. eliciting from consumers evaluations of the likelihood of purchasing items for uses associated with the items;

i. eliciting from consumers an identification of which items each consumer would potentially purchase for uses associated with the items;

j. performing an independence factor analysis of the attributes in order to form clusters of related attributes;

k. performing with respect to uses identified by consumers in step e a squeeze analysis of the attributes whereby (i) a matrix of factors is created wherein points representing the Euclidean distances between each item and the ideal item are plotted based upon the attribute evaluations associated with each respective item and (ii) an importance weight is assigned to each attribute so that the Euclidean distances between the points on the matrix representing each item and the point representing the ideal item are re-ranked into the same order as the likelihoods of purchasing each item;

l. computing for each consumer a purchase elasticity curve determined by points defined by (i) the relative proportion which the likelihood of purchasing each item identified as a potential purchase bears to the sum of all of such likelihoods for all such items and (ii) the Euclidean distances between the points on the squeezed matrix representing the items identified as potential purchases and the point representing the ideal item;

m. determining for each consumer the propensity of a given change in an attribute evaluation to effect a change in the likelihood of purchasing each item relative to the concomitant change in the likelihood of purchasing each other item whereby a propensity weight is assigned to the attribute evaluations for each consumer which weight corresponds to the slope of the portion of the purchase elasticity curve computed for said consumer falling between the points which represent the items for which the propensity to change is being determined;

n. calculating a purchase probability for each item by summing the relative proportions computed for each item for all consumers in step l(i);

o. measuring the change in the purchase probability for each item which results from a given change in an attribute evaluation for all consumers for a given item by (i) calculating for each consumer the concomitant adjustments in other attribute evaluations in proportion to the degree to which said other attribute evaluations respectively correlate to the changed attribute evaluation, (ii) re-valuing these changes and adjustments in proportion to the importance weights and propensity weights, assigned to each attribute evaluation, respectively, and (iii) summing for all consumers the re-valued changes and adjustments associated with each item.

10. A method for evaluating consumer response comprising:

a. eliciting descriptors of functionally-related items from consumers including rational, stereotype and personality descriptors;

b. submitting the descriptors to consumers in qualitative interviews in order to reduce the number thereof by eliminating descriptors which the interviews indicate are insufficient as bases for the consumers to distinguish between the items;

c. eliciting from consumers evaluations of the extent to which the non-eliminated descriptors are attributable to each item;

d. calculating a discrimination index of the evaluated descriptors whereby descriptors, which provide the most discrimination between items and which account for a level of behavioral variance of about 90% or greater among the interviewed consumers, are identified as attributes;

e. eliciting from consumers identification of the uses which consumers associate with items;

f. eliciting from consumers evaluations of the extent to which each attribute is attributable to each item for each use identified by consumers in step e;

g. eliciting from consumers evaluations of the extent to which each attribute ideally should be possessed by an item when used for a use identified by consumers in step e;

h. eliciting from consumers evaluations of the likelihood of purchasing each of the items for each use identified by consumers in step e;

i. eliciting from consumers an identification of which items each consumer would potentially purchase for each use identified in step e;

j. performing an independence factor analysis of the attributes in order to form clusters of related attributes;

k. performing with respect to each use identified by consumers in step e a squeeze analysis of the attributes whereby (i) a matrix of factors is created wherein points representing the Euclidean distances between each item and the ideal item are plotted based upon the attribute evaluations associated with each respective item and (ii) an importance weight is assigned to each attribute so that the Euclidean distances between the points on the matrix representing each item and the point representing the ideal item are re-ranked into the same order as the likelihoods of purchasing each item;

l. computing for each consumer a purchase elasticity curve determined by points defined by (i) the relative proportion which the likelihood of purchasing each item identified as a potential purchase bears to the sum of all of such likelihoods for all such items and (ii) the Euclidean distances between the points on the squeezed matrix representing the items identified as potential purchases and the point representing the ideal item;

m. determining for each consumer the propensity of a given change in an attribute evaluation to effect a change in the likelihood of purchasing each item relative to the concomitant change in the likelihood of purchasing each other item whereby a propensity weight is assigned to the attribute evaluations for each consumer which weight corresponds to the slope of the portion of the purchase elasticity curve computed for said consumer falling between the points which represent the items for which the propensity to change is being determined;

n. calculating a purchase probability for each item by summing the relative proportions computed for each item for all consumers in step l(i);

o. measuring the change in the purchase probability for each item which results from a given percent change for each consumer in an importance weight for a given attribute by (i) calculating concomitant adjustments in the importance weights associated with other attributes in proportion to the degree to which the attributes associated with the adjusted importance weights respectively correlate with the attribute associated with the changed importance weight, (ii) revaluing for each consumer the Euclidean distances calculated in step k in proportion to the respective importance weights using the adjustments calculated in step o(i), (iii) calculating new preference evaluations on the basis of the respective revalued Euclidean distances and (iv) summing the new preference evaluations for each item.

