- Email: [email protected]
A Method for the Construction of User Targeting Knowledge for B2B Industry Website
Available online at www.sciencedirect.com
ScienceDirect Procedia Computer Science 96 (2016) 1147 – 1155
20th International Conference on Knowledge Based and Intelligent Information and Engineering Systems
A Method for the Construction of User Targeting Knowledge for B2B Industry Website Takumi Ozawaa*, Akiyuki Sekiguchia, Kazuhiko Tsudaa a
Graduate School of Business Sciences, University of Tsukuba, 3-29-1, Otsuka, Bunkyo-ku, Tokyo 112-0012, Japan
Abstract To improve own software and service awareness in enterprise B2B Software industry often have the marketing event. Improving the number of the event attendees is equal to improving the software and service awareness. Therefore, it is required to develop the effective web site to promote the event and secure the registration number of event attendance, also asked for cost reduction at the same time from ROI (Return on Investment) perspective. It requires to show the right contents to the right users because each user have several different motivation which is according to their business scheme, business model, and own role and responsibilities. In this study, we could figure out user targeting knowledge through the A/B testing, such as wording effectiveness between the several industries to make the website visitors register the event registration. Also it showed several results of A/B testing to figure out effective schemes and knowledges of user targeting for B2B industry website. © by Elsevier B.V. This is an open access article under the CC BY-NC-ND license ©2016 2015Published The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of KES International. Peer-review under responsibility of KES International Keywords: B2B industry, Marketing event, Website, Web analysis, A/B testing, User Targeting,
1. INTRODUCTION To improve own products and service awareness in enterprise B2B (Business to Business) software company such as Adobe Systems, Microsoft, Oracle, IBM, etc. generally have the marketing event. Improving the number of event attendees is equal to improving the software and service awareness. The sales representatives invite enterprise customers at the onsite meeting, but for small or mid-sized customers it does not work with the same approach. The strategy of assigning sales representatives which works for the enterprise customers is not realistically for small and
* Corresponding author. Tel. +81-80-4173-0701 E-mail address: [email protected]
1877-0509 © 2016 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of KES International doi:10.1016/j.procs.2016.08.157
1148
Takumi Ozawa et al. / Procedia Computer Science 96 (2016) 1147 – 1155
medium-sized customers from the resource and cost perspectives of the sales. There are many opportunities for us to generate revenue from the small and medium-sized customers, and it must be considered for further business expansion. Therefore most of the enterprise B2B software companies are required to develop the effective web site for improving event awareness and number of customer registration through the website with efficiency and cost savings. In figure 1 it shows scheme and positioning of B2B software event. To develop the effective website in order to achieve its purpose, it should deliver the right contents to the right visitors because they have different motivation according to their business scheme, business model, and own role and responsibilities. It should figure out what kind of the customers / industries visit to the event website and what kind of contents motivate them to register B2B event attendance.
Fig. 1. Scheme of B2B Software Event
2. ABOUT USER TARGETING & A/B TESTING IN B2B INDUSTRY 2.1. Purpose It requires to show the right contents to the right visitors on the B2B event registration website because each visitor have different motivation according to their business scheme, business model, business requirement, and role and responsibilities of visitors. It is possible to figure out several information through the connecting between visitor’s IP address and customer database which has company name, address, phone number, industry, number of employee, revenue, business model, business scheme, etc. When the visitors visit the website, user targeting system shows the personalized contents using the data from the customer database in milliseconds. We can expect significant of the business impact by running the A/B testing at the most effective page and/or content with user targeting, and figure out what the components and contents of the website fit to the purpose of the visitor’s motivation at low risk. The risks are that company do not invest unnecessary resources and make the visitors have wrong understanding of products and services. It is important to avoid those risks through the user targeting and A/B testing.
Takumi Ozawa et al. / Procedia Computer Science 96 (2016) 1147 – 1155
1149
2.2. Previous Research There were studies about B2B web analytics and A/B testing. Virtsonis [1] conducted a study about B2B web site from branding perspectives. Furthermore, Wilson [2] used clickstream data to analyze B2B web site performance and concluded that web analytics can be applied in a B2B domain. In addition, Sekiguchi et al [3] studied for B2B web analytics for such a critical web conversion metric as web form registration and factors of web user registration for own website. Sekiguchi et al [4] studied for user segmentation knowledge of B2B manufacture website. Also there were general studies about web analysis, web optimization and personalization [5] - [16]. However, there was no study done for user targeting and A/B testing for improving event awareness and attendance registration in B2B industry.
