• No results found

Does the type of access device influence the online buying behavior? An explorative study

N/A
N/A
Protected

Academic year: 2021

Share "Does the type of access device influence the online buying behavior? An explorative study"

Copied!
8
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Does the type of access device influence the online buying behavior? An explorative study

   

Author: EFTHYMIOS CONSTANTINIDES - Email: e.constantinides@utwente.nl  

 

University: UNIVERSITY OF TWENTE  

 

Co-author(s): Nicolai Fabian (University of Groningen) / Sjoerd de Vries (University of Twente) / Petra de Weerd-Nederhof (University of Twente)

Access to this paper is restricted to registered delegates of the EMAC 2018 Conference.

   

(2)

Does the type of access device influence the online buying behavior?

An explorative study

Abstract:

The purpose of this study is to assess whether the device type (smartphone, tablet or desktop) used to access an online shop, has any effect on the customer’s buying behavior and

interaction with the site. A field experiment was conducted on the web site of an online retailer, collecting data with web analytics tools over a period of six weeks. To validate the findings and reveal underlying reasons for certain behavioral patterns, a customer survey was used. Our main findings are that online customers use different types of devices to perform different activities at different stages of the customer journey. The device type is essential for the information amount customers receive during a shopping session. Besides, smartphone and tablet usage behavior are not similar although both are touch-screen devices. The study provides new insights into the online customer behavior, presents a new way of observational research and allows practitioners to better optimize their websites.

Keywords: Digital Marketing, E-commerce, online behavior Digital Marketing Track

(3)

1. Introduction

The online distribution of goods and services by means of internet-based platforms,

commonly called Ecommerce, is expected to reach about 14 % of the entire retail volume by 2020 (eMarketer, 2016a) while in 2018, 5.7 billion people will be in possession of at least one internet capable device such as smartphone, tablet or desktop computer. These devices make Ecommerce ubiquitous and re-shape the customer journey (Forrester, 2011). The effects of Ecommerce on consumer behavior and the implications for marketing is an issue that is often debated among academics and practitioners (Marketing Science Institute, 2016) but there is limited research attention on the effects of device types on buying behavior of online shoppers. This study addresses this research gap by means of a field experiment making use of a commercial web site. The main aim of this approach is to assess whether the device type used in accessing web shops (smartphone, tablet, desktop) has any effect on purchase behavior in a real life observational setting.

2. Literature Review

The issue of consumer behavior as well as influencing factors of consumer behavior in in web environments is one widely discussed in the literature, but relatively little research attention is given on the effects of using different device types for accessing commercial web sites and the implications on consumer behavior. Such knowledge can help online marketers optimize their online approaches and proposition in order to increase customer acquisition and

retention.

In web environments privacy is an important factor to consider. Research by Tsai et al. (2011) indicated that perceived privacy invasion results in negative attitude towards a brand or a web shop; privacy issues result in postponing a purchase decision or move to another web shop (Kim, Ferrin and Rao, 2008). Aguirre et al. (2015) identify a similar effect when remarketing is used: if customers are often exposed to certain ads, they feel vulnerable and avoid clicking on the ad. With regard to the website itself a number of elements have been identified as affecting consumer behavior online. Usability, interactivity and aesthetics of web sites can have an effect on customer preference for an e-shop against another (Constantinides and Geurts, 2005). Even factors like the time of the day influence customer online behavior: Presman (2015) found that customers are most likely react to emails with shopping intent in the morning and in the late afternoon.

With regard to the effects of the device type on customer behavior, Ghose et al., (2013a) found that smaller screen size is associated with higher time investment and leads therefore to higher search costs. The same effect was observed by Liebe et al. (2015) who compared time spending of different devices when filling out a survey. They discovered that smartphone users need more time compared to other devices, but this has no effect on the survey quality. Regarding tablets as device type, Burford and Park (2015) argue that the strong focus of apps, common in such devices, limits the amount of information customers receive. Xu et al. (2016) found that tablets encourage casual browsing behavior which also effects online sales.

