UV2972

t
Rev. Apr. 17, 2017

os
Portland Trail Blazers

rP
Teaching Note

Author Contact Information

yo
If you have questions about how to teach this case, you may contact Ron Wilcox at
wilcoxR@darden.virginia.edu or 434-243-5558.

Synopsis

The Portland Trail Blazers, a franchise in the National Basketball Association, is struggling to increase
op
attendance. It has contracted a marketing research firm to conduct a conjoint analysis focusing on several
aspects of its season-ticket offerings with the hope of improving the attractiveness of what it offers its fans,
and to do so profitably. The case presents the results of this research.

Use of the Case

tC

This case introduces students to conjoint analysis in a management decision setting. It is most effective
when paired with the technical note “A Practical Guide to Conjoint Analysis” (UVA-M-0675) and its
accompanying PowerPoint presentation (UVA-M-0675TNP).1 The case is appropriate both for a required
marketing management course at the MBA level as well as a marketing research elective at the MBA or
undergraduate levels. It is meant both to teach the technique of conjoint analysis and to help develop some
constructive skepticism about interpreting the output of such an analysis.
No

Suggested Discussion Questions

1. Which attribute does the conjoint analysis indicate is most important in the overall purchase
decision?
2. Are the conjoint results useful in making any pricing decisions? What useful information can be
gleaned from the research?
Do

1 Ronald T. Wilcox, “A Practical Guide to Conjoint Analysis,” UVA-M-0675 (Charlottesville, VA: Darden Business Publishing, 2003).

This teaching note was prepared by Professor Ronald Wilcox. Copyright  2009 by the University of Virginia Darden School Foundation,
Charlottesville, VA. All rights reserved. To order copies, send an e-mail to sales@dardenbusinesspublishing.com. No part of this publication may be reproduced,
stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the
permission of the Darden School Foundation.

This Teaching Note is authorized for use only by Michael Naraine, Brock University until Apr 2021. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or
617.783.7860.
 Page 2 UV2972

t
os
3. What about the promotional items? Are they worth giving away (free) to season-ticket holders?
4. What about the size of ticket packages? What does the conjoint analysis suggest?
5. What should the management of the Portland Trail Blazers do?

rP
Flow of the Class

The flow of the class will largely be governed by whether the case is taught in a traditional case study
environment or whether lecture material will be mixed in with the case discussion. In the latter instance, a
significant portion of the class may be spent discussing the aforementioned technical note. In the former
instance, or if the material in the technical note is sufficiently mastered before the case is taught, the
instructor can concentrate more of the effort on reasons why some conjoint analysis results provide more

yo
actionable information to management than others. The decision of how much time to allocate to the
technical note should be governed by your knowledge of the class. With a group of highly motivated students
who can be counted on to prepare the material prior to class, it is possible to do the entire case and answer
any potential questions about the technical note in a single (1.5-hour) class session. If this is not a safe
assumption, and you believe that you will need to spend considerable time lecturing the conjoint analysis
material, a single class session is not sufficient.
op
The flow presented below assumes a traditional case study environment and follows the assigned
discussion questions.

Analysis of the Case

1. Opening cold call: Which attribute does the conjoint analysis indicate is most important in the overall purchase
tC

decision?

This question involves computing attribute importances. Attribute importances are a standard metric in
conjoint analysis and are discussed in “A Practical Guide to Conjoint Analysis” (UVA-M-0675), which is
available for purchase through Darden Business Publishing (sales@dardenbusinesspublishing.com). This
opening question is essentially probing whether the called-upon student read and understood the technical
note, given that the answer itself is a rather mechanical application of that material to the case facts. The
No

importances measure the overall decision weight (importance) given to each attribute when the average
respondent is choosing their preferred season-ticket option.2 Applying the common metric used to measure
attribute importances (as given in the technical note):

Number of games = (0.24 + 0.28) ÷ ((0.24 + 0.28) + (1.00 + 0.65) + (0.73 + 1.01) +
(0.17 + 0.32)) = 0.12, or 12%

Ticket price = 38%

Location = 39%
Do

Promotions = 11%

2 The “part-worths” (utilities) in this conjoint analysis were estimated from a choice-based conjoint analysis in which respondents were shown two

full profiles and asked to select their most preferred. An example of the choice task is shown in the case Appendix.

This Teaching Note is authorized for use only by Michael Naraine, Brock University until Apr 2021. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or
617.783.7860.
 Page 3 UV2972

t
os
The qualitative results flow directly from the quantitative results. Within the tested range of the attributes,
ticket price and seat location are considered very important, while the number of games in a package and
ancillary promotions are less so. One interesting observation to make at this point is that the estimated
importance of the number of games in a package might be low because the descriptions of the attribute levels
are somewhat vague. For example, what does “one elite team and two very good teams” mean? The technical

rP
note emphasizes the need to make attribute levels unambiguous in their interpretation, and this particular
element of the conjoint design, the interpretability of the “number of games” levels, appears to fall short of
the ideal. One might reasonably suspect that attribute levels that are inappropriately vague might shift the
focus of the respondent toward attributes they understand, and hence show up in the form of a downward-
biased estimated importance for that attribute.