3. APPROACH OF USER TARGETING & A/B TESTING 3.1. Theme of User Targeting & A/B testing It is necessary for the enterprise B2B Software Company to make the web visitors have strong motivation for improving number of the event registration, also need to deliver right communication with right wording at the event web page. We executed two type of communication, such as A/B testing of a) comparing effectiveness of main message between concrete wording (Specific date of the event) and/versus abstract wording (Time pressure of registration deadline), and also user targeting of b) comparing effectiveness of main message between 3 types of industry (Retail & Distribution / Media & Entertainment / Automotive & Financial Service). Before executing A/B Testing and User Targeting, we found several studies and researches related to a), Rik Pieters et al [18] studied the impact of time pressure and task motivation, Ravi Dhar and Stephen M. Nowlis [19] showed the effect of time pressure on consumer choice deferral, and Ola Svenson et al [20] studied the effectiveness of time pressure and stress in human judgment and decision making. Also CF Köhler et al [17] showed the comparison of the consumer acceptance of recommendations with concrete communication versus abstract communications. There are several patent document that includes user targeting delivery system and methods with user behaviors and/or IP addresses [18-20], but it could not find any studies and researches related to b) user targeting by specific industries in the enterprise B2B software industry. 3.2. Plan for User Targeting & A/B testing We planned two types of marketing creatives, a-1) solicitation of specific name / date / place of the event, (concrete wording “Adobe Digital Marketing Symposium, 12th April, Grand Hyatt Tokyo”) and a-2) solicitation of registration time pressure (abstract wording “Hurry! Register before fully booked!”). We executed the A/B testing with those creatives to comparing effectiveness of event registration from 15 th April to 30th April, 2014 at the top page of the event website where the biggest number of visits and visitors had in the website. After the A/B testing of a), we set the user targeting with using a system and method for selectively acquiring and targeting web contents based on users' Internet Protocol (IP) addresses which shows in Figure 2. Websites are made aware of IP addresses of interest, determined by matching attributes of current internet users for whom IP addresses are known and targeting attributes of online marketing. The websites choose whether to show the contents for each visitor from one of said IP addresses in the immediate demand. In figure 2, this scheme also provides for targeting of online contents based on updated user IP addresses and some data provided by customer database which has company name, address, phone number, industry, number of employee, revenue, business model, business scheme, etc. According to the targeting scheme we executed three types of user targeting contents which related to three different industries, such as b-1) Retail & Distribution (targeting wording “Event speakers from Rakuten Inc, and DeNA Co., Ltd.” which are one of the most famous retail & distribution company in Japan), b-2) Media & Technology (targeting wording “Event speakers from Yahoo!JAPAN and Panasonic”) and b-3) Automotive & Financial Services (targeting wording “Event speakers from Nissan Motors and Mitsui Sumitomo Credit Card”) with evaluating effectiveness of event registration comparing to default wording (wording “Adobe Digital Marketing Symposium, 12 th April, Grand Hyatt Tokyo”).
1150
Takumi Ozawa et al. / Procedia Computer Science 96 (2016) 1147 – 1155
In Table 1 & 2, it shows all experiences of those user targeting and A/B testing.
Table 1. Experiences of A/B Testing.
A/B Testing Experiences : a) Concrete versus Abstract wording a-1) Default (Concrete wording) “Adobe Digital Marketing Symposium, 12th April, Solicitation of specific name / date / place of the event Grand Hyatt Tokyo” “Hurry! Register before fully booked!”
a-2) Abstract wording Solicitation of registration time pressure
Table 2. Experiences of User Targeting.
User Targeting Experiences : b) User Targeting with 3 industries b-1) Default “Adobe Digital Marketing Symposium, 12th April, Solicitation of name / date / place of the event Grand Hyatt Tokyo” b-2) Retail & Distribution Solicitation of company name of event speakers
“Event speakers from Rakuten Inc, and DeNA Co., Ltd.”
b-3) Media & Technology Solicitation of company name of event speakers
“Event speakers Panasonic”
b-4) Automotive & Financial Services Solicitation of company name of event speakers
“Event speakers from Nissan Motors and Mitsui Sumitomo Credit Card”
from
Yahoo!JAPAN
and
Fig. 2. Scheme of User Targeting.
In the user targeting and A/B testing, same experience was displayed as long as visitors did not delete the cookies of the web browser, and the test experience to be displayed randomly and equality for first time visitors.
1151
Takumi Ozawa et al. / Procedia Computer Science 96 (2016) 1147 – 1155
4. RESULT OF USER TARGETING & A/B TESTING 4.1. Result of A/B testing It is the results of user targeting and A/B testing in Table 3. The conversion rate is calculated "Number of Success Registrations divides by Visits". The test result was Experience a-2) that Abstract wording ĀHurry! Register before fully booked!” has become the largest value in the conversion rate. From the previous researches the time pressure effects user behaviors and decision making. This A/B testing showed positive effectiveness of event registration, but this is one of the abstract wording. Therefore we need to continue A/B testing with other candidates of abstract wording, and figure out what experience is more effective than others. Table 3. Result of A/B Testing: Concrete versus Abstract wording.
