Incremental Clicks Impact Of Mobile Search Advertising Shaun Lysen Google Inc.

Abstract In this research, we examine how the number of mobile organic clicks changes when advertisers significantly change their mobile ad spend. This continues the line of research of search ads pause by applying it to the mobile platform. We utilize a statistical model to estimate the fraction of clicks that can be attributed to mobile search advertising. A metastudy of 327 advertisers reveals that 88% of ad clicks are incremental, in the sense that the visits to the advertiser’s site would not have occurred without the mobile ad campaigns.

1

Introduction Mobile advertising is largely driven by smartphones. The use of smartphones in the United States is on the rise with 56% penetration as of the first quarter of 2013. Of those smartphone owners, 61% perform searches on their smartphones every day while 77% have researched a product or service online. Smartphones also drive users who begin research on their phones to make purchases through other channels with 40% purchasing via a computer and 38% purchasing offline in a store [3].

In two previously published papers, we established that 89% of visits to an advertiser’s site are incremental to clicks on organic results on average [1], and when broken down by organic result rank, 50% are incremental when an organic result appears at the top position, 82% when an organic result appears at ranks 2-4, and 95% when an organic result appears at ranks 5 and higher [2]. This paper continues this line of research by focusing exclusively on the mobile platform – only considering ad spend, ad clicks, organic clicks and organic impressions for mobile. The two prior studies combined data across the desktop, mobile and tablet platforms. As the layout of the search results page on mobile differs from desktop, we might expect the incrementality of ad clicks to vary on the mobile platform.

US mobile online advertising spend is one of the fastest growing areas with a projected spend of $7.7 billion in 2013 growing to an estimated $28 billion in 2017 [4]. There are advantages search advertising has over traditional media advertis-

1

2.1

Implementation Details

2

METHODOLOGY

ing. One involves access to direct metrics of impact, such as the number of clicks achieved. Another is search advertising allows advertisers to pay only when a user clicks on an ad. And yet another is that since the ads are triggered by search terms, they tend to be highly relevant to the user. Additionally, search advertising on the mobile platform allows advertisers to reach users when they are out of their homes. 83% of smartphone users do not leave home without their device [3].

are paused. Conversely, when the campaign is restarted, the IAC represents the fraction of paid clicks that are incremental. Since we do not assume a positive interaction between paid and organic search in our analysis, the IAC estimate is bounded at 100%. For example, consider the following scenario:

However, measuring the number of mobile ad clicks alone does not provide information on the incrementality of mobile search advertising. That is, the question “how many of the clicks are incremental to clicks that would have occurred on natural search results in the absence of paid ad results?” is not answered. Advertisers that pause their mobile search advertising campaigns sometimes cite concerns about how much of the traffic to the sites is truly incremental to clicks on natural search results.

(B) Subsequently, they cut their ad spend to $0 and find there are 500 organic clicks a month and 0 paid clicks a month. Under (A), there are 200 incremental clicks, thereby giving an IAC of (700-500)/(300-0) = 66.7%.

2

ceding the ad spend change, a meta-analysis of all the Mobile Search Ad Pause studies provides insight into the average IAC from mobile search advertising.

(A) An advertiser spends $1,000 a month and receives 400 organic and 300 paid clicks a month.

In the above example, we do not consider external factors which could also affect the organic clicks before and after the spend change. To control for this, we employ the statistical model deThe incrementality is dependent on factors such scribed in [1]. as the organic search result ranking and how This estimate of 200 incremental clicks depends similar the paid and organic listings are to each on factors leading to the ad spend drop and the other. By measuring the incremental click im- state of the account and competitive environpact from search advertising, the advertiser is ment around the time of the spend change. Alable to make more informed decisions regarding though the estimate of the IAC should always their advertising spend. be considered in the context of the changes pre-

Methodology

In order to determine the incremental clicks related to mobile search advertising, we quantify the impact that a pronounced search ad spend change (increase or decrease) has on total mobile clicks. Indirect navigation to the advertiser site is not considered. Each study produces an estimate of the incremental clicks attributed to search advertising for an advertiser. To make comparison across multiple studies easier, we express the incremental clicks as a percentage of the change in paid clicks. This metric is labeled “Incremental Ad Clicks”, or “IAC” for short.

2.1

Implementation Details

The studies are implemented by leveraging the automated pipeline from [1] and customizing it to only extract spend, clicks and impressions for the mobile platform. The pipeline extracted ad and organic data between March 2012 and April 2013 for each company that was identified to have either a significant increase or decrease in spend. Also identified are a pre-period (a relatively stable period prior to the spend change), IAC represents the percentage of paid clicks that and a post-period (a relatively stable period afare not made up for by organic clicks when ads ter the spend change). The results are compared 2

Google Inc.

3

META-ANALYSIS RESULTS

3.1

against validation checks on data integrity and model quality. Validation checks are used to ensure confidence in the statistical model and its predictions. Only studies passing the validation flags are included in this meta-analysis. For more details on the statistical model, see [1].

3

Meta-Analysis Results

The meta-analysis is based on 327 validated studies conducted between March, 2012 to April, 2013.

