Media Consumption and Cross Media Synergies
Transcript of Media Consumption and Cross Media Synergies
Media Consumption and CrossMedia Synergies
Kalyan RamanNorthwestern University, Evanston, IL
Presented at the Center for MeasurableMarketing, Measurable Marketing in aDigital World, NYU, May 27th , 2011
And now…
GPS is enabled by Einstein’sTime-Dilation formula
Our Worldis driven byHigh-Tech& ExoticScience
From the EMAC 2007Conference Theme (Verbatim)
• Now the challenge is to fightsimultaneously on several frontsagainst fragmentation, saturationand a storm of novelties that appeardaily in increasingly diversifiedmarkets.
IMC is increasingly importantbecause…
Integrated MarketingCommunications (IMC)
– 2 + 2 = 5 phenomenon
An IMC program plans and executes variousmarketing activities with consistency so thatits total impact exceeds the sum of eachactivity’s impact.
An IMC program takes optimal advantageof synergies between media
Media Planning Issues• Budgeting $$• Allocation Across Media• Allocation Over Time—Increase, Decrease, Constant,
Pulse, Other?• Allocation under time-varying effectiveness of media• Online and Offline media• Consumers control media consumption• Consumer-created synergy• Social media
A Deterministic Dynamic Model of IMC (Oligopoly)(Prasad Naik, Kalyan Raman and Russ Winer,Marketing Science 2005)
A Stochastic Dynamic Model of IMC (Duopoly & Uncertainty)(Kalyan Raman and Prasad Naik,Review of Marketing Science 2004)
Research on IMC
A Deterministic Dynamic Model of IMC (Monopoly)(Prasad Naik and Kalyan Raman,Journal of Marketing Research 2003)
Temporal Allocation in IMC
• In Automatica (2006), I showed that constant, increasingor decreasing spending patterns can be optimal for thesame response function
• But the assumption of a single medium is restrictive• IMC uses multiple media• And the assumption of constant effectiveness is limiting
because media effectiveness vary over time• How should budgeting and allocation be optimized over
multiple media with time-varying effectiveness?• The answer is relevant to managing IMC optimally during
recessions
Marketing EffectivenessCan Vary over Time
• Changes over product lifecycles, e.g., Competitors enter or drop out Consumer values shift and behavior patterns
change Different customer segments are attracted to
the product at different stages in productcycle
Changes over business cycles Economic upturns or downturns can create
changes in demand16
Empirical Evidence of Time-VaryingMarketing Effectiveness
• Advertising elasticity changes over time (Parsons1975, Erickson and Montgomery 1980)
• Advertising copy quality (Naik, Mantrala, Sawyer1998)
• Price elasticity and promotion effects (Jedidi, Mela,Gupta 1999, Krishnamurthi and Papatla 2003)
17
18
Dynamic Model ofTime-Varying Effectiveness
Advertising Sales
PersonalSelling
2 t1 t
Profitm(t)c(t) T
Sales Responseparametersvary over time
A Dynamic SalesResponse Model
19
ttttt vtutSS )()( 211
Advertisingspending
Sales forcespending
SalesCarryover
Time-varyingAdvertisingeffectiveness
Time-varyingsales forceEffectiveness
Random Error
•β1 (t) and β2(t) are continuous and differentiable at least once
20
Profit Maximizationwith Time-Varying Parameters
T
0
22t dt))t(cv)t(u)t(mS(eMaximize
subject to
Find optimal strategy (u*, v*) to
)()()()()( 21 tvttuttSdtdS
salvage value: S(T) =e- t mSθ
+S(T)
Empirical study
• Data:– 5 years of monthly data on a major pharmaceutical
product in the US: detailing, journal advertising effort
– Typical life for a patented pharmaceutical product is 12years. Data period starts at the beginning of the 7th yearafter product launch.
21
• Selling effort effectiveness decreases over time,consistent with literature (Albers, Mantrala and Sridhar2008; Manchanda and Chintagunta 2004)
• Advertising effectiveness increases over time
Kalman Filter Results
22
Implications for OptimalSpending Trajectories
• At t=50, investment emphasis would switch from sales force toadvertising.
23
0
20
40
60
80
100
120
140
160
180
0 20 40 60
ten
thou
sand
time
sales force
advertising
Comparison with a ConstantParameters Model
• Optimal policy will always call for more spending onsales effort (u) than on advertising (v).
24
(Raman, Mantrala, Sridhar & Tang,Forthcoming, JIM, 2011)
The World Marketers Inhabit Today
Marketer Customers/Prospects
Competitors
Competitors
Competitors
Competitors
Messages and Incentives
Products and Services
Word-of-Mouth New Forms of Media
Web Search
Empl
oyee
s/R
ecom
men
ders
/Frie
nds/
Influ
ence
rs
Agency Media Sales Force Retail
59
Responses Occur Behind the Marketing Lines
Organization Agencies Messages Consumer
BrandExperiences
Internet Systems
Electronic SystemsBrand
Experiences
• Web sites• Customer Service• Tech support• Retailers• Distributors• User communities• Blogs• RSS• Influencers• Recommenders
60
Response
MediaExposure
MediaExposure
MediaExposure
M ediaConsum er
Potentia l SynergyV ia S im ultaneous M edia Usage
(M essages Reinforced)
Potentia l F ragm entationO f A ttention
(M essage Im pact D iluted)
Foreground/Background Media
Tim e A llocated To Each M edia Form
A Media Consumption Model
Schultz and Pilotta, ESOMAR, 2004 106
Four Critical Measures in MediaConsumption
• What media form is accessed• Amount of time spent with each media form• Media combinations – what media used
together and simultaneously• Which media form has the greatest influence
on product purchase
Time US Consumers Spend With Each MediaForm
Time US Consumers Spend With Each MediaForm
Avg Minutesper Day
Internet 146.3Email 132.3TV 124.1Radio 70.7Direct Mail 54.0Video Games 48.7IM 48.5Magazines 46.7Newspapers 40.5Satellite Radio 19.6Web Radio 18.7Blogs 14.4
Source: 1st Half, 2010 – BIGresearch, Inc.
Measures #1 and #2