Experiences
Visits
Success Registrations
Conversion Rate
Total
2,628
310
11.8%
a-1) Default (Concrete Wording) a-2) Abstract Wording
1,321
129
9.8%
1,307
181
13.8%
4.2. Result of User Targeting It set three types of user targeting with these industries, Retail & Distribution, Media & Technology, and Automotive & Financial Services. It could be defined the industry of each visitors with scheme of user targeting in Figure 2. If the visitor’s industry was Retail or/and Distribution, user targeting scheme showed targeted main message banner which said “Event speakers from Rakuten Inc, and DeNA Co., Ltd.”. Then we compared by A/B Testing for clarifying effectiveness of user targeting. In Table 4, the test of b-2) user targeting of Retail & Distributions, there was significant difference between Experience Default which is solicitation of name / date / place of the event (“Adobe Digital Marketing Symposium, 12th April, Grand Hyatt Tokyo”) and Experience b-2) Retail & Distribution (“Event speakers from Rakuten Inc, and DeNA Co., Ltd.”). Also in figure 3, it could be found that Experiences b-2) became a winner consistently during testing period. Both of result were good enough to identify the winner therefore it could be said that user targeting with IP address. We need to continue the user targeting with other wording for this industry to have more clear findings. Table 4. Result of User Targeting: Retail & Distribution.
Experiences
Visits
Success Registrations
Conversion Rate
Total
742
179
-
b-1) Default
366
68
18.6%
b-2) Retail & Distribution
376
111
29.5%
1152
Takumi Ozawa et al. / Procedia Computer Science 96 (2016) 1147 – 1155
Fig. 3. Result of User Targeting: Retail & Distribution.
Also if the visitor’s industry was Media or/and Technology, user targeting scheme showed targeted main message banner which said “Event speakers from Yahoo!JAPAN and Panasonic”. In Table 5, the test of b-3) user targeting of Retail & Distributions, there is difference between Experience Default which is solicitation of name / date / place of the event (“Adobe Digital Marketing Symposium, 12th April, Grand Hyatt Tokyo”) and Experience b-3) Retail & Distribution (“Event speakers from Yahoo!JAPAN and Panasonic”). Also in figure 4, it could be found that Experiences b-3) became a winner consistently during testing period. Both of result were good enough to identify the winner therefore it could be said that user targeting with IP address. Table 5. Result of User Targeting: Media & Technology
Experiences Total
Visits
Success Registrations
Conversion Rate
1441
319
-
b-1) Default
714
98
13.7%
b-3) Media & Technology
727
221
30.4%
1153
Takumi Ozawa et al. / Procedia Computer Science 96 (2016) 1147 – 1155
Fig. 4. Result of User Targeting: Automotive & Distribution.
If the visitor’s industry was Automotive or/and Financial Services, user targeting scheme showed targeted main message banner which said “Event speakers from Nissan Motors and Mitsui Sumitomo Credit Card”. In Table 6, the test of b-4) user targeting of Retail & Distributions, there is difference between Experience Default which is solicitation of name / date / place of the event (“Adobe Digital Marketing Symposium, 12th April, Grand Hyatt Tokyo”) and Experience b-4) Retail & Distribution (“Event speakers from Nissan Motors and Mitsui Sumitomo Credit Card”). Also in figure 5, it could be found that Experiences b-4) became a winner consistently during testing period. Both of result were good enough to identify the winner therefore it could be said that user targeting with IP address. Table 6. Result of User Targeting: Automotive & Financial Services.
Experiences
Visits
Success Registrations
Conversion Rate
Total
617
126
-
b-1) Default
301
43
14.3%
b-4) Auto & Fin
316
83
26.3%
1154
Takumi Ozawa et al. / Procedia Computer Science 96 (2016) 1147 – 1155
Fig. 5. Result of User Targeting: Automotive & Distribution.
5. CONCLUSIONS To improve own products and services awareness in the B2B business industry is generally often used by way of the marketing event. Improving the number of event attendees is equal to improving the products and services awareness with low cost investment. Therefore, many of the B2B companies are required to develop the effective web site for improving event awareness and number of customer registration from the cost reduction perspective. To develop the effective web site in order to achieve its purpose, it should deliver the right contents to the right visitors because they had different motivation according to their business scheme, business model, and own role and responsibilities. It means that it should figure out what kind of the customers / industries came to the website and what kind of contents motivated them to register B2B event attendance through the not only web analysis, but also the A/B testing with user targeting. We could expect significant of the business impact by running the user targeting and A/B testing at the web page which had biggest number of visits, and figure out what the experiences it should show to the targeted customers. In addition it could be found that there are differences of visitor behaviour by wordings of the main message between concrete wording and abstract wording through the A / B testing. It showed that abstract wording with time pressure of event registration deadline was effective more than abstract wording with the name / date / place of the event. Also it was executed three type of user targeting with these experiences of industries, b-1) Retail & Distribution, b-2) Media & Technology, b-3) Automotive & Financial Services using IP address and targeting scheme. Through the result of those studies we understood that there are differences in the number of conversion rate of success registration. However we did not able to fully understand the factors of those influences for website visitors because it needed to figure out the effectiveness of other wording and segment of the user targeting. We will continue to find the effectiveness of the wording and user targeting with other experiences. Acknowledgements We thank Adobe Systems Incorporated, Japan for providing the data and test results of the B2B event website.
