Presentation is loading. Please wait.

Presentation is loading. Please wait.

1 OPSM 405 Service Management Class 9: Service System Design Tools: Service Blueprinting Conjoint Analysis Koç University Zeynep Aksin

Similar presentations


Presentation on theme: "1 OPSM 405 Service Management Class 9: Service System Design Tools: Service Blueprinting Conjoint Analysis Koç University Zeynep Aksin"— Presentation transcript:

1 1 OPSM 405 Service Management Class 9: Service System Design Tools: Service Blueprinting Conjoint Analysis Koç University Zeynep Aksin

2 2 Service blueprinting  Activities  Decision points  Precedence relations  Line of visibility  Resources

3 Low complexity, high divergence

4 High complexity, low divergence

5 High complexity, high divergence

6 Defining terminology complexity vs. divergence what is done? how is it done?

7 Understanding the link between positioning and service structure Structural change: reduce divergence  positioning: economies of scale  + : perceived increase in reliability  - : conformity, inflexibility

8 Understanding the link between positioning and service structure Structural change: increase divergence  positioning: niche  + : prestige, customization, personalization  - : difficult to manage and control

9 Understanding the link between positioning and service structure Structural change: reduce complexity  positioning: specialization  + : expert image, easy control  - : stripped down image

10 Understanding the link between positioning and service structure Structural change: increase complexity  positioning: wallet share  + : maximize revenue generation / customer  - : customer confusion, decline in service quality

11 11 Example: Structural Alternatives Lower Complexity/DivergenceCurrent ProcessHigher Complexity/ Divergence No reservationsTake reservationSpecific table selection Self seat, menu on blackboardSeat guests, give menuRecite menu: describe choices EliminateServe water and breadAssortment of meze & bread Customer fills out formTake orders, prepare ordersAt table Pre prepared-no choiceSalad (4 choices)Individual prep at table Limit to 4 choicesMain dish (15 choices)Expand choices, bone fish at table etc. Ice cream bar-self serviceDessert (6 choices)Expand choices Serve salad and main dish; Dessert and bill together Serve ordersSeparate service or orders; change plates Cash only, pay when leavingCollect paymentChoice of payment, serve karanfil & kolonyali mendil

12 12 Conjoint Analysis: Motivation  Objective: max profits=revenues-costs  Positioning (or repositioning) impacts both profits and costs  We said earlier: in a service concept all details matter –What do customers value? –How are trade-offs between attributes made? –Etc.

13 13 Conjoint Analysis  Conjoint: joined together, combined  CONsidered JOINTly

14 14 What’s So Good about Conjoint?  More realistic questions: Would you prefer Horsepower or 140 Horsepower 17 MPG 28 MPG  If choose left, you prefer Power. If choose right, you prefer Fuel Economy  Rather than ask directly whether you prefer Power over Fuel Economy, we present realistic tradeoff scenarios and infer preferences from your product choices

15 15 Conjoint Analysis  Basic idea: the service can be broken down into a set of relevant attributes  Have consumers react to a number of alternatives  Infer –Importance –Most desired level  Estimation of an individual’s value system  Overall product judgements lead to value system through some data analysis technique

16 16 Services broken down into attributes  Credit card Brand + Interest Rate + Annual Fee + Credit Limit  On-line brokerage Brand + Fee + Speed of Transaction + Reliability of Transaction + Research/Charting Options  Ski area for ski resort pysical setting, distance, snow base, new snow, vertical drop, type of runs, challenge, size of area, facilities, ticket price, wait for lifts, type of lift, snowboards

17 17 Attributes have levels  Levels are mutually exclusive  Have unambiguous meaning  Keep number of levels low (3-5)  Try to balance number of levels across attributes

18 18 Example adapted from: J. Curry  Golf balls: driving distance, ball life, price  Alternatives –275 yards, 54 holes, $1.25 –250 yards, 36 holes, $1.50 –225 yards, 18 holes, $1.75  Market’s ideal ball?  Ideal ball for manufacturing costs?

19 19 Rank the balls  Distance –275 yards Rank 1 –250 yards Rank 2 –225 yards Rank 3  Ball Life –54 holes Rank 1 –36 holes Rank 2 –18 holes Rank 3  Doesn’t really tell us anything we didn’t know

20 20 Take 2 features conjointly Buyer 154 holes yards Buyer 254 holes yards Note: different tradeoffs made by each buyer. Only best and worst are the same.

