University of Alberta Marketing Camp,

1
Predicting Consumers Adoption and Use of New and
Really New Entertainment and Communication
Technologies

• David Alexander John Lynch
• Duke University
• Qing Wang, Warwick Business School

2
Really New Products

• GO Corp Pen PCs in late 80shyped 75 MM failure
• MSI Priority Understanding how to design,
  market, and forecast Really New Products
  (Lehmann 1994)
• Definitions of newness
• Chronology, Technology, Psychology
• New to market v. new to firm (Booz, Allen,
  Hamilton 1982)
• e.g., Goldenberg, Lehmann, Mazurski (2001) find
  that moderately new-to-market products more
  successful
• Psychology
• compatibility, complexity, communicability
  (Rogers 1983)
• loss aversion from change (Gourville 2005)
• Inapplicable existing category knowledge (Moreau,
  Lehmann, Markman 2001 Moreau et al. 2001 Wood
  and Lynch 2002)
• Hoeffler (2003) Newness as uncertainty and
  constructive preferences

3
Really New Steve Hoeffler
4
Differences Between Really New Products and Just
New Products

• developed from Hoeffler 2003, JMR
• RNPs are, in general, products about which
  consumers have little personal knowledge and
  little access to new information

5
Differences Between Really New Products and Just
New Products

• developed from Hoeffler 2003, JMR
• RNPs are, in general, products about which
  consumers have little personal knowledge and
  little access to new information

6
Research Questions

• How do characteristic differences between
  Really-New Products (RNPs) and Just-New Products
  (JNPs)
• affect the likelihood of forming long-term
  product acquisition intentions?
• affect the likelihood of following through on
  those intentions?
• affect the timing of follow-through?
• affect how much products are used once they are
  adopted?

7
Newness Scale Items

• I feel quite certain of the benefits I could
  expect to get if I bought (adopted) this
  product/service (reverse coded).
• Im quite sure of what the relevant tradeoffs are
  among the costs and benefits of buying and using
  this product/service (reverse coded).
• Ill have to change my behavior significantly to
  attain the potential benefits of this new
  product/service.
• Using this new product/service would allow me to
  do things that I cant easily do now.

8
Newness Pretest

• 12,237 email invitations sent to members of CBS
  Television City Online Panel in August 2004
• 2,693 replied and participated mean age 37 58
  Female
• Respondents asked if they had already acquired 28
  new entertainment communication tech products
  and services identified by CBS (Yes 24,925 No
  50,479)
• If No, then asked
• Intend to acquire in next 6 months?
• Yes (5,207) asked to rate product on 4 newness
  dims (a .95)
• Mean newness across respondents used to classify
  products, not peoplea measure of aggregate
  psychological response by people in the market
  rather than a measure of individual perception
  (cf. LaBay and Kinnear 1981 Moreau, Lehmann,
  Markman 2001)

9
Newness Correlations Individual Level
10
Newness Correlations Product Level
11
Classify Product Newness

• RNP (12 w/ highest newness scores)
• e.g., Blogging service, Streaming TV, PDA,
  Digital video recorder (DVR), DVD-by-mail service
• JNP (10 w/ lowest newness scores)
• e.g., Flat screen TV, DVD player, home theatre
  system, broadband internet service
• neither RNP nor JNP (6 w/ intermediate scores)
• e.g., Camcorder, Video Game Player, MP3 player
• These 6 products were dropped in Study 2

12
Pretest JNPs
13
Pretest RNPs (newest last)
14
Pretest Neither JNP nor RNP
15
Validation of Newness

• Average Newness as rated by those who do not now
  own but who intend to buy in next 6 months
  predicts proportion of people in the panel who
  already own the product.
• Binary logistic regression (1own, 0no own)
• f(newness, participant dummies) shows ownership
  declines with mean newness
• Repeating similar analysis replacing the
  composite mean newness with each component shows
  ownership
• Declines with average uncertainty_of_benefits
• Declines with average uncertainty_of_cost-benefit_
  tradeoffs
• Declines with average need_to_change_behavior
• Increases with average new_things_possible