2. Are the conjoint results useful in making any pricing decisions? What useful information can be gleaned from the
research?

yo
The conjoint analysis can be used to explore the value that customers get from different seat
location/price combinations. They are the most important attributes these individuals consider, and by
looking at the utilities of the current offerings we can get a sense of where there are opportunities for
improvement.

The utilities of the current ticket location/price combinations are as follows:

op
 300BB + $15 = −0.076
 300C + $25 = −0.217
 300M + $35 = 0.283
 200M + $60 = 0.009
tC

So if they have any leeway on the upside, they might look to raise the price of 300M tickets, since this is
currently offering the highest utility to customers. In terms of lowering price, it appears that the 300C tickets
would be the best bet.

An interesting, albeit more advanced, analysis would be to examine how much they could raise the 300M
ticket prices to give them the same level of attractiveness as the next-best alternative (200M). That can be
accomplished by computing the spread in utilities between $35 and $60, which equals 1.129, and then
No

dividing this by (60 − 35) = $25. This calculation, 1.129 ÷ $25 = 0.045, provides us with the utility per dollar
within that range. The difference between the utility of the current 300M seats is 0.283, while the 200M is
0.009. This means that we can decrease utility by 0.283 − 0.009 = 0.274 to make them equally attractive. To
translate this into dollars, we divide the total change possible in utility by utility per dollar:
0.274 ÷ 0.045 = $6.07. This represents the maximum amount ticket prices could be raised while keeping this
seating option at least as preferred as the next-best current location/price combination.

Does this mean they should raise the price by this amount? Probably not, given current lagging ticket
sales. But it does mean that when management is considering raising ticket prices, something like a $2.50
Do

increase in the 300M tickets would be the first place to look.

3. What about the promotional items? Are they worth giving away (free) to season-ticket holders?

This analysis flows directly from the utilities given for the various promotional items and the cost of
those same items.

This Teaching Note is authorized for use only by Michael Naraine, Brock University until Apr 2021. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or
617.783.7860.
 Page 4 UV2972

t
os
a. Priority for playoff tickets is an obvious yes. It costs you nothing and you are probably going to
sell those games out no matter what.
b. The hot dog and soda combination costs you $3.25, per a contract with the concessions service.

rP
The value of this “gift” is 0.174 + 0.318 = 0.492. The spread in utilities across the entire range of
possible prices (a range of $45) is 0.656 + 1.002 = 1.658. So, on average, a free hot dog and soda
is worth about the same as 0.492 ÷ 1.658 = about 30% × $45 = $13.50, a drop in ticket price of
$13.50. This is a rather astounding number, far greater than the $3.25 it costs. This recalls some
of the classic results in the marketing academic literature arising from mental accounting that
suggest the value of small bonuses and prices is sometimes quite large. Bottom line, the data
strongly suggest the hot dog and soda idea is a very good one.

yo
i. A common question that occurs during this calculation is some version of “Why use
0.174 + 0.318 instead of simply 0.174?” Because the conjoint utilities are scaled to add up to
zero within an attribute, we can only interpret these utilities relative to some other attribute
level. The natural thing is to compare the hot dog and soda combination to an attribute level
in which nothing is given as a promotional item since this offers the least utility (−0.318)
The most reasonable way to interpret the value of the hot dog and soda combination in this
case is to difference it from the least-desired level, “no promotional item”; hence, we get
0.174 + 0.318.
op
ii. I think the above (i) makes a great class discussion point when students make an error in this
calculation (which they almost inevitably do). The point that needs to be emphasized to
students is that conjoint analysis never gives you the absolute utility of anything, only the
utility relative to the other levels that were tested. As such, the estimated utility of any
particular level of any attribute will be sensitive to the number of levels specified and the
value or description of those levels. Likewise, the estimated attribute importances—
tC

calculated earlier—are quite sensitive to the experimental design as well. If you add or delete
attributes, or even change the number of levels of a given attribute, the estimated attribute
importances will likely change.
c. The same analysis for the Trail Blazers jersey yields 0.318 + 0.002 = 0.320. Again, dividing by the
range of the price utilities: 0.320 ÷ 1.658 = 0.19, and converting to dollars, 19% × $45 = $8.55.
Given that the variable cost of the jersey is $12, the conjoint suggests that giving away a jersey is
No

not worth the cost. It might surprise students that a jersey is less valued than a hot dog and soda.
One possible explanation is that among the group sampled for this marketing research a
disproportionate number of them already owned Trail Blazers jerseys.
i. Issues of team branding and cannibalization are also appropriate issues to consider here. The
team jersey acts as a form of advertisement, so the team may receive benefits beyond what
the conjoint utilities suggest. This benefit may be particularly useful for a team currently
struggling with brand image.
ii. Free jerseys may also cannibalize regular jersey sales on which the team probably earns
Do

substantial margins. Taken in totality, the short-term financial argument stacks up strongly
against including the jersey as a free gift. The only argument in its favor is harder-to-measure
long-term brand-building benefits.