Takumi Ozawa et al. / Procedia Computer Science 96 (2016) 1147 – 1155
1155
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.
Virtsonis, Sally Harridge-March, 2009. Brand positioning in the B2B online environment: A case from the UK print industry, s.l.: Journal of Brand Management (16), p.556–570. Wilson, R. Dale, 2010. Using clickstream data to enhance businessϋtoϋbusiness web site performance, s.l.: Journal of Business & Industrial Marketing Vol. 25 Iss: 3, p.177 - 187. Akiyuki Sekiguchi, Tadao Katsunuma, Itsumi Hokao, Yusuke Yamada, and Kazuhiko Tsuda, “Web Analytics for B to B Marketing in Semiconductor Industry”, International Journal of e-Education, e-Business, e-Management and e-Learning, Vol. 2, No. 5, October 2012 Akiyuki Sekiguchi, Kazuhiko Tsuda, 2014. Study on Web Analytics Utilizing Segmentation Knowledge in Business to Business Manufacturer Site, s.l.: KES. Ejiri T., 2008. Web Analytics and Web Marketing: Access Log Analytics Realized Web Marketing Kaizen Cycle Management systems, s.l.: Journal of Japan Industrial Management Association 18(1), p.8-43. K. Rebecca, P. Justin, P. Graeme, 2012. Ethical considerations and guidelines in web analytics and digital marketing: a retail case study, s.l.: Proceedings of the 6th Australian Institute of Computer Ethics conference 2012, p.5-12. KYODO, 2014. Benesse data leak may affect 20 million customers, including children. The Japan Times, 9 July, http://www.japantimes.co.jp/news/2014/07/09/national/benesse-data-leak-may-affect-20-million-customers-including-children/. Sampath P., 2012. An efficient weighted rule mining for web logs using, s.l.: Advances in Engineering, Science and Management (ICAESM), International Conference, p.432-436. Pascual-Cid V., 2008. An information visualisation system for the understanding of web data, s.l.: Visual Analytics Science and Technology, VAST '08. IEEE Symposium, p.183-184. Nasraoui, O. Soliman, M. Saka, E. Badia, A. Germain, R., 2008. A Web Usage Mining Framework for Mining Evolving User Profiles in Dynamic Web Sites, s.l.: Knowledge and Data Engineering, IEEE Transactions, p.202-215. X. Wang, D. Shen, H. Chen, L. Wedman, 2011. Applying web analytics in a K-12 resource inventory, s.l.: Electronic Library, Vol. 29 Iss: 1, p.20-35. S. Otsuka, M. Toyoda, M. Kitsuregawa, 2003. A Study for Analysis of Web Access Logs with Web Community, s.l.: Information Processing Society of Japan (IPSJ) Database44, p.432-436. Y. Ichikawa, M. Nakamura, Y. Kishimoto, T. Kobayashi, 2012. A Proposal of Extracting Innovative Users with Web Access Log of an ECommerce Site, s.l.: Information Processing Society of Japan, IPSJ SIG Notes 2012-GN-83(2). Aivalis C.J., 2011. Log File Analysis of E-commerce Systems in Rich Internet Web 2.0 , s.l.: Informatics (PCI), 2011 15th Panhellenic Conference, p.222-226. J. Park, K. Jung, Y. Lee, G. Cho, J. Kim, J. Koh, 2009. The Continuous Service Usage Intention in the Web Analytics Services, s.l.: HICSS '09. 42nd Hawaii International Conference, p. 1-7. Xiao-Gang W., 2009. Web mining based on user access patterns for web personalization, s.l.: Computing, Communication, Control, and Management-ISECS International Colloquium, 2009.p. 20-35. Clemens F. Köhlera, Els Breugelmansb & Benedict G. C. Dellaertc, 2014. Consumer Acceptance of Recommendations by Interactive Decision Aids: The Joint Role of Temporal Distance and Concrete Versus Abstract Communications, s.l: Journal of Management Information Systems, Volume 27, p.231-260. Kevin Wueste, Concentric user-targeting delivery system and methods, US20030115318 A1. Christina Yip Chung, Joshua M. Koran, Long-Ji Lin, Hongfeng Yin, Model for generating user profiles in a behavioral targeting system, US 7809740 B2. David A. Gerken, System and method for selectively acquiring and targeting online advertising based on user IP address, US 7376714 B1.