According to Xu et al. (2016) and Lee et al. (2017) smartphones and tablets complement each other in the purchase journey and customers typically spend less time per page when using smartphones compared to desktop devices (Chaffey, D. 2017).

3. Methodology

Xu et al. (2016) used data sets and statistical analysis to determine the influence of tablets on consumer buying behavior while surveys are commonly used to identify consumer

(4)

behavior which is according to Chandon et al.(2005) a feasible approach as intended behavior is a good predictor of actual behavior. Field experiments meant to observe actual customer behavior during online interactions are also a method often used for observing online customer behavior (Ghose et al., 2013a). A widely used tool for analyzing online customer behavioral data is the Google Analytics which allows the collection of data on gender, age, device type, visit frequency, search behavior or bounce rate (DeMers, J. 2014; Pakkala et al. 2012). Kaur and Singh (2015) analyzed customer click behavior in order to identify points of interest on a web page. Choros (2011) used the tool Hotjar, and applied heat, tap and scroll maps to draw inferences about customer behavior and web site usability. Hence these methods allow to observe consumer behavior right on page and this information can then be utilized to help online marketers to improve the page as well as marketing messages.

In previous studies about online device types archival data and surveys have been used (Bart et al., 2014; Xu et al., 2016). Rarely the field experimental approach, as utilized by Ghose et al, (2013a) providing insights on behavior by observing customers live on webpages has been used. In this study a new approach is used that has advantages against surveys based on customer intentions and methods based on analysis of old archival data. In cooperation with a Dutch online commercial organization, a real live experimental setting with a three-step research approach was used: observational elements on the web page as well as a customer survey are used in combination in order to gain deeper insights in customer actions and motivations.

• In the first step Google Analytics was installed on the website of the company to track demographic data as well as general data such as time on page or device type used to access the site.

• The second step consists of the tool Hotjar which is used to generate different heat, scroll and tap maps based on customer click behavior. Those are used to draw inferences about consumer interaction with the page on different devices. Both tools complement each other providing information about how consumers behave per page on individual and collective levels.

• In the third step a customer survey was used to explain the underlying motivation for certain behavioral patterns of the visitors and to further validate the findings of the previous steps such as demographic data or device type.

The analysis was focused to a single category page (different products of the same category in the online shop) in which the tooling was installed. The page was chosen because of the product type displayed: a seasonal commodity item, attracting a diversified flow of visitors from all age groups. In a timeframe of 6 weeks, this page was observed, and data was collected. To minimize the potential bias, the emailing list of the company was used to distribute the survey among all web site users. The survey itself was then send out in week 5 of the observation. To reduce the threat of bias in the survey it was announced as a website improvement survey. For ensuring validity and reliability of the collected data, both samples were compared using different observed variables such as device type, gender and age. Main goal of the survey was to first validate the data about demographics (age/gender) and device type. Secondly purchase motivation with certain device type was asked to compare actual observed behavior to intended behavior.

Due to the nature of the research and the used data collection methods, there is a discrepancy in sample size between the different methods. For Google Analytics (GA) the sample consists of 805 customers in the 6 weeks’ time frame. In Hotjar (HJ) the sample consists out of 530 visitors which equals 65% of the GA sample. This is due to the fact that HJ automatically draws a sample of the visitors. In the survey there was a total number of 395 responses. As

(5)

only complete answers and only people who have visited the website before we counted, the final sample consisted of 286 respondents.

4. Results

The analysis of data collected indicates that there are differences in time spend on a e-shop page for customers using different devices. While smartphone users on average spent 0.53

minutes per page, desktop users spent 1.06 min and tablet users 1.04 min. This indicates that on average customers using smartphones are staying on average 12 seconds less on a page. Scroll maps created with HJ were created where the scroll depth per device type was measured: how “deep” users scrolled on a page. To account for differences in page designs across devices, the number of products users have scrolled through was counted (see Table 1). We observe here that smartphone users typically see 20% less of the products (or that tablet/desktop users scrolled 20% deeper). This indicates that smartphone users saw indeed less products compared to bigger screen devices. Next to that using click maps allows to observe the click behavior: we found that smartphone users did not apply filter or sorting functions on page.