16
Study 1 Newness Affects Acquisition Intentions

• How does product newness affect whether people
  form acquisition intentions?
• Intention to acquire should be lower for
  psychologically newer products. Follows from  
• Missing information is a negative (Simmons and
  Lynch Jaccard Wood Johnson Levin Meyer)
• Uncertainty negatively correlated with feature
  importance (Hoeffler, Moreau,
  Kubowicz-Malhotra)
• If benefits of RNPs less certain, then temporal
  construal theory implies that intent to buy in
  distant future should be lower for RNPs
  (Liberman, Trope, and Stephen)

17
Study 1Method

• Same 2,693 members of CBS Television City Online
  Panel from pretest
• Respondents asked if they had already acquired 28
  new entertainment tech products and services
  (Yes 24,925 No 50,479)
• For 50,479 No responses, asked if Intend to
  acquire in next 6 months? (DV 1 Yes, 0 No)
• Yes (5,207), No (45,272)
• Intenders mean newness scores used to predict
  proportion intending to acquire across entire
  sample (10.3 on average)

18
Study 1 Analysis

• Logistic regression predicting prob of stating
  intent to acquire in next 6 months (1 Yes, 0
  No)
• Individual Respondent dummies
• Product level IV Newness using aggregate mean
  newness score for a given product from subset of
  sample intending to acquire
• Results
• Probability of stating intention to acquire
  declines sharply with newness (t -30.9)
• Predicted prob of intending to acquire 4-5 times
  as high for least new products as for most new
  products

19
Consumers less likely to express intention to
acquire newer products.
?2(1) 760.2, p lt 0.001
20
Components of Newness

• Similar analyses run replacing overall newness
  with the component allowed to be weighted
  unequally
• All components have negative effect when entered
• If just new things possible and have to change
  behavior entered, new things possible is
  positive and change behavior negative

21
Conclusion of Study 1

• Psychological newness discourages people from
  intending to acquire new technology products
• I feel quite certain of the benefits I expect to
  get from my new product (reversed)
• I can easily evaluate whether the benefits of
  acquiring and using product outweigh the costs in
  money, time, and hassle (reversed)
• I'll have to change my behavior significantly to
  get the full benefit of product
• Using product will allow me to do things that I
  can't easily do now

22
Study 1 Conclusions (cont.)

• Our results on adoption of entire technologies
  consistent with others work looking at
  individual new products
• Goldenberg, Lehmann, Mazursky find that radically
  new-to-market products fail more than moderately
  new
• Hoeffler, Moreau, Kubowicz-Malhotra find that
  positioning a really new product on less new
  attributes leads to more acceptance

23
Study 2 Follow-Through

• Stated intentions often used in new product
  forecasting (e.g., BASES,Morwitz)
• Are people less likely to follow through on
  stated intentions to acquire a new technology if
  it is really new?
• Test effect of product newness on
• follow-through on stated intention to acquire
• acquisition timing

24
Effect of Newness on Follow-Through

• Intentions at a temporal distance may appear
  unwise when the event draws nearer in time
• Zauberman and Lynch (2005) and the YesDamn!
  effect Soman (1998) on rebate non-redemption
• Second-guessing more likely for more
  psychologically distant events
• Construal theory (Liberman, Trope, Stephen)
• Newer more abstract, distant (shown later)
• As acquisition draws near
• Can do new things gets less weight
• behavior change cons get more weight
• Hence, lower follow-through for RNP than JNP

25
Temporal Construal Follow-Through
26
Study 2

• Follow up with Study 1 respondents, 4 months
  later
• 1,622 (60) had indicated intent to acquire at
  least 1 RNP or JNP in next 6 months and were
  invited to participate in Study 2
• 38 (620) agreed to participate (52.7 women
  mean age 38 (range 11 to 77) owning an average
  of 4.3 JNPs and 3.4 RNPs at pretest
• Respondents were shown each product/service that,
  in Study 1, they had expressed an intention to
  acquire in next 6 months
• Asked if they had yet acquired
• If Yes, asked what month (if no, recruited for
  Study 5)
• Acquisition predicted by aggregate pretest
  newness scores