ScienceDirect Procedia Computer Science 96 (2016) 1147 – 1155
20th International Conference on Knowledge Based and Intelligent Information and Engineering Systems
A Method for the Construction of User Targeting Knowledge for B2B Industry Website Takumi Ozawaa*, Akiyuki Sekiguchia, Kazuhiko Tsudaa a
Graduate School of Business Sciences, University of Tsukuba, 3-29-1, Otsuka, Bunkyo-ku, Tokyo 112-0012, Japan
Abstract To improve own software and service awareness in enterprise B2B Software industry often have the marketing event. Improving the number of the event attendees is equal to improving the software and service awareness. Therefore, it is required to develop the effective web site to promote the event and secure the registration number of event attendance, also asked for cost reduction at the same time from ROI (Return on Investment) perspective. It requires to show the right contents to the right users because each user have several different motivation which is according to their business scheme, business model, and own role and responsibilities. In this study, we could figure out user targeting knowledge through the A/B testing, such as wording effectiveness between the several industries to make the website visitors register the event registration. Also it showed several results of A/B testing to figure out effective schemes and knowledges of user targeting for B2B industry website. © by Elsevier B.V. This is an open access article under the CC BY-NC-ND license ©2016 2015Published The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of KES International. Peer-review under responsibility of KES International Keywords: B2B industry, Marketing event, Website, Web analysis, A/B testing, User Targeting,
1. INTRODUCTION To improve own products and service awareness in enterprise B2B (Business to Business) software company such as Adobe Systems, Microsoft, Oracle, IBM, etc. generally have the marketing event. Improving the number of event attendees is equal to improving the software and service awareness. The sales representatives invite enterprise customers at the onsite meeting, but for small or mid-sized customers it does not work with the same approach. The strategy of assigning sales representatives which works for the enterprise customers is not realistically for small and
* Corresponding author. Tel. +81-80-4173-0701 E-mail address: [email protected]
1877-0509 © 2016 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of KES International doi:10.1016/j.procs.2016.08.157
1148
Takumi Ozawa et al. / Procedia Computer Science 96 (2016) 1147 – 1155
medium-sized customers from the resource and cost perspectives of the sales. There are many opportunities for us to generate revenue from the small and medium-sized customers, and it must be considered for further business expansion. Therefore most of the enterprise B2B software companies are required to develop the effective web site for improving event awareness and number of customer registration through the website with efficiency and cost savings. In figure 1 it shows scheme and positioning of B2B software event. To develop the effective website in order to achieve its purpose, it should deliver the right contents to the right visitors because they have different motivation according to their business scheme, business model, and own role and responsibilities. It should figure out what kind of the customers / industries visit to the event website and what kind of contents motivate them to register B2B event attendance.
Fig. 1. Scheme of B2B Software Event
2. ABOUT USER TARGETING & A/B TESTING IN B2B INDUSTRY 2.1. Purpose It requires to show the right contents to the right visitors on the B2B event registration website because each visitor have different motivation according to their business scheme, business model, business requirement, and role and responsibilities of visitors. It is possible to figure out several information through the connecting between visitor’s IP address and customer database which has company name, address, phone number, industry, number of employee, revenue, business model, business scheme, etc. When the visitors visit the website, user targeting system shows the personalized contents using the data from the customer database in milliseconds. We can expect significant of the business impact by running the A/B testing at the most effective page and/or content with user targeting, and figure out what the components and contents of the website fit to the purpose of the visitor’s motivation at low risk. The risks are that company do not invest unnecessary resources and make the visitors have wrong understanding of products and services. It is important to avoid those risks through the user targeting and A/B testing.
Takumi Ozawa et al. / Procedia Computer Science 96 (2016) 1147 – 1155
1149
2.2. Previous Research There were studies about B2B web analytics and A/B testing. Virtsonis [1] conducted a study about B2B web site from branding perspectives. Furthermore, Wilson [2] used clickstream data to analyze B2B web site performance and concluded that web analytics can be applied in a B2B domain. In addition, Sekiguchi et al [3] studied for B2B web analytics for such a critical web conversion metric as web form registration and factors of web user registration for own website. Sekiguchi et al [4] studied for user segmentation knowledge of B2B manufacture website. Also there were general studies about web analysis, web optimization and personalization [5] - [16]. However, there was no study done for user targeting and A/B testing for improving event awareness and attendance registration in B2B industry.
