nach oben

Quantitative Marketing and Economics

Erschienen in:

01.09.2015

Effect of temporal spacing between advertising exposures: Evidence from online field experiments

verfasst von: Navdeep S. Sahni

Erschienen in: Quantitative Marketing and Economics | Ausgabe 3/2015

Einloggen

Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.

search-config

KI-gestützte Suche

Aus

Abstract

This paper aims to understand the impact of temporal spacing between ad exposures on the likelihood of a consumer purchasing the advertised product. I create an individual-level data set with exogenous variation in ad exposure and its spacing by running online field experiments. Using this data set, I first show that (1) ads significantly increase the likelihood of the consumers purchasing from the advertiser and (2) this increase carries over to future purchase occasions. Importantly, I also find evidence for the spacing effect: the likelihood of a product’s purchase increases if it’s ads are spread apart rather than bunched together, even if spreading apart involves shifting some ads away from the purchase occasion. Accounting for the spacing effect is important to detect the effects of repeated advertising. Because the traditional models of advertising do not explain the data patterns, I build a new memory-based model of how advertising influences consumer behavior. Using a nested test, I reject the restrictions imposed by the canonical goodwill stock model (Nerlove and Arrow, Economica, 29(114):129–142, 1962), in favor of the memory-based model I propose. Additionally, I use the estimated parameters to simulate counterfactual scenarios and show that the advertisers’ profits might be lower if the features of the memory model are not accounted for.

Vorheriger Artikel Consumer learning and evolution of consumer brand preferences

Nächster Artikel Erratum to: Effect of temporal spacing between advertising exposures: Evidence from online field experiments

Sie haben noch keine Lizenz? Dann Informieren Sie sich jetzt über unsere Produkte:

Springer Professional "Wirtschaft+Technik"

Online-Abonnement

Mit Springer Professional "Wirtschaft+Technik" erhalten Sie Zugriff auf:

über 102.000 Bücher
über 537 Zeitschriften

aus folgenden Fachgebieten:

Automobil + Motoren
Bauwesen + Immobilien
Business IT + Informatik
Elektrotechnik + Elektronik
Energie + Nachhaltigkeit
Finance + Banking
Management + Führung
Marketing + Vertrieb
Maschinenbau + Werkstoffe
Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Jetzt informieren

Springer Professional "Wirtschaft"

Online-Abonnement

Mit Springer Professional "Wirtschaft" erhalten Sie Zugriff auf:

über 67.000 Bücher
über 340 Zeitschriften

aus folgenden Fachgebieten:

Bauwesen + Immobilien
Business IT + Informatik
Finance + Banking
Management + Führung
Marketing + Vertrieb
Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Jetzt informieren

Nur mit Berechtigung zugänglich

The experiments I report in this study are all the experiments that were conducted. Prior to these experiments, I conducted one pilot study which was a before and after comparison to judge the magnitude of the effect of ads. One experimental restaurant closed between planning and the data collection, therefore the focal restaurant had to be changed. After this paper, the data have been used for the analysis of the competitive effects of the experimental ads, described in Sahni (2014).

See, e.g., Varian (2007), Agarwal et al. (2011), Jerath et al. (2011), Varian (2007), Katona and Sarvary (2010), Narayanan and Kalyanam (2015).

I use the term “learning” here to describe retaining of information, as opposed to gaining new information / learning about quality as, e.g., in Erdem et al. (2008).

E.g., Mahajan and Muller (1986), Hahn and Hyun (1991), Villas-Boas (1993), Feichtinger et al. (1994), Feinberg (2001), Dubé et al. (2005), Doganoglu and Klapper (2006)

The website claims to cover all restaurants in these markets; indeed, for the markets this study considers, the number of restaurants in this website’s database is roughly 40 % higher than those returned by searching on the next major competitor.

If the advertiser is a chain, clicking on the ad banners takes the user to a page with links to web pages for the chain’s outlets in the filtered geographic area.