The findings indicate that that smartphone users spend less time on page, see less products and do not use filter functions. Hence, we can affirm that customers on smartphones make a more limited and superficial use of the information available in an e-shop compared to users of bigger screen devices such as tablet and desktop users.

By employing click map analysis on visitor journeys we looked for similarities in click behavior between smartphones and tablets as touch devices as well as differences of both to desktops (see Fig. 1). Using different points of interest at the page (product description, product picture, CTA button) as unit of analysis, we found that there was a big variance between those points and the different devices. The variance in clicks was on average between 5%-13%. Besides that, in Google Analytics data there were no similarities between the touch devices

concerning time on page or other metrics which would indicate that usage is similar. Hence, one can conclude that even though both smartphones and tablets are touch devices, there are no similarities in usage between these devices. There were also no similarities between tablet and desktop a fact indicating that these devices complement each other.

As to the findings based on comparison of the data analysis and the customer survey, the most important findings are the following.

Comparing whether customer intentions for certain behaviors matches actual customer behavior, we compared our findings from Google Analytics to the survey data. In the survey we found that customers indicated a strong preference (50%) to conduct a purchase activity on the internet with their desktop device rather than with smartphones or tablets. However, in the observation data based on actual behavior desktop devices only accounted for 37.64% of the total visits. In the question what device is most likely to be used for information search we found that intentions and actual behavior were more closely together. To further test how the differences in purchase device usage arise, we analyzed device usage during a day period. We found that during working hours (9-17) tablet and desktop usage is

(6)

significantly higher than during evenings and commuting time. In the evenings, mostly smartphone devices are used. One can therefore conclude that in the customer journey, different devices are more dominantly used for different tasks while the time of the day is a possible determinant for the choice of device type. This means that customers have different touch points over the day with a brand with different devices.

To account for differences in measurement across different data collection method and to ensure validity and reliability of the study we cross-compared the results of the different methods. Between Google Analytics and HJ the variance of devices types was minus 1 percentage point for tablets, plus eight percentage points for desktops and minus seven percentage points for smartphone. While between Google Analytics and the survey the values were identical to the previous presented values. Concerning demographics between Google Analytics and the survey, there was a variance for gender of plus eight percentage points more in the survey while for age the total variance for all groups was around one to three percentage points. The differences can partly be explained by the differences in measurement methods of the different programs. The differences observed are insignificant and indicate that the data obtained for the two different methods belong to the same

population which further strengthens the validity of the findings.

Conclusions and Practical Implications

This study observes customers in a real-life setting: interacting with a web page, while controlling their intentions and motivations in the same moment. We found that smartphones limit the number of products customers see when browsing through a web shop site. Next to that we did not find similarities in usage behavior between different devices which indicates that each device type could have a specific purpose in the customer journey. Which is supported by our finding of dominantly used devices at certain times of day. Finally, our method provides interesting insight in customer behavior by means of observation on page rather than analyzing secondary data.

Implications of this findings could be either way useful to make improvements in web design and customer friendly web shopping. Otherwise companies can also exploit that knowledge to further limit consumers and present items first which have higher profit margins for the company. The societal relevance of point can be found in litigation cases about search engine results pages, where those practices and their possible malicious implications on behavior, in this case by Google Shopping, are already under question (European Commission, 2017). We could imagine that further experimental research could reveal how strong the effect of limitation between devices in different situations is.

Secondly, we found further evidence that there are indeed differences in consumer behavior among certain devices and across the customer journey. We rely on the findings from previous research which suggest that intention is a good predictor of actual behavior (Chandon, Morwitz and Reinartz 2005). The self-reported data in a one wave survey does not provide further insights about causality which could better be tested in experimental settings or with a longitudinal setting. We also did not account for factors such as trust due to the scope of this study (Tsai et al. 2011). Next to that we rely on a very limited setting to first test the implications of this new methodology. In the current explorative setting we believe that already these findings are relevant and provide insights from another angle. By further tests on bigger scale and as complement to other studies even more relevant insight can be made.