This Teaching Note is authorized for use only by Michael Naraine, Brock University until Apr 2021. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or
617.783.7860.
 Page 5 UV2972

t
os
4. What about the size of ticket packages? What does the conjoint analysis suggest?
a. At its most basic level, the research suggests that the six-ticket package is most preferred and the
ten-ticket package is least preferred; however, there are reasons to be skeptical about the results:
i. How are “elite” teams and “very good teams” going to be defined by management? The way

rP
the experimental design is constructed appears to guarantee that all teams in all packages are
going to be at least “very good.” Given management’s previous practice of setting up
packages that contained a small number of elite teams and some other less desirable teams, it
seems unlikely that all packages would or could be constructed such that all teams are very
desirable match-ups.
ii. The result that three-game packs are less desirable (0.03) than six-game packs (0.24) lacks
some face validity. Some students will probably realize that, given the way the data are

yo
presented, a six-game pack could be constructed from two three-game packs. The number of
“elite” versus “very good” teams would be exactly the same whether one six-game pack was
chosen or two three-game packs were chosen. What this suggests is that respondents were
probably making some inferences about the price of tickets in this situation, particularly that
the price per ticket would be lower in the six-game pack relative to the three-game pack. Yet,
the conjoint analysis does not present results for the interaction between game- pack size
and price. Because respondents’ views of this might be different from what management
decides to offer, results on ticket pack size are not particularly helpful for decision making.
op
5. What should the management of the Portland Trail Blazers do?

This final, open-ended question is meant to start a discussion about the takeaways from the conjoint
analysis. First, if it has not come up previously in the discussion, it should be noted that this analysis was
performed by people who had already demonstrated an interest in the Portland Trail Blazers. In that sense,
the results should be viewed more as a way to increase engagement with those who already have some
tC

positive disposition toward coming to a Trail Blazers game. The results do not speak to the person on the
street who has no interest in attending an NBA basketball game, nor to current fans who continue to
patronize the team. What we have are some results from the “middle group,” interested but not tremendously
interested in the product we are offering.

You should probe students about what results deserve to be taken seriously by management and what
results might just be an artifact of this particular conjoint design. In my view:
No

 Exploring ways to raise/lower prices on different seat locations is a good use of this analysis.
The research suggests that these attributes are important and the differences in utilities between
various location/price combinations are large enough to warrant action by management. At a
minimum, they should probably lower prices on 300BB seats.
 The analysis of the promotional items is also useful. The free food items are a cheap and easy
way to build some goodwill with the fans. The jersey seems too expensive given its value to the
customers. Management should offer priority for playoff tickets to individuals who purchase
ticket packages, particularly the larger packages.
Do

 For reasons discussed above, students should be skeptical about using the results of the conjoint
analysis to make decisions about the size of the ticket package.

This Teaching Note is authorized for use only by Michael Naraine, Brock University until Apr 2021. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or
617.783.7860.
 Page 6 UV2972

t
os
Epilogue

The Portland Trail Blazers used the results of this research to implement a more robust program of
promotional items. The real story, though, is how the team has improved both its performance and its
image—which no doubt helped to draw more fans. The following is an excerpt of an update that Trail

rP
Blazers management sent to me:

Four years after the completion of the conjoint analysis and case study, the Portland Trail Blazers are one
of the NBA’s biggest success stories.

Three years removed from a 21-win season, the team is on a pace to win 50 games and reach the
postseason for the first time since 2003, despite [being] one of the NBA’s youngest teams.

yo
Players more notorious for their off-court transgressions than their on-court contributions were
jettisoned, and the team was reconstructed with young, socially conscious and community-minded players
who happily answer the team’s call to “Make It Better” in the Portland community.

No longer in bankruptcy, the Rose Garden is rocking. The building has hosted 55 consecutive sellouts
[through February 2009], and is one of the NBA’s loudest and best-attended arenas.

In 2008, the Trail Blazers won the University of Massachusetts esteemed PRISM award for “Best Sports
op
Team.”
tC
No
Do

This Teaching Note is authorized for use only by Michael Naraine, Brock University until Apr 2021. Copying or posting is an infringement of copyright. Permissions@hbsp.harvard.edu or
617.783.7860.