(A) contained different randomized sub-conditions where characteristics of ads, such as their position, were randomized. For analysis in this paper, I pool data from all sub-conditions of (A) to estimate average effects of ads.

These banners show the name of the advertiser and are not too informative; some displayed the restaurant’s logo and included a short three to four-word phrase describing it.

Calls were tracked for 8 out of 11 experiments. The callers were notified of this before the call was forwarded.

Nine out of the 11 experimental restaurants are chains. When the users click on ads, they are shown links to restaurants from the chain located in the geographic area of search. Hence, clicking on ads might not always lead to a visit to the restaurant page.

None of the effects in this paper are driven by just one market. All effects reported in this paper hold even when data from any one of the markets are excluded from analysis.

Recall that clicking on an ad on this website does not imply a restaurant page visit, if the ad is for a chain. Clicking on the ad takes the user to a page with links to the chain outlets in the area of search. The user may then choose to visit the restaurant page.

For example, if an individual gets exposed to the experimental ad on 3 pages, then n E x p = 3.

This issue is not discussed in previous marketing research on the impact of online search advertising (e.g., Ghose and Yang (2009) and Ghose and Yang (2010)).

For illustration, consider a set of individuals who browsed n pages. Within this set, individuals with higher ad exposure are the ones who continued browsing the website after being exposed to the ad a few times. Therefore, among the individuals who browsed n pages, ones with higher ad exposure might be the ones that have a lower preference for the advertiser. This possibility can lead to an underestimation of the effect of multiple ad exposures.

Also, I focus on individuals who have more than one day’s gap between sessions, so that the next session is likely to be a different purchase occasion. Because the number of users with multiple sessions is fewer, the dependent measure I use is a visit to the advertiser’s page; I take this step because the base level of sales lead is small, making precise estimation of the effects difficult.

An implicit assumption here is that consumers’ decisions to revisit the website are not correlated with past advertising exposure. I find no evidence for such correlation in the data, presented in Appendix B.

For analyzing carryover, I don’t include data from one experiment in which the advertiser was advertising before the experiments started.

The means I show here are not conditioned on n E x p ₁ for the purpose of simplicity. In the data, the effects are larger when n E x p ₁>0, as one would expect.

Estimating linear probability models using OLS instead of logit regression yields similar results for the model-free analysis throughout the paper.

I also estimated the specification in Column IV separately for the subsamples with (a) d a y s ₁₋₂≤7 and (b) d a y s ₁₋₂>7. The coefficient of n E x p ₂ is positive and significant for (b), but statistically indistinguishable from zero for (a).

Prior lab evidence of the spacing effect has been shown for time intervals ranging up to a week (Janiszewski et al. 2003).

I use OLS in this case, to show the average slope. Running a logit regression and estimating the implied marginal effect gives similar results.

More recent research builds on this approach by allowing for wearin and wearout of advertising - repeated ad exposure decreases the attention consumers pay to it (Naik et al. 1998; Bass et al. 2007). However, the data patterns I observe do not comply with these models; columns V and VI of Table 6 show that an increase in d a y s ₁₋₂ affects the carryover, but not the contemporaneous effect of n E x p ₂. Therefore, the empirical evidence suggests an underlying data-generating process relating to long-term retention of ad effects as opposed to attention.

The age of an ad at a particular point in time is the time elapsed since the ad exposure.

This assumption is made to apply the model because in the context of advertising, one exposure may not lead to the consumers noticing the ad, unlike forced exposure occasions in the lab or general learning situations in which agents deliberately rehearse at every occasion. Therefore, in this setting, I define a session as a learning occasion, and the number of ad exposures represent the memory strength due to the ads at that occasion.

I fix β _i = 0 (Eq. 2) because in the data, it is indistinguishable from the impact of $A_{it}^{M}$ on choice.

I use data for individuals who repeat their website visits with a time interval of more than a day. Also, I did not use data from the experiment where the experimental restaurant had advertised before the time of the study. This step is taken to avoid the measurement problems caused by unknown initial value of any prior ad effect.