27
Effects of Newness on Follow-Through

• Binary logit (1acquired, 0 not acquired)
  f(product newness participant random effects)
• Consistent with H2, participants who in August
  had stated an intent to acquire in next 6 months
  less likely to have followed through the newer
  the product
• Twice as likely to follow through for least new
  as for most new products

28
(No Transcript)
29
Components of Newness

• Similar analyses run including all four newness
  components as separate predictors
• 2 of 4 components are significant and signed
  correctly
• uncertainty about benefits and uncertainty about
  benefit-cost tradeoffs have no effect
• Have to change behavior significant and negative
  (feasibility / constraints loom larger when
  behavior draws near than at temporal distance)
• Can do new things is significant and negative
  (desirability / benefits seem less compelling
  when behavior draws near)

30
Implications

• In market research, we often use intention-to-buy
  to predict actual adoption, using standard
  deflators
• Really New products require more deflating
• Consistent with construal account
• When newness replaced in model by each component
• Must change behavior sig neg effect (feasibility
  looms larger)
• New things possible has neg effect (desirability
  looms less large than at 4 months)

31
Study 2 Newness and Timing of Follow-Through

• For those who do follow through, how does newness
  affect the timing?
• 620 respondents who said they intended to acquire
  one or more of our new products in 6 months. 4
  months later, asked
• If they had acquired, and if so
• What month? (1, 2, or 3 months after stated
  intent)

32
Study 2

• Discrete-time non-proportional hazard rate
  function for intention follow through
• Allison 1995

33
Each month after intentions stated, people become
more likely to follow through on their JNP
intentions, less likely to follow through on RNP
intentions
Conditional Acquisition Probability
34
Implications

• Marketers of JNPs can build momentum with
  consumers by marketing well in advance of product
  launch
• Marketers of RNPs cannot
• In using intentions for market research,
  appropriate deflators more different for really
  new v. just new products the farther out in
  time

35
Why Would Probability of Follow through Change
Over Time?

• If people state an intent, they have
• Favorable attitude
• Temporarily accessible in memory
• Accessible attitudes lead to selective perception
  of new information, so that new info only
  confirms initial attitudes (Fazio Williams 86)
• What determines whether boost in attitude
  accessibility stays or dissipates?
• Frequency of exposure to new info
• JNP Frequent ? sustained attitude accessibility
  (Dholakia Morwitz) ? only attitude consistent
  new info comes in
• RNP Infrequent ? intention-induced spike in
  attitude accessibility dissipates ? no filter on
  new info

36
Timing of Follow-Through

• Forming intentions affects attitude accessibility
  
• Exposure to new information affects attitudes and
  beliefs and attitude accessibility

JNPs
Accessibility Threshold
RNPs
Accessibility Threshold
37
The Effect of New Information on the Timing of
Acquisition Intention Follow-Through
38
Study 3

• Test effect of product newness across time on
  attitude accessibility and exposure to new
  product information
• exposure to new information 4-item scale
  standardized within subjects across products
  (a0.57)
• attitude accessibility Response latencies
  captured for responses to attitude measure
  (a0.81) (Fazio et al. 1989)
• For products participants indicated intent to
  acquire
• product newness same product newness scale as
  Study 1, 22 new products (a0.91)
• 0.87 correlation of newness scores for the 11
  common products

39
Study 3

• 107 Daytime Duke MBA students recruited for 2
  session research study and paid 15 for
  completing both sessions
• 74 (79) completed 2nd session
• First session was conducted prior to students
  Fall break
• Second session was conducted in first week after
  Fall break