IAC Statistics by Vertical

Industry Vertical Retail Education & Government Technology Finance Services All Verticals Consumer Packaged Goods Automotive Business & Industrial Markets Healthcare Media & Entertainment Travel Classifieds & Local

N 79 30 28 26 20 19

Mean 86.3% 93.6% 89.7% 87.3% 82.4% 85.7%

Sd 17.6% 11.7% 19.1% 16.6% 19.4% 17.1%

Median 93.9% 98.4% 97.0% 96.2% 91.7% 90.8%

18 18

85.6% 94.0%

15.9% 5.5%

88.3% 95.1%

14 14 12 9

82.8% 86.4% 84.8% 96.9%

12.2% 23.8% 22.5% 4.3%

87.0% 96.7% 92.0% 98.7%

Table 1: IAC Statistics by Industry Vertical Figure 1 is a histogram plot of the IAC across all 327 studies. The average IAC across all studies is Figure 2 is a boxplot of the IAC for each of 12 87.7%, with the median rate at 96%. More than different verticals arranged alphabetically. The 63% of the studies had an IAC value greater 90%, IAC is quite consistent across verticals with most with only a few studies showing a low IAC value. verticals having a median IAC above 90%. Incrementality of Ad Clicks by Vertical Histogram of Incremental Ad Click Percentage

100

● ● ● ●

150

75







50

● ● ●

● ● ●

100



25

● ●

● ●



50

40

60

80

3.1

IAC Statistics by Vertical

Table 1 and Figure 2 summarize the IAC statistics by industry vertical. We have omitted industry verticals with fewer than 5 studies from the table and boxplot. Google Inc.

Travel

Classifieds & Local

Media & Entertainment

Healthcare

Vertical

Percent Incremental Ad Clicks

Figure 1: Histogram of Incremental Ad Clicks

Automotive

Finance

Technology

100

Business & Industrial Markets

20

Consumer Packaged Goods

0

Services All Verticals

0

Retail

0 Education & Government

Frequency

Incrementality



Figure 2: Boxplot of Incremental Ad Clicks by Vertical Table 2 and Figure 3 summarize the IAC by whether the spend change was an increase or decrease. Spend change Increase Decrease

N 89 235

Mean 87.0% 88.7%

Sd 17.7% 16.0%

Median 95.4% 96.1%

Table 2: IAC Statistics by Spend Change Type 3

REFERENCES

REFERENCES

mate. Being able to track these factors for each study will allow us to better understand their influence on the IAC estimate.

Incrementality of Ad Clicks by Spend Change Type 100

75

Incrementality

Acknowledgments Spend Change Type

● ●

50

Decrease

● ● ● ● ● ● ●

Increase

● ● ●

25

● ● ● ●



Increase

Decrease

0

Spend Change Type

Figure 3: Boxplot of Incrementality by Ad Spend Change Type

4

Concluding Remarks

We have examined those accounts which have exhibited a pronounced spend change for which our models produce valid results. The metaanalysis is not representative of all the possible factors which could drive ad spend change. However, given the large volume of studies produced, across multiple industry verticals, our analysis does provide a reasonable cross section of expected IAC. A more rigorous approach to determining IAC would be to conduct a randomized experiment. A test group would be exposed to the pull back in paid search ads while search spend would be held constant in a control group. A comparison of the paid and organic click volumes in the two groups would then yield an IAC estimate. One approach would be to conduct a geo-experiment [5]. Ultimately, advertisers are interested in how much income can be attributed to their search advertising campaigns. Our analysis does not include an estimate for incremental conversions. Other factors such as the ranking of the organic search result or the strength of brand awareness of the search term could influence the IAC esti4

We would like to thank David Chan for all of his insightful discussion with the existing search pause infrastructure and Jeffrey Ai for all of his help in extracting mobile data. We would also like to thank Tony Fagan and Penny Chu for their encouragement and support and a big thanks to Jim Koehler for his constructive feedback. Thank you also to the many others at Google who made this paper possible.

References [1] D. Chan, Y. Yuan, J. Koehler, D. Kumar Incremental Clicks Impact of Search Advertising Journal of Advertising Research, 51(4):58-62, 2011 [2] D. Chan, D. Kumar, S. Ma, J. Koehler, Impact of Ranking of Organic Search Results on the Incrementality of Search Ads www.google.com/think/researchstudies/impact-of-ranking-of-organicsearch-results-on-the-incrementality-ofsearch-ads.html [3] Google Think Insights Our Mobile Planet: United States, Understanding the Mobile Consumer http://www.google.com/think/researchstudies/our-mobile-planet-unitedstates.html [4] eMarketer US Mobile Advertising Spending & Growth. eMarketer, June 2013. [5] J. Vaver, J. Koehler Measuring Ad Effectiveness Using Geo Experiments http://googleresearch.blogspot.com/2011/12/measuringad-effectiveness-using-geo.html

Google Inc.

Incremental Clicks Impact Of Mobile Search ... - Research at Google

[2]. This paper continues this line of research by focusing exclusively on the .... Figure 2: Boxplot of Incremental Ad Clicks by ... ad-effectiveness-using-geo.html.