Other specifications provide a similar inference, and are presented in the accompanying Online Appendix.

Online ad networks such as DoubleClick and Facebook can directly apply this approach by estimating the models as above and performing counterfactual simulations because they can control ad exposure at the individual level in real time. Some advertisers who have access to the advertising medium can also follow directly (e.g., Amazon.com, Ebay.com). For other firms, advertising planning involves managing aggregate levers that can be set in a similar manner by making assumptions about consumer media consumption (see Dubé et al. (2005)). In practice, media consumption data available through third-party companies such as Nielsen, comScore and HitWise empirically support these assumptions.

I assume constant spacing between all occasions. For these calculations, I assume the profits resulting from the restaurant page visit to be $60 and the marginal cost of an ad impression to be $0.1.

If the consumer browses five pages in each of the eight sessions in the eight weeks of the month, the number of possible choices of a freq. cap is 6, from 0 to 5, and 2⁸ combinations of switching the advertising on or off in any of the weeks are possible.

The integration is implemented using a Monte Carlo simulation method; I take draws of $\theta _{i}^{\psi }$ and $\xi _{i}^{\psi }$ from the distribution and compute the average value of $\pi _{i}^{\psi }$ for these draws.

Agarwal, A., Hosanagar, K., & Smith, M.D. (2011). Location, location, location: An analysis of profitability of position in online advertising markets. Journal of Marketing Research, 48(6), 1057–1073.CrossRef

Anderson, J.R. (1996). ACT: A simple theory of complex cognition. American Psychologist, 51, 355–365.CrossRef

Anderson, J.R., & Milson, R. (1989). Human memory: an adaptive perspective. Psychological Review, 96(4), 703–719.CrossRef

Anderson, J.R., Bothell, D., Byrne, M.D., Douglass, S., Lebiere, C., & Qin, Y. (2004). An integrated theory of the mind. Psychological Review, 111(4), 1036–1060.CrossRef

Bagwell, K. (2007). The economic analysis of advertising. Handbook of Industrial Organization, 3, 1701–1844.CrossRef

Bass, F.M. (1969). A simultaneous equation regression study of advertising and sales of cigarettes. Journal of Marketing Research, 6(3), 291–300.CrossRef

Bass, F.M., & Clarke, D.G. (1972). Testing distributed lag models of advertising effects. Journal of Marketing Research, 9(3), 298–308.CrossRef

Bass, F.M., Bruce, N., Murthi, B.P.S., & Majumdar, S. (2007). Wearout effects of different advertising themes: A dynamic bayesian model of the ad-sales relationship. Marketing Science, 26(2), 179–195.CrossRef

Blake, T., Nosko, C., & Tadelis, S. (2015). Customer heterogeneity and paid search effectiveness: A large scale field experiment. Econometrica (forthcoming).

Braun, M., & Moe, W.W. (2013). Online display advertising: Modeling the effects of multiple creatives and individual impressions histories. Marketing Science, 32(5), 753–767.CrossRef

Bruce, N. (2008). Pooling and dynamic forgetting effects in multitheme advertising: Tracking the advertising sales relationship with particle filters. Marketing Science, 27 (4), 659–673.CrossRef

Clarke, D.G. (1976). Econometric measurement of the duration of advertising effect on sales. Journal of Marketing Research, 13(4), 345–357.CrossRef

Danaher, P.J. (2008). Advertising models. Handbook of Marketing Decision Models.

Doganoglu, T., & Klapper, D. (2006). Goodwill and dynamic advertising strategies. Quantitative Marketing and Economics, 4(1), 5–29.CrossRef

Dubé, J.-P.H., Hitsch, G.J., & Manchanda, P. (2005). An empirical model of advertising dynamics. Quantitative Marketing and Economics.