3. APPROACH OF USER TARGETING & A/B TESTING 3.1. Theme of User Targeting & A/B testing It is necessary for the enterprise B2B Software Company to make the web visitors have strong motivation for improving number of the event registration, also need to deliver right communication with right wording at the event web page. We executed two type of communication, such as A/B testing of a) comparing effectiveness of main message between concrete wording (Specific date of the event) and/versus abstract wording (Time pressure of registration deadline), and also user targeting of b) comparing effectiveness of main message between 3 types of industry (Retail & Distribution / Media & Entertainment / Automotive & Financial Service). Before executing A/B Testing and User Targeting, we found several studies and researches related to a), Rik Pieters et al [18] studied the impact of time pressure and task motivation, Ravi Dhar and Stephen M. Nowlis [19] showed the effect of time pressure on consumer choice deferral, and Ola Svenson et al [20] studied the effectiveness of time pressure and stress in human judgment and decision making. Also CF Köhler et al [17] showed the comparison of the consumer acceptance of recommendations with concrete communication versus abstract communications. There are several patent document that includes user targeting delivery system and methods with user behaviors and/or IP addresses [18-20], but it could not find any studies and researches related to b) user targeting by specific industries in the enterprise B2B software industry. 3.2. Plan for User Targeting & A/B testing We planned two types of marketing creatives, a-1) solicitation of specific name / date / place of the event, (concrete wording “Adobe Digital Marketing Symposium, 12th April, Grand Hyatt Tokyo”) and a-2) solicitation of registration time pressure (abstract wording “Hurry! Register before fully booked!”). We executed the A/B testing with those creatives to comparing effectiveness of event registration from 15 th April to 30th April, 2014 at the top page of the event website where the biggest number of visits and visitors had in the website. After the A/B testing of a), we set the user targeting with using a system and method for selectively acquiring and targeting web contents based on users' Internet Protocol (IP) addresses which shows in Figure 2. Websites are made aware of IP addresses of interest, determined by matching attributes of current internet users for whom IP addresses are known and targeting attributes of online marketing. The websites choose whether to show the contents for each visitor from one of said IP addresses in the immediate demand. In figure 2, this scheme also provides for targeting of online contents based on updated user IP addresses and some data provided by customer database which has company name, address, phone number, industry, number of employee, revenue, business model, business scheme, etc. According to the targeting scheme we executed three types of user targeting contents which related to three different industries, such as b-1) Retail & Distribution (targeting wording “Event speakers from Rakuten Inc, and DeNA Co., Ltd.” which are one of the most famous retail & distribution company in Japan), b-2) Media & Technology (targeting wording “Event speakers from Yahoo!JAPAN and Panasonic”) and b-3) Automotive & Financial Services (targeting wording “Event speakers from Nissan Motors and Mitsui Sumitomo Credit Card”) with evaluating effectiveness of event registration comparing to default wording (wording “Adobe Digital Marketing Symposium, 12 th April, Grand Hyatt Tokyo”).
1150
Takumi Ozawa et al. / Procedia Computer Science 96 (2016) 1147 – 1155
In Table 1 & 2, it shows all experiences of those user targeting and A/B testing.
Table 1. Experiences of A/B Testing.
A/B Testing Experiences : a) Concrete versus Abstract wording a-1) Default (Concrete wording) “Adobe Digital Marketing Symposium, 12th April, Solicitation of specific name / date / place of the event Grand Hyatt Tokyo” “Hurry! Register before fully booked!”
a-2) Abstract wording Solicitation of registration time pressure
Table 2. Experiences of User Targeting.
User Targeting Experiences : b) User Targeting with 3 industries b-1) Default “Adobe Digital Marketing Symposium, 12th April, Solicitation of name / date / place of the event Grand Hyatt Tokyo” b-2) Retail & Distribution Solicitation of company name of event speakers
“Event speakers from Rakuten Inc, and DeNA Co., Ltd.”
b-3) Media & Technology Solicitation of company name of event speakers
“Event speakers Panasonic”
b-4) Automotive & Financial Services Solicitation of company name of event speakers
“Event speakers from Nissan Motors and Mitsui Sumitomo Credit Card”
from
Yahoo!JAPAN
and
Fig. 2. Scheme of User Targeting.
In the user targeting and A/B testing, same experience was displayed as long as visitors did not delete the cookies of the web browser, and the test experience to be displayed randomly and equality for first time visitors.
1151
Takumi Ozawa et al. / Procedia Computer Science 96 (2016) 1147 – 1155
4. RESULT OF USER TARGETING & A/B TESTING 4.1. Result of A/B testing It is the results of user targeting and A/B testing in Table 3. The conversion rate is calculated "Number of Success Registrations divides by Visits". The test result was Experience a-2) that Abstract wording ĀHurry! Register before fully booked!” has become the largest value in the conversion rate. From the previous researches the time pressure effects user behaviors and decision making. This A/B testing showed positive effectiveness of event registration, but this is one of the abstract wording. Therefore we need to continue A/B testing with other candidates of abstract wording, and figure out what experience is more effective than others. Table 3. Result of A/B Testing: Concrete versus Abstract wording.