The outcomes of this study are a relevant contribution to the study of the online customer behavior due to a variety of reasons; the study further shed light on the impact of an

(7)

indicate that smartphones have an influence on purchase behavior since they limit the customer is a number of ways. With further use of sophisticated tracking technology or experimental settings, more relevant findings on the magnitude of these limitations can be studied. We also found evidence that there are differences in the customer journey when taking device category into account: different devices are likely more dominantly used for certain tasks. Further research can examine what the exact influence of each device type in the customer journey is or to shed light on other areas as done in previous research (Pakkala et al. 2012/ Choros, 2011).

The explorative character of this study suggests that we have to be careful with the

generalizability of the study findings. The geographical area where the study is conducted (The Netherlands) and the experimental character in a particular type of online are limiting factors here. However, we believe that examining differences in consumer behavior on small scale with new means of data collection can contribute to the development of new

methodologies making better use of available analytics tools for online marketing situations. This could make the limited usage of observational approaches in web environments

especially considering the evolving capacities of tracking technology, a thing of the past. With the use of observational research in web environments, the effects of certain marketing actions on customers can be better examined and explained. This is especially societally relevant due to further digitalization and with regards to litigation cases in which practices of digital companies are already under question.

REFERENCES

1. Aguirre, Elizabeth M., Dominik Mahr, Dhruv Grewal, Ko de Ruyter, and Martin Wetzels (2015). “Unraveling the Personalization Paradox: The Effect of Information Collection and Trust-building Strategies on Online Advertisement Effectiveness,” Journal of Retailing, 91, 1, 34–49.

2. Chaffey, Dave (2017). “ Mobile Marketing Statistics compilation”. Retrieved on April 13, 2017 from

http://www.smartinsights.com/mobile-marketing/mobile-marketing-analytics/mobile-marketing-statistics/

3. Choroś K. (2011). Further Tests with Click, Block, and Heat Maps Applied to Website

Evaluations. In: Jędrzejowicz P., Nguyen N.T., Hoang K. (eds) Computational

Collective Intelligence. Technologies and Applications. ICCCI 2011. Lecture Notes in Computer Science, vol 6923. Springer, Berlin, Heidelberg

4. Constantinides, E. and Geurts, P. (2005). “The Impact of Web Experience on Virtual Buying Behaviour: An Empirical Study”, Journal of Customer Behaviour, 4/3, 307-336

5. DeMers, J. (2014). Forbes, “15 Google Analytics Tricks to Maximize your Marketing Campaign”. Retrieved April 13, 2017 from

https://www.forbes.com/sites/jaysondemers/2014/08/20/15-google-analytics-tricks-to-maximize-your-marketing-campaign/#99f871f3fb9d

6. eMarketer, (2015). Tablet Users to Surpass 1 Billion Worldwide in 2015. (2015, January 8). Retrieved April 13, 2017 from https://www.emarketer.com/Article/Tablet-Users-Surpass-1-Billion-Worldwide-2015/1011806

7. eMarketer, (2016a). Worldwide Retail Ecommerce Sales Will Reach $1.915 Trillion This Year. (2016, August 22). Retrieved April 13, 2017, from

https://www.emarketer.com/Article/Worldwide-Retail-Ecommerce-Sales-Will-Reach-1915-Trillion-This-Year/1014369

8. eMarketer, (2016b). Mobile Phone, Smartphone Usage Varies Globally. (2016, November 23). Retrieved April 13, 2017 from https://www.emarketer.com/Article/Mobile-Phone-Smartphone-Usage-Varies-Globally/1014738

(8)

9. European Commission (2017), “Antitrust: Commission fines Google €2.42 billion for abusing dominance as search engine by giving illegal advantage to own comparison shopping pages” retrieved July 7 from http://europa.eu/rapid/press-release_IP-17-1784_en.htm

10. Forrester (2011), Welcome to the era of Agile Commerce, Brian Walker, Forrester Research Inc, Cambridge, MA

11. Ghose, Anindya, Avi Goldfarb, and Sand P. Han (2013a). “How is the Mobile Internet

Different? Search Costs and Local Activities,” Information Systems Research, 24, 3, 613– 31.