40
Study 3

• In first session, participants were presented
  with list of 22 new entertainment/tech
  products/services
• Identified those they owned and those they
  intended to acquire in the next 6 months
• Responded to product newness and accessibility
  items
• For each product/service not owned
• Indicated how informed they felt about the
  product/service
• Davidson et al. 1985
• Responses excluded where participants indicated
  feeling very uninformed about product

41
Study 3

• In 2nd session, participants saw the
  products/services they said they did not own and
  that they felt at least moderately informed about
  in 1st session
• Responded to same attitude accessibility items
• Responded to exposure to new information items

42
Study 3

• Univariate ANOVA
• Product newness
• b -0.33, p lt 0.001

43
As time passes after intentions are formed,
people encounter more new information for JNPs
than RNPs
44
Implication

• Supports Hoefflers (2003) claim that RNPs have
  less developed information networks
• Perhaps this explains why a heightened sense of
  attitude accessibility persists for JNPs but not
  RNPs

45
Study 3 Attitude Accessibility

• Does attitude accessibility decline more steeply
  over time for really new products?
• Ln (Attitude Accessibility (t2))
• f(average newness rating of product, ln(Attitude
  Accessibility (t1)), participant dummies)
• Longer res
• Consistent with our hypotheses, attitude
  accessibility declined more rapidly the newer the
  product (b 0.15, p .03)

46
Study 3 Attitude Accessibility
47
A Problem

• Our story is that attitude accessibilty was less
  at time 2 for really new products than for just
  new products because of the greater exposure to
  product-relevant information in the intervening
  time for JNPs
• Alternatively, RNPs get more very-short term
  boost in accessibility because of greater
  elaboration necessary to form intention judgment
  for RNPand this dissipates in seconds for
  reasons unrelated to info exposure between t1 and
  t2.

48
Study 4 Attitude Accessibility Redux

• Replicate attitude accessibility findings in a
  3-wave study with 48 MBAs
• Wave 1 Measure intentions, newness, attitude
  accessibility (newness means a .84, means
  correlate with Study 3 .81)
• Wave 2 (1-2 days later). Measure attitude
  accessibility again for all products not owned
• Wave 3 (4 weeks later). Measure attitude
  accessibility again
• Ln (RT3) f(product mean newness, ln(RT2),
  participant dummies, Time 1 Intention (1 or 0).
• Newness effect significant and positive (b .14,
  plt.001)

49
Study 4
50
Conclusions from Studies 3 and 4

• Forming intentions for really new products
  heightens attitude accessibility, which has been
  shown in past research to affect fulfillment of
  stated intentions (Fitzsimons Morwitz)
• Spike in attitude accessibility declines without
  energizing events (Dholakia Morwitz)
• exposure to product info between intent and
  decision
• Less post-intent exposure for RNPs
• More rapid decline in exposure for RNPs
• Implies tough to use pre-launch hype to produce
  much-delayed purchase of an RNP

51
How Does Newness Affect Use of New Technologies
Once Acquired?
52
After the Box Has Been Opened Consumers Use of
RNPs

• All published work on really new products (RNPs)
  has focused on what consumers think about such
  products at the time of purchase
• We know nothing about the drivers of actual use
  of RNPs after the sale
• It is obviously to the sellers (and the buyers)
  advantage if consumers make extensive use of
  really new products after purchasing them
• Many high tech gadgets wind up unused

53
Study 5 Abstract Thinking about Use of RNPs

• Do people think more abstractly about RNPs
• 1 week before acquisition, we asked about
  expected use in the first week after acquisition
• 10 days after acquisition, we asked about
  expected use in the next week
• Open endedcoded for abstractness, activity by
  description v. description by activity (Trope
  Liberman 2003)

54
Abstractness of Codings

• Ps generated several lines describing anticipated
  use of to-be-acquired produc
• Decomposed into sentences
• If sentence describes something that can be done
  by using the new product, superordinate 1
• If sentence describes new product being used by
  doing some activity, subordinate -1
• Ambiguous or unclassifiable 0
• Scores for all sentences by P averaged to provide
  construal level score (M .42)
• Abstractness greater for newer products (p lt
  .002)