Experiences
Visits
Success Registrations
Conversion Rate
Total
2,628
310
11.8%
a-1) Default (Concrete Wording) a-2) Abstract Wording
1,321
129
9.8%
1,307
181
13.8%
4.2. Result of User Targeting It set three types of user targeting with these industries, Retail & Distribution, Media & Technology, and Automotive & Financial Services. It could be defined the industry of each visitors with scheme of user targeting in Figure 2. If the visitor’s industry was Retail or/and Distribution, user targeting scheme showed targeted main message banner which said “Event speakers from Rakuten Inc, and DeNA Co., Ltd.”. Then we compared by A/B Testing for clarifying effectiveness of user targeting. In Table 4, the test of b-2) user targeting of Retail & Distributions, there was significant difference between Experience Default which is solicitation of name / date / place of the event (“Adobe Digital Marketing Symposium, 12th April, Grand Hyatt Tokyo”) and Experience b-2) Retail & Distribution (“Event speakers from Rakuten Inc, and DeNA Co., Ltd.”). Also in figure 3, it could be found that Experiences b-2) became a winner consistently during testing period. Both of result were good enough to identify the winner therefore it could be said that user targeting with IP address. We need to continue the user targeting with other wording for this industry to have more clear findings. Table 4. Result of User Targeting: Retail & Distribution.
Experiences
Visits
Success Registrations
Conversion Rate
Total
742
179
-
b-1) Default
366
68
18.6%
b-2) Retail & Distribution
376
111
29.5%
1152
Takumi Ozawa et al. / Procedia Computer Science 96 (2016) 1147 – 1155
Fig. 3. Result of User Targeting: Retail & Distribution.
Also if the visitor’s industry was Media or/and Technology, user targeting scheme showed targeted main message banner which said “Event speakers from Yahoo!JAPAN and Panasonic”. In Table 5, the test of b-3) user targeting of Retail & Distributions, there is difference between Experience Default which is solicitation of name / date / place of the event (“Adobe Digital Marketing Symposium, 12th April, Grand Hyatt Tokyo”) and Experience b-3) Retail & Distribution (“Event speakers from Yahoo!JAPAN and Panasonic”). Also in figure 4, it could be found that Experiences b-3) became a winner consistently during testing period. Both of result were good enough to identify the winner therefore it could be said that user targeting with IP address. Table 5. Result of User Targeting: Media & Technology
Experiences Total
Visits
Success Registrations
Conversion Rate
1441
319
-
b-1) Default
714
98
13.7%
b-3) Media & Technology
727
221
30.4%
1153
Takumi Ozawa et al. / Procedia Computer Science 96 (2016) 1147 – 1155
Fig. 4. Result of User Targeting: Automotive & Distribution.
If the visitor’s industry was Automotive or/and Financial Services, user targeting scheme showed targeted main message banner which said “Event speakers from Nissan Motors and Mitsui Sumitomo Credit Card”. In Table 6, the test of b-4) user targeting of Retail & Distributions, there is difference between Experience Default which is solicitation of name / date / place of the event (“Adobe Digital Marketing Symposium, 12th April, Grand Hyatt Tokyo”) and Experience b-4) Retail & Distribution (“Event speakers from Nissan Motors and Mitsui Sumitomo Credit Card”). Also in figure 5, it could be found that Experiences b-4) became a winner consistently during testing period. Both of result were good enough to identify the winner therefore it could be said that user targeting with IP address. Table 6. Result of User Targeting: Automotive & Financial Services.
Experiences
Visits
Success Registrations
Conversion Rate
Total
617
126
-
b-1) Default
301
43
14.3%
b-4) Auto & Fin
316
83
26.3%
1154
Takumi Ozawa et al. / Procedia Computer Science 96 (2016) 1147 – 1155
Fig. 5. Result of User Targeting: Automotive & Distribution.
5. CONCLUSIONS To improve own products and services awareness in the B2B business industry is generally often used by way of the marketing event. Improving the number of event attendees is equal to improving the products and services awareness with low cost investment. Therefore, many of the B2B companies are required to develop the effective web site for improving event awareness and number of customer registration from the cost reduction perspective. To develop the effective web site in order to achieve its purpose, it should deliver the right contents to the right visitors because they had different motivation according to their business scheme, business model, and own role and responsibilities. It means that it should figure out what kind of the customers / industries came to the website and what kind of contents motivated them to register B2B event attendance through the not only web analysis, but also the A/B testing with user targeting. We could expect significant of the business impact by running the user targeting and A/B testing at the web page which had biggest number of visits, and figure out what the experiences it should show to the targeted customers. In addition it could be found that there are differences of visitor behaviour by wordings of the main message between concrete wording and abstract wording through the A / B testing. It showed that abstract wording with time pressure of event registration deadline was effective more than abstract wording with the name / date / place of the event. Also it was executed three type of user targeting with these experiences of industries, b-1) Retail & Distribution, b-2) Media & Technology, b-3) Automotive & Financial Services using IP address and targeting scheme. Through the result of those studies we understood that there are differences in the number of conversion rate of success registration. However we did not able to fully understand the factors of those influences for website visitors because it needed to figure out the effectiveness of other wording and segment of the user targeting. We will continue to find the effectiveness of the wording and user targeting with other experiences. Acknowledgements We thank Adobe Systems Incorporated, Japan for providing the data and test results of the B2B event website.