12. Kaiquan Xu, Jason Chan, Anindya Ghose, Sang Pil Han (2017) Battle of the Channels: The Impact of Tablets on Digital Commerce. Management Science 63(5):1469-1492.

13. Kaur, Kawaljit., Singh, Hardeep (2015). “Analysis of Website using Click Analytics”, IJCSET, 5,

14. Kim, D. J., Ferrin, D. L., & Rao, H. R. (2008). A trust-based consumer decision-making model in electronic commerce: The role of trust, perceived risk, and their antecedents Decision support systems,44(2), 544-564.

15. Lee et al. (2017). “An empirical analysis of tablet PC diffusion”, Telematics and Information, 34, 2, 518 - 527

16. Liebe, Glenk, Oehlmann and Meyerhoff, J. (2015). „Does the use of mobile devices (tablets and smartphones) affect survey quality and choice behaviour in web surveys?, Journal of Choice Modelling, 14, 17-31

17. Marketing Science Institute (2016). “Research Priorities 2016-2018.” Cambridge, Marketing Science Institute. Accessed on 13 Nov. 2017

http://www.msi.org/research/2016-2018-research-priorities/

18. Pakkala, H., Presser, K., Christensen T. (2012). “Using Google Analytics to measure visitor statistics: The case of food composition websites” International Journal of Information Management 32 (2012) 504–512

19. Patel, Neil. (n.a.). How to get actionable data from google analytics in 10 minutes, retrieved 25.05.2017 from: http://neilpatel.com/blog/how-to-get-actionable-data-from-google-analytics-in-10-minutes/

20. Pierre Chandon, Vicki G. Morwitz, Werner J. Reinartz (2005) Do Intentions Really Predict Behavior? Self-Generated Validity Effects in Survey Research. Journal of Marketing: April 2005, Vol. 69, No. 2, pp. 1-14.

21. Presman, J. (2015). Remarkety, What´s the difference? Email Marketing Vs Email Remarketing (2015, May 29). Retrieved April 13, 2017 from

https://www.remarkety.com/whats-the-difference-email-marketing-vs-email-remarketing 22. Sally Burford, Sora Park, (2014). “The impact of mobile tablet devices on human

information behavior”, Journal of Documentation, Vol. 70 Issue: 4, pp.622-639,

23. Tsai, Janice, Serge Egelman, Lorrie Cranor, and Alessandro Acquisti (2011). “The Effect of Online Privacy Information on Purchasing Behavior: An Experimental Study,”

Referenties

GERELATEERDE DOCUMENTEN

According to the results of gender, age, education and beliefs of consumers, there is no relationship with consumers’ buying behavior of bottled water.. CHAPTER 6 Conclusions

H2: The effect of motivation to buy the meat substitute is stronger when a congruent food label (animal welfare vs climate change vs healthy lifestyle) is used than when an

Validity refers to the extent to which the data collection method measures what it was intended to measure (Saunders et al., 2007) In other words: did the

Individuals behave in more unethical ways when they have a high love of money as opposed to a low love of money and this effect is stronger when one has a

Furthermore, adolescents are believed to be more receptive to peer influence (Berndt, 1979; Steinberg & Silverberg, 1986). This could be interesting whether these effects

To test the hypothesis that consumers under time pressure are more likely to make an impulse purchase, particularly when having a fast LHS (H3), a 2 (time stress; high vs.

Website Quality: • Usefulness • Ease of Use • Entertainment Online Customer Experience: Cognitive Emotional Type of device: mobile vs.. fixed

Personalities don’t seem to have a large impact on hedonic and utilitarian shopping motives overall, but when these are split up into multiple underlying shopping motives,