11. A method for measuring and evaluating consumer response for the development of consumer products comprising:

a. eliciting from consumers in qualitative open-ended interviews descriptors of related items including rational, stereotype and personality descriptors;

b. including the descriptors in questionnaires and submitting said questionnaires to consumers in qualitative interviews whereby the number of descriptors is reduced by eliminating from consideration descriptors which the responses to such interviews indicate are insufficient as bases for the consumers to distinguish between the items;

c. eliciting from consumers in quantitative interviews evaluation of the extent to which the non-eliminated descriptors are attributable to each item;

d. performing a discriminant analysis of the evaluated descriptors whereby a discrimination index of the evaluated descriptors is calculated and whereby descriptors, which provide the most discrimination between items and systematically account for the greatest amount of behavioral variance over 70% among the interviewed consumers, are identified as attributes;

e. eliciting from consumers using questionnaires in qualitative interviews identification of uses which consumers associate with items;

f. eliciting from consumers using questionnaires in quantitative interviews evaluations of the extent to which each attribute is attributable to each item for uses associated with the items;

g. eliciting from consumers using questionnaires in quantitative interviews attribute evaluations of the extent to which each attribute ideally should be possessed by an item for uses associated with the items;

h. eliciting from consumers evaluation of purchase interest, for each item when used for uses associated with the items, using questionnaires in quantitative interviews;

i. eliciting from consumers using questionnaires in qualitative interviews identification of potential purchases for uses associated with the items;

j. performing an independence factor analysis of the attributes whereby clusters of related attributes are formed;

k. performing with respect to each use associated with the itemsy a squeeze analysis, iteratively according to St. James Theorum, of the attributes whereby (i) a matrix is created wherein points representing each item and the ideal item are plotted based upon the attribute evaluations associated with each respective item and (ii) an importance weight is assigned to each attribute by adjusting the scale used to measure each attribute evaluation in the matrix which adjustment relocates the position of each point on the matrix to obtain the closest fit of the Euclidean distances between the points on the matrix representing each item and the point representing the ideal item to the purchase interest for the respective item;

l. computing for the evaluations elicited from each consumer a purchase elasticity curve defined by points on a graph which are plotted with respect to the relative proportions which the purchase interest for each potential purchase bears to the sum of purchase interests for all potential purchases on the vertical axis and the Euclidean distances between the points on the squeezed matrix representing each potential purchase and the point representing the ideal item on the horizontal axis;

m. calculating a purchase probability for each item by summing among consumers the proportions of purchase interest associated with each item as a potential purchase;

n. determining the propensity of a given change in an attribute evaluation to effect a change in the purchase interest associated with each potential purchase relative to the concomitant change in the purchase interest associated, with each other potential purchase, respectively, whereby a propensity weight is assigned to the attribute evaluation for each consumer which corresponds to slope of the portion of the purchase elasticity curve computed with respect to said consumer falling between the two points on said curve which represent the potential purchase with respect to which the propensity to change is being determined.

o. measuring the change in the purchase probability for each item which results from a given percent change for all consumers in an attribute evaluation for a given item whereby (i) concomitant percent adjustments in other attribute evaluations are for each consumer calculated in proportion to the degree to which said attribute evaluations respectively correlate to the changed attribute evaluation, (ii) said changes and adjustments are modified in accordance with the importance weights and propensity weights assigned to each attribute evaluation respectively and (iii) the modified changes and adjustments associated with each item are summed for all consumers.

12. A method for measuring and evaluating consumer response for the development of consumer products comprising:

a. eliciting from consumers in qualitative open-ended interviews descriptors of related items including rational, stereotype and personality descriptors;

b. including the descriptors in questionnaires and submitting said questionnaires to consumers in qualitative interviews whereby the number of descriptors is reduced by eliminating from consideration descriptors which the responses to such interviews indicate are insufficient as bases for the consumers to distinguish between the items;

c. eliciting from consumers in quantitative interviews evaluation of the extent to which the non-eliminated descriptors are attributable to each item;

d. performing a discriminant analysis of the evaluated descriptors whereby a discrimination index of the evaluated descriptors is calculated and whereby descriptors, which provide the most discrimination between items and which account for a level of behavioral variance of about 90% or greater among the interviewed consumers, are identified as attributes;

e. eliciting from consumers using questionnaires in qualitative interviews identification of uses which consumers associate with items;

f. eliciting from consumers using questionnaires in quantitative interviews evaluations of the extent to which each attribute is attributable to each item when used for a use identified by consumers in step e;