Takumi Ozawa et al. / Procedia Computer Science 96 (2016) 1147 – 1155
1155
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.
Virtsonis, Sally Harridge-March, 2009. Brand positioning in the B2B online environment: A case from the UK print industry, s.l.: Journal of Brand Management (16), p.556–570. Wilson, R. Dale, 2010. Using clickstream data to enhance businessϋtoϋbusiness web site performance, s.l.: Journal of Business & Industrial Marketing Vol. 25 Iss: 3, p.177 - 187. Akiyuki Sekiguchi, Tadao Katsunuma, Itsumi Hokao, Yusuke Yamada, and Kazuhiko Tsuda, “Web Analytics for B to B Marketing in Semiconductor Industry”, International Journal of e-Education, e-Business, e-Management and e-Learning, Vol. 2, No. 5, October 2012 Akiyuki Sekiguchi, Kazuhiko Tsuda, 2014. Study on Web Analytics Utilizing Segmentation Knowledge in Business to Business Manufacturer Site, s.l.: KES. Ejiri T., 2008. Web Analytics and Web Marketing: Access Log Analytics Realized Web Marketing Kaizen Cycle Management systems, s.l.: Journal of Japan Industrial Management Association 18(1), p.8-43. K. Rebecca, P. Justin, P. Graeme, 2012. Ethical considerations and guidelines in web analytics and digital marketing: a retail case study, s.l.: Proceedings of the 6th Australian Institute of Computer Ethics conference 2012, p.5-12. KYODO, 2014. Benesse data leak may affect 20 million customers, including children. The Japan Times, 9 July, http://www.japantimes.co.jp/news/2014/07/09/national/benesse-data-leak-may-affect-20-million-customers-including-children/. Sampath P., 2012. An efficient weighted rule mining for web logs using, s.l.: Advances in Engineering, Science and Management (ICAESM), International Conference, p.432-436. Pascual-Cid V., 2008. An information visualisation system for the understanding of web data, s.l.: Visual Analytics Science and Technology, VAST '08. IEEE Symposium, p.183-184. Nasraoui, O. Soliman, M. Saka, E. Badia, A. Germain, R., 2008. A Web Usage Mining Framework for Mining Evolving User Profiles in Dynamic Web Sites, s.l.: Knowledge and Data Engineering, IEEE Transactions, p.202-215. X. Wang, D. Shen, H. Chen, L. Wedman, 2011. Applying web analytics in a K-12 resource inventory, s.l.: Electronic Library, Vol. 29 Iss: 1, p.20-35. S. Otsuka, M. Toyoda, M. Kitsuregawa, 2003. A Study for Analysis of Web Access Logs with Web Community, s.l.: Information Processing Society of Japan (IPSJ) Database44, p.432-436. Y. Ichikawa, M. Nakamura, Y. Kishimoto, T. Kobayashi, 2012. A Proposal of Extracting Innovative Users with Web Access Log of an ECommerce Site, s.l.: Information Processing Society of Japan, IPSJ SIG Notes 2012-GN-83(2). Aivalis C.J., 2011. Log File Analysis of E-commerce Systems in Rich Internet Web 2.0 , s.l.: Informatics (PCI), 2011 15th Panhellenic Conference, p.222-226. J. Park, K. Jung, Y. Lee, G. Cho, J. Kim, J. Koh, 2009. The Continuous Service Usage Intention in the Web Analytics Services, s.l.: HICSS '09. 42nd Hawaii International Conference, p. 1-7. Xiao-Gang W., 2009. Web mining based on user access patterns for web personalization, s.l.: Computing, Communication, Control, and Management-ISECS International Colloquium, 2009.p. 20-35. Clemens F. Köhlera, Els Breugelmansb & Benedict G. C. Dellaertc, 2014. Consumer Acceptance of Recommendations by Interactive Decision Aids: The Joint Role of Temporal Distance and Concrete Versus Abstract Communications, s.l: Journal of Management Information Systems, Volume 27, p.231-260. Kevin Wueste, Concentric user-targeting delivery system and methods, US20030115318 A1. Christina Yip Chung, Joshua M. Koran, Long-Ji Lin, Hongfeng Yin, Model for generating user profiles in a behavioral targeting system, US 7809740 B2. David A. Gerken, System and method for selectively acquiring and targeting online advertising based on user IP address, US 7376714 B1.