Eastlack, J.O. Jr, & Rao, A.G. (1989). Advertising experiments at the campbell soup company. Marketing Science, 8(1), 57–71.CrossRef

Ebbinghaus, H. (1913). Memory: A contribution to experimental psychology (translated by henry a. ruger and clara e. bussenius; original german work published 1885). New York: Teachers College, Columbia University.

Erdem, T., Keane, M.P., & Sun, B. (2008). A dynamic model of brand choice when price and advertising signal product quality. Marketing Science, 27(6), 1111–1129.CrossRef

Eskin, G., & Baron, P. (1977). Effects of price and advertising in test-market experiments. Journal of Marketing Research, XIV, 499–508.CrossRef

Feichtinger, G., Hartl, R.F., & Sethi, S.P. (1994). Dynamic optimal control models in advertising: Recent developments. Management Science, 40, 195–226.CrossRef

Feinberg, F.M. (2001). On continuous-time optimal advertising under s-shaped response. Management Science, 47(11), 1476–1487.CrossRef

Ghose, A., & Yang, S. (2009). An empirical analysis of search engine advertising: Sponsored search in electronic markets. Management Science, 55(10), 1605–1622.CrossRef

Ghose, A., & Yang, S. (2010). Modeling cross-category purchases in sponsored search advertising: SSRN eLibrary.

Goldfarb, A., & Tucker, C. (2010). Online display advertising: Targeting and instrusiveness.

Goldfarb, A., & Tucker, C. (2011). Search engine advertising: Channel substitution when pricing ads to context. Management Science, 57(3), 458–470.CrossRef

Hahn, M., & Hyun, J.-S. (1991). Advertising cost interactions and the optimality of pulsing. Management Science, 37, 157–169.CrossRef

Hanssens, D., Parsons, L.J., & Schultz, R.L. (2001). Market response models: Econometric and time series analysis, 2nd edn. Kluwer Academic Publishers.

Heflin, D.T.A., & Haygood, R.C. (1985). Effects of scheduling on retention of advertising messages. Journal of Advertising, 14(2), 41–64.CrossRef

Janiszewski, C., Noel, H., & Sawyer, A.G. (2003). A meta-analysis of the spacing effect in verbal learning: Implications for research on advertising repetition and consumer memory. Journal of Consumer Research, 30(1), 138–149.CrossRef

Jerath, K., Ma, L., Park, Y.-H., & Srinivasan, K. (2011). A “position paradox” in sponsored search auctions. Journal of Marketing Research, 30(4), 612–627.

Katona, Z., & Sarvary, M. (2010). The race for sponsored links: Bidding patterns for search advertising. Marketing Science, 29(2), 199–215.CrossRef

Krishnamurthi, L., Narayan, J., & Raj, S.P. (1986). Intervention analysis of a field experiment to assess the buildup effect of advertising. Journal of Marketing Research, 23(4), 337–345.CrossRef

Lewis, R., & Reiley, D. (2014). Replace Online ads and offline sales: measuring the effect of retail advertising via a controlled experiment on Yahoo!. Quantitative Marketing and Economics, 12(3), 235–266.CrossRef

Lodish, L.M., Abraham, M., Livelsberger, J., Lubetkin, B., Richardson, B., & Stevens, M.E. (1995a). A summary of fifty-five in-market experimental estimates of the long-term effect of tv advertising. Marketing Science, 14(3), G133–G140.CrossRef

Lodish, L.M., Abraham, M., Kalmenson, S., Livelsberger, J., Lubetkin, B., Richardson, B., & Stevens, M.E. (1995b). How t.v. advertising works: A meta-analysis of 389 real world split cable t.v. advertising experiments. Journal of Marketing Research, 32(2), 125–139. ISSN 00222437.CrossRef

Mahajan, V., & Muller, E. (1986). Advertising pulsing policies for generating awareness for new products. Marketing Science, 5(2), 89–106.CrossRef

Naik, P.A., Mantrala, M.K., & Sawyer, A.G. (1998). Planning media schedules in the presence of dynamic advertising quality. Marketing Science, 17, 214–235. ISSN 1526-548X.CrossRef