g. eliciting from consumers using questionnaires in quantitative interviews attribute evaluations of the extent to which each attribute ideally should be possessed by an item when used for a use identified by consumers in step e;

h. eliciting from consumers evaluation of purchase interest, for each item when used for a use identified in step e, using questionnaires in quantitative interviews;

i. eliciting from consumers using questionnaires in qualitative interviews identification of potential purchases for each use identified in step e;

j. performing an independence factor analysis of the attributes whereby clusters of related attributes are formed;

k. performing with respect to each use identified by consumers in step e a squeeze analysis, iteratively according to St. James Theorum, of the attributes whereby (i) a matrix of factors is created wherein points representing each item and the ideal item are plotted based upon the attribute evaluations associated with each respective item and (ii) an importance weight is assigned to each attribute by adjusting the scale used to measure each attribute evaluation in the matrix which adjustment relocates the position of each point on the matrix to obtain the closest fit of the Euclidean distances between the points on the matrix representing each item and the point representing the ideal item to the purchase interest for the respective item;

l. computing for the evaluations elicited from each consumer a purchase elasticity curve defined by points on a graph which are plotted with respect to the relative proportions which the purchase interest for each potential purchase bears to the sum of purchase interests for all potential purchases on the vertical axis and the Euclidean distances between the points on the squeezed matrix representing each potential purchase and the point representing the ideal item on the horizontal axis;

m. calculating a purchase probability for each item by summing among consumers the proportions of purchase interest associated with each item as a potential purchase;

n. determining the propensity of a given change in an attribute evaluation to effect a change in the purchase interest associated with each potential purchase relative to the concomitant change in the purchase interest associated with each other potential purchase, respectively, whereby a propensity weight is assigned to the attribute evaluation for each consumer which corresponds to slope of the portion of the purchase elasticity curve computed with respect to said consumer falling between the two points on said curve which represent the potential purchase with respect to which the propensity to change is being determined;

o. measuring the change in the purchase probability for each item which results from a given percent change for each consumer in an importance weight for a given attribute by (i) calculating concomitant adjustments in the importance weights associated with other attributes in proportion to the degree to which the attributes associated with the adjusted importance weights respectively correlate with the attribute associated with the changed importance weight, (ii) revaluing for each consumer the Euclidean distances calculated in step k in proportion to the respective importance weights using the adjustments calculated in step o(i), (iii) calculating new preference evaluations on the basis of the respective revalued Euclidean distances and (iv) summing the new preference evaluations for each item.

13. A method for the development of consumer products comprising:

a. creating or obtaining a set of fragrance notes sufficient for a perfumer to identify and distinguish between fragrances on the basis of preparation and composition thereof;

b. identifying a set of reference fragrances corresponding to the fragrance notes;

c. conducting interviews of consumers whereby rational, stereotype and personality descriptors of the reference fragrances are elicited and descriptors which the interviews indicate are insufficient as bases to distinguish between reference fragrances are eliminated;

d. eliciting from consumers evaluations of the extent to which the non-eliminated descriptors are attributable to reference fragrances;

e. calculating a discrimination index of the evaluated descriptors whereby descriptors, which provide the most discrimination between reference fragrances and which systematically account for the greatest amount of behavioral variance over 70% among the interviewed consumers, are identified as attributes;

f. eliciting from consumers evaluations of the extent to which attributes are attributable to reference fragrances;

g. creating a perceptual map wherein points representing attributes are plotted with respect to the ability of the respective attributes to provide a basis for consumers to distinguish one reference fragrance from other reference fragrances and points representing reference fragrances are plotted with respect to the attribute evaluations associated with each respective reference fragrance;

h. eliciting from consumers evaluation of the degree of preference for a reference fragrance as associated with a use;

i. eliciting from consumers evaluations of the extent to which attributes are attributable to given fragrances;

j. performing an independence factor analysis of the attributes in order to form clusters of related attributes;

k. plotting points on the perceptual map created in step h representing the given fragrances which points are plotted with respect to the attribute evaluations associated with each given fragrance respectively;

l. eliciting from consumers evaluations of the degree of preference for a given fragrance as associated with a given use;

m. calculating a relative preference for each given fragrance as associated with a given use which corresponds to the proportion which the preference evaluation for each given fragrance bears to the sum of preference evaluations for all given fragrances;

n. measuring the change in the relative preference for each given fragrance which results from a given change for all consumers in an attribute evaluation for a given fragrance which resulting change is reflected by the resultant relocation of the point representing the given fragrance on the perceptual map and is measured by the Euclidean distances of that point from the points representing the reference fragrances on the perceptual map.

14. A method according to claim 13 wherein the attributes achieve a level of behavioral variance of about 90% or greater.