GoConqr has a new look! Let us know what you think.

Sign up free

Flashcards

2.3 Parameter estimation

Martin Reichmuth Mops German 29 cards

Aggregation Entity (3x3)

Cards in this deck (29)

  1. 1ObjectiveA) Statistical analysis
    -> Aggregation
    -> Comparability
    B) Experiments
    -> Characteristics
    -> Types/Assessment
    -> Evaluation
    C) Analogies & meta analyses
    -> Comparability
    -> Results
  2. 2Objective Aggregation (4)- Time
    -Space
    - Entity
    - Marketing Elements
  3. 3Aggregation Time (5)• Yearly data
    • Quarterly data (Q)
    • Monthly data (M)
    • Weekly data
    • Time-accurate data (D)
  4. 4Aggregation Space (3)• States (St)
    • Districts (Di)
    • Sales territories (Te)
  5. 5Aggregation Entity (3x3)Aggregation Entity (3x3)
  6. 6Aggregation Marketing Elements (3)Aggregation Marketing Elements (3)
  7. 7Aggregation Suggestionmarket response should be estimated using the most disaggregated data available
  8. 81. Temporal comparability (3)• Adjustment for trend and Season
    • Adjustment for inflation
    • Adjustment for demographic developments over time
  9. 92. Spatial comparability (2)• Correction for transactions outside of a sales territory
    • Inclusion of all factors which explain structural differences of sales territories
    Measures
  10. 103. Comparability ofentities (2)• City versus ZIP code
    • Brand versus individual stock keeping units (SKUs)
  11. 114. Comparability of marketing Elements (2)• Use different voucher codes for different marketing elements (e.g., on print coupons, in TV commercials, on the internet)
    • Asking customer how she/he was reached by marketing
  12. 125. Comparability: Other• Jackknife method:
    Calculation of parameters for a lot of subsamples and creation of a stable parameter value as a mean value of the subsample parameter values (i.e., coefficients).
  13. 13Characteristics of Experiments (2)Characteristics of Experiments (2)
  14. 14Types of experiments used to estimate market response functions (3)1. Laboratory experiments (artificial environment)
    2. Simulated field experiments
    3. Field experiments
  15. 15Types of Experiments 1. Laboratory experiments (3)a. Money equivalents for products
    b. Estimation of preferences by rank-ordering of product stimuli (conjoint analysis)
    c. Recall of commercials in manipulated reading material
  16. 16Types of Experiments 2. Simulated field Experiments (2)a. Test market simulations
    b. Mini test market
  17. 17Types of Experiments 3. Field Experiments (4)Types of Experiments 3. Field Experiments (4)
  18. 18Evaluation of types of Experiments (3)Evaluation of types of Experiments (3)
  19. 19Evaluation of experiments by managers (5)1. field experiments are expensive & time-consuming.
    2. Results are also informative for competitors.
    3. Field experiments need to take place during day-to-day Business, thus, not all external influences can be controlled for.
    4. Consequently experimentally reduced sales activities, decreases in sales & customer relations which have been established over a long time can suffer.
    5. After ending experimentally increased sales activities, customers may react negatively when the level of sales activities return to the former level.
  20. 20Evaluation of experiments by managers Nevertheless:Nevertheless:
    Even though extensive field experiments might not be necessarily advantageous per se, it is still recommended to continuously run small experiments which allow management to permanently stay updated and gain new customer insights.
  21. 21AnalogiesAnalogies
  22. 22Meta-analysesMeta-analyses
  23. 23Subjective response estimation- In many cases only one point of the market response function is known
    - managers make decisions based on intuitive knowledge
    - market response function not estimated directly
    - parameters derived from point estimates (ADBUDG )Estimation of the ADBUDG function .
  24. 24Estimation of the ADBUDG function Step 1Estimation of the ADBUDG function Step 1
  25. 25Estimation of the ADBUDG function Step 2Estimation of the ADBUDG function Step 2
  26. 26Evaluation 1. Statistical estimation (non-experimental)Evaluation 1. Statistical estimation (non-experimental)
  27. 27Evaluation 2. Experimental estimationAdvantages:
    • Market response is directly attributable to the variation of marketing Instruments
    • Error can be quantified
    Disadvantages:
    • A lot of requirements regarding the experimental design
    • Expensive and time-consuming
  28. 28Evaluation 3. Analogies and meta-analysesAdvantages:
    • Applicable even to new products
    Disadvantages:
    • Low reliability
  29. 29Evaluation 4. Subjective estimationAdvantages:
    • In terms of implementation the easiest
    • Intuitive knowledge is disclosed
    • Estimation process triggers learning
    Disadvantages:
    - errors cannot be quantified

Information

Description
Mops Flashcards on 2.3 Parameter estimation, created by Martin Reichmuth on 18/07/2018.
Views
1
Rating

Feedback

What is it about?

What do you think of the new look?

How do you like it?