256KB Sizes 3 Downloads 471 Views

Recommend Documents

Incremental Clicks Impact Of Search Advertising - Research at Google
Google Inc. Abstract. In this research ... search advertising has over traditional media ad- vertising. ... across multiple studies easier, we express the in- cremental ...

Impact Of Ranking Of Organic Search Results ... - Research at Google
Mar 19, 2012 - average, 50% of the ad clicks that occur with a top rank organic result are ... In effect, they ... Below is an illustration of a search results page.

Incremental Crawling - Research at Google
Part of the success of the World Wide Web arises from its lack of central control, because ... typically handled by creating a central repository of web pages that is ...

Query Suggestions for Mobile Search ... - Research at Google
Apr 10, 2008 - suggestions in order to provide UI guidelines for mobile text prediction ... If the user mis-entered a query, the application would display an error ..... Hart, S.G., Staveland, L.E. Development of NASA-TLX Results of empirical and ...

TechWare: Mobile Media Search Resources - AT&T Labs Research
weka): A collection of machine learn- ing algorithms in Java for data mining tasks. It contains data preprocessing, classification, regression, clustering, association rules, and visualization. □ LIBSVM (www.csie.ntu.edu. tw/~cjlin/libsvm): A libra

The Incremental Reach and Cost Efficiency of ... - Research at Google
ads and whether a cross-media campaign would be more cost-efficient for reaching their target ... media campaign would have delivered at least the reach of the TV campaign. ...... [3] US TV Ad Spending(2009-2015), eMarketer, March 2011.

Incorporating Clicks, Attention and Satisfaction ... - Research at Google
the model predict click vs. satisfaction events? RQ2 Does an offline ... ing a SERP item [26] and even the satisfaction reported by the user [27], based on mouse ...

Data Enrichment for Incremental Reach Estimation - Research at Google
small data set and B for big data set. Our methods for improvement are based on the decomposition θ = Pr(Z = 1) Pr(Y = 0 | Z = 1). Under the IDA, we can use the ...

Large-scale Incremental Processing Using ... - Research at Google
language (currently C++) and mix calls to the Percola- tor API with .... 23 return true;. 24. } 25. } 26 // Prewrite tries to lock cell w, returning false in case of conflict. 27 ..... set of the servers in a Google data center. .... per hour. At thi

Google Search by Voice - Research at Google
May 2, 2011 - 1.5. 6.2. 64. 1.8. 4.6. 256. 3.0. 4.6. CompressedArray. 8. 2.3. 5.0. 64. 5.6. 3.2. 256 16.4. 3.1 .... app phones (Android, iPhone) do high quality.

Google Search by Voice - Research at Google
Feb 3, 2012 - 02/03/2012 Ciprian Chelba et al., Voice Search Language Modeling – p. 1 ..... app phones (Android, iPhone) do high quality speech capture.

On the Impact of Kernel Approximation on ... - Research at Google
termine the degree of approximation that can be tolerated in the estimation of the kernel matrix. Our analysis is general and applies to arbitrary approximations of ...

Query-Free News Search - Research at Google
Keywords. Web information retrieval, query-free search ..... algorithm would be able to achieve 100% relative recall. ..... Domain-specific keyphrase extraction. In.

Google Search by Voice - Research at Google
Kim et al., “Recent advances in broadcast news transcription,” in IEEE. Workshop on Automatic ... M-phones (including back-off) in an N-best list .... Technology.

Voice Search for Development - Research at Google
26-30 September 2010, Makuhari, Chiba, Japan. INTERSPEECH ... phone calls are famously inexpensive, but this is not true in most developing countries.).

The Impact of Memory Subsystem Resource ... - Research at Google
sources on five Google datacenter applications: a web search engine, bigtable ... havior and discover that the best mapping for a given ap- plication changes ... Machines have multiple sockets hosting proces- sors with multiple ... 9 10. 3 MB L2. 11.

Do Viewers Care? Understanding the impact of ... - Research at Google
TV advertising, and that these reactions can then be used to rank creatives by their apparent relevance to the viewing audience. 1 Why viewers tune away.

On the Difficulty of Nearest Neighbor Search - Research at Google
plexity to find the nearest neighbor (with a high prob- ability)? These questions .... σ is usually very small for high dimensional data, e.g., much smaller than 0.1).

Topical Clustering of Search Results - Research at Google
Feb 12, 2012 - that the last theme is easily identifiable even though the last three ..... It goes without saying that we have to add the cost of annotating the short ...

Retroactive Answering of Search Queries - Research at Google
May 23, 2006 - [email protected]. ABSTRACT. Major search engines currently use the history of a user's actions ... H.3.4 [Information Systems]: Information Storage and ...... users' standing interests and needs, and existing technology intended to ...

Learning Battery Consumption of Mobile Devices - Research at Google
Social networking, media streaming, text mes- saging, and ... chine Learning, New York, NY, USA, 2016. ... predict the battery consumption of apps used by ac-.

On the Predictability of Search Trends - Research at Google
Aug 17, 2009 - various business decisions such as budget planning, marketing ..... the major characteristics of the underlying time series are maintained.