Narayanan, S., & Kalyanam, K. (2015). Position effects in search advertising and their moderators: A regression discontinuity approach. Marketing Science, 34(3), 388–407.CrossRef

Nerlove, M., & Arrow, K.J. (1962). Optimal advertising policy under dynamic conditions. Economica, 29(114), 129–142.CrossRef

Noel, H., & Vallen, B. (2009). The spacing effect in marketing: A review of extant findings and directions for future research. Psychology and Marketing, 26(11), 951–969.CrossRef

Pavlik, P.I., & Anderson, J.R. (2003). An ACT-R model of the spacing effect. In Proceedings of the Fifth International Conference on Cognitive Modeling (pp. 177–182).

Pavlik, P.I., & Anderson, J.R. (2005). Practice and forgetting effects on vocabulary memory: An activation-based model of the spacing effect. Cognitive Science, 29, 559–586.CrossRef

Pavlik, P.I., & Anderson, J.R. (2008). Using a model to compute the optimal schedule of practice. Journal of Experimental Psychology, 14(2), 101–117.

Rutz, O., & Trusov, M. (2011). Zooming in on paid search ads - a consumer-level model calibrated on aggregated data. Marketing Science, 30(5), 789–800.CrossRef

Rutz, O.J., & Bucklin, R.E. (2011). From generic to branded: A model of spillover in paid search advertising. Journal of Marketing Research, 48(1), 87–102.CrossRef

Rutz, O.J., Trusov, M., & Bucklin, R.E. (2011). Modeling indirect effects of paid search advertising: Which keywords lead to more future visits? Marketing Science, 30(4), 646–665.CrossRef

Sahni, N.S. (2014). Advertising spillovers: Field experiment evidence and implications for returns from advertising. Working Paper.

Sawyer, A.G., & Ward, S. (1979). Carry-over effects in advertising communication. Research in Marketing, 2, 259–314.

Sethuraman, R., Tellis, G.J., & Briesch, R.A. (2011). How well does advertising work? generalizations from meta-analysis of brand advertising elasticities. Journal of Marketing Research, 48(3).

Simester, D., Hu, Y., Brynjolfsson, E., & Anderson, E. (2009). Dynamics of retail advertising: Evidence from a field experiment. Economic Enquiry, 47(3), 482–499.CrossRef

Singh, S.N., Mishra, S., Bendapudi, N., & Linville, D. (1994). Enhancing memory of television commercials through message spacing. Journal of Marketing Research, 31, 384–392.CrossRef

Tellis, G.J. (2009). Generalizations about advertising, effectiveness in markets. Journal of Advertising Research, 49(2), 240–245.CrossRef

Terui, N., Ban, M., & Allenby, G.M. (2011). The effect of media advertising on brand consideration and choice. Marketing Science, 30(1), 74–91.CrossRef

Varian, H.R. (2007). Position auctions. International Journal of Industrial Organization, 25.

Villas-Boas, J.M. (1993). Predicting advertising pulsing policies in an oligopoly: A model and empirical test. Marketing Science, 12(1), 88–102.CrossRef

Villas-Boas, J.M., & Winer, R.S. (1999). Endogeneity in brand choice models. Management Science, 45(10), 1324–1338.CrossRef

Winer, R.S. (1993). Using single-source data as a natural experiment for evaluating advertising effects. Journal of Marketing Science, 2(2), 15–31.

Yao, S., & Mela, C.F. (2011). A dynamic model of sponsored search advertising. Marketing Science, 30(3), 447–468.CrossRef

Titel: Effect of temporal spacing between advertising exposures: Evidence from online field experiments
verfasst von: Navdeep S. Sahni
Publikationsdatum: 01.09.2015
Verlag: Springer US
Erschienen in: Quantitative Marketing and Economics / Ausgabe 3/2015
Print ISSN: 1570-7156
Elektronische ISSN: 1573-711X
DOI: https://doi.org/10.1007/s11129-015-9159-9