55
Study 5 Unrealistic Expectations of Use of RNPs

• Are People Worse at Forecasting Use of Really New
  Products?
• 1 week before acquisition, we asked about
  expected use in the first week after acquisition
• Closed-ended
• What of available features will you use in week
  1?
• On how many separate occasions will you use the
  product in week 1?
• How many hours will you spend using the product
  in week 1?
• How many hours on instructions in week 1.
• 10 days after acquisition, asked how much actual
  use in that first week
• overestimation (estimated actual)/estimated

56
Inaccurate Estimates of Features Used for RNPs
57
Inaccurate Estimates of Time Spent Using RNPs
58
r(expected use occasions, actual use
occasions). Z -2.13
59
Implication

• Consumers acquiring really new products are not
  well calibrated in expectations about early use
• This can cause shocks that lead to negative word
  of mouth I didnt use it as much as I
  expected.

60
Differences Between Really New Products and Just
New Products

• developed from Hoeffler 2003, JMR
• RNPs are, in general, products about which
  consumers have little personal knowledge and
  little access to new information

61
Summary Implications for Marketing of RNPs

• Study 1Anything that a firm can do that can make
  an RNP seem less new will help
• Study 2Cant create buzz for RNP far in advance
  and hope that intentions will be formed that will
  carry through
• Studies 3 4. The problem for RNPs is that
  information is too sparse to keep attitude
  accessibility elevated. Hence, if you have a
  monster budget (e.g., X-Box), pre-launch hype may
  maintain
• Study 5Consumers poorly calibrated about degree
  of use of RNPs expectations likely to be violated

62
Summary Implications for Market Research on RNPs

• Study 2
• Follow through on intention far less for RNPs
  need larger standard deflators
• Differences in follow through more evident after
  several months
• Study 5
• Consumer estimates of expected use have very
  little map to reality for RNPs
• Much better calibrated for JNPs

63
Why Are These Effects Happening?

• Liberman, Trope, and Stephen
• Distance? focus on desirability, not feasibility
• RNPs more abstract and distant
• Integrating Hoeffler with Trope Liberman
• RNPs more uncertainty about benefit about
  benefits / costs
• RNPs must change behavior to get benefit
  (feasibility looms larger when behavior near)
• RNPs allow me to do new things (desirability
  looms less large when behavior near)

64
Thank you
65
Components of Newness (1Intend, 0 No Intend)

• Standard Wald
• Parameter DF
  Estimate Error Chi-Square PrgtChiSq
  Exp(Est)
• Intercept 1 -2.1519
  0.0186 13443.7568 lt.0001 0.116
• avg_uncertainty_adj 1 -2.1990
  0.1674 172.5341 lt.0001 0.111
• avg_change_behavior_ 1 -0.8222
  0.1379 35.5224 lt.0001 0.439
• avg_new_things_possi 1 -0.4735
  0.0585 65.6235 lt.0001 0.623
• respage_adj 1 -0.0204
  0.00160 163.5973 lt.0001 0.980
• respsex 0 1 -0.1808
  0.0176 105.7070 lt.0001 0.835
• innovativeness_adj 1 0.4621
  0.1614 8.1973 0.0042 1.587
• own_tot_adj 1 0.1033
  0.00452 522.7176 lt.0001 1.109
• innovativown_tot_ad 1 0.0972
  0.0339 8.2098 0.0042 1.102
• Partial effects show that all components of
  newness predict lower stated intent.
• Also, intent higher for younger, male, already
  owning lots of 28 technologies, especially those
  with high newness scores.

66
Logistic Regression (1Acquired 0 Not Acquired
by 4-month mark)
67
As time passes after intentions are formed,
people become more likely to follow through on
their JNP intentions, less likely to follow
through on RNP intentions
Conditional Acquisition Probability