21 21 Illustration by example (source: Dolan 1999)  Fitness facility design –Towel service: yes or no –Locker service Small storage lockers permanently assigned plus large hanging ones for daily use Mid-size only permanently assigned No permanently assigned locker, large hanging locker with mirror inside door

22 22 Rank from most to least preferred YesNo Small storage, large daily Rank 2Rank 4 Medium storage only Rank 1Rank 3 Large daily with mirror only Rank 5Rank 6 Towel Service Locker

23 23 Give utility points YesNo Small storage, large daily 42 Medium storage only 53 Large daily with mirror only 10 Towel Service Locker Avg

24 24 Value system ProductValue System Score Value system score rank Stated original rank MSO+Towel4+3.33= SSLD+Towel3+3.33= MSO+No Towel = SSLD+No Towel = LDMO + Towel = LDMO+ No Towel =2.1766

25 25 Question you can answer  Would this customer trade-off a storage locker on a daily basis for towel service?  Loss:  Gain:

26 26 In sum  Collect tradeoffs  Estimate buyer value system  Make choice prediction

27 27 Example: Output analysis (source: Montgomery and Wittink, 1979) Business Travel <= 1 night nights.109 >=6 nights Geographic Area East.070 Midwest South West.449 Opportunity for Advance Rapid.216 Moderate Range:.436 Range:.770Range:.432 Attribute importance for business travel:.436/( ) Importance analysis only relevant if attributes are in relevant ranges

28 28 What we can’t say about the utilities (part worths)..  >= 6 nights is unattractive to respondents  West is almost 7 times more attractive than East  <=1 night is more attractive than East  Why? –Arbitrary scaling within each attribute –Here utilities are scaled to sum to 0 within each attribute –Interval data does not support ratio operations –If count based then can say West is chosen 7 times more than East

29 29 Conjoint Importances  Measure of how much influence each attribute has on people’s choices  Best minus worst level of each attribute, percentaged: Vanilla - Chocolate ( ) = % 25¢ - 50¢( ) = % Totals: %  Importances are directly affected by the range of levels you choose for each attribute

30 30 Output analysis: PC Example (source Dolan) Weight <= 2 lbs lbs.9 >5lbs 0.0 BatteryLife 1 hr 0.0 2hrs 0.2 4hrs 1.5 8hrs 1.5 Resolution Below avg 0.0 Avg..4 Above avg..5 Price Product A: 2 lbs 1hr below average 2000 Product B: 5 lbs 4hrs average 3000 ProductC: >5lbs 8 hrs average 1000 Value of A= =1.7 Value of B = 1.9 Value of C = 3.0 Sum = 6.6 Share of preference approach: Prob. of choosing A: 1.7/6.6=26% Prob of choosing B: 1.9/6.6=29% Prob. of choosing C: 3.0/6.6=45% Market share: average purchase probability across all subjects

31 31 Output analysis  Aggregate analysis  Segmentation analysis  Scenario simulations

32 32 Market Simulation Example  Predict market shares for 35¢ Vanilla cone vs. 25¢ Chocolate cone for Respondent #1: Vanilla (2.5) + 35¢ (3.2) = 5.7 Chocolate (1.8) + 25¢ (5.3)= 7.1  Highest value choice (first choice rule): Respondent #1 “chooses” 25¢ Chocolate cone!  Repeat for rest of respondents...

33 33 Market Simulation Results  Predict responses for 500 respondents, and we might see “shares of preference” like:  65% of respondents prefer the 25¢ Chocolate cone

34 34 Example source: sawtoothsoftware  9 cards, ranked by 2 volunteers  Copy of Excel spreadsheet available from course web site

35 35 Traditional Conjoint Designs  Full profile: each service concept is defined using all attributes being studied  Full factorial: a design in which all possible product combinations are shown  Fractional Factorial: a fraction of the full factorial that permits efficient estimation of the parameters of interest) –From design catalogs –From software programs

36 36 Study design  Step 1: determine relevant attributes  Step 2: choose stimulus representations (how products will be described to respondents, full or partial)  Step 3: Choose response type (choice, ranking, rating)  Step 4: Choose criterion (liking, preference, likelihood of purchase)  Step 5: Choose method of data analysis

37 37 Summary  Blueprints for documentation  Analyze for complexity & divergence for positioning  Understand links between positioning and costs (service delivery system)  Conjoint analysis to assess customer valuations  Use output from conjoint analysis to link valuation, purchase, aggregate market share and profitability

38 38 Next time  Will continue Conjoint Analysis  Class will be held in the computer lab SOS Z13  Be on-time! Counts as in-class activity.  Will practice doing conjoint analysis via regression using Excel


Download ppt "1 OPSM 405 Service Management Class 9: Service System Design Tools: Service Blueprinting Conjoint Analysis Koç University Zeynep Aksin"

Similar presentations


Ads by Google