INTRODUCTION:

The creation and use of intelligent agents using a computer software and machines capable of addressing cognitive problems by simulating human intelligence is termed as Artificial Intelligence or Computation Intelligence¹⁹.

AI provides variety of add-ons to the businesses that promise to enhance relationship with the stakeholders, to seamlessly upgrade and promote marketing initiatives and to positively impact the consumer buying behavior process. Over the last few years, online interactions have become preferred mode to obtain customer support⁴. The Pandemic of 2019 has further enhanced the role of digital technologies and artificial intelligence as an important means of interaction between the buyers (consumers) and the sellers (brands). As marketers strive to combine human intellect with technology, there has been an increasing interest in the nature of the interaction of the chatbots with the consumers at different stages of their buying decision processes.

CHATBOT combines the words ‘chat’ and ‘robot’ and can be defined as a computer program which is capable of interacting like a human with the help of text-based dialogue system. Chatbot(or a chatterbot) is an artificial intelligence(AI) feature that can be embedded and used through various OTT Platforms. Chatbots can also referred to as Talkbots, Bot, IM Bot. Key examples are Facebook messenger or virtual assistants like amazon, Alexa, GoogleSiri.

The first chatbot was created by Joseph Wiesenbaum, and was called Eliza in year 1966. It all started when Alan Turing printed an editorial named “Computer Machinery and Intelligence”, and raised an intriguing question, “Can machine think?”, and since then, we've witnessed multiple instances of computers capable of exhibiting human-like intellect.

A report by Gartner claims that by the year 2022, 70% of white-collar employees will interact with colloquial platforms on a daily basis and that 80% of companies are expected to own some version of Chabot automationby202⁶.

The objective of this research paper is to identify and to analyze the impact of AI enabled Chabots on the pre-purchase, purchase, post purchase behaviours of consumers in Delhi NCR and to suggest implications of the study to the marketers. Further, this research paper aims to Understand how chatbots interactions are viewed by consumers across different demographic categories such as age, gender, income and education.

LITERATURE REVIEW:

During the early days of the pandemic the streets were quiet and many businesses across the countries were struggling for survival. However, the businesses also recognized the absolute need and opportunity in the moment to capitalize on the use of digital technologies to adapt modify and initiate newer ways of interacting with and influencing the consumers. Digital technologies seem to be quickly changing the environment within which our businesses are operating. The most important step in analyzing the interactions between digital technologies and the consumers starts with the investigation of the consumer behavior which is ever evolving both in the online and mobile context¹⁴. Maintaining healthy and mutually beneficial consumer relationships is the main maneuver of businesses today²¹. Therefore, it is necessary for businesses to identify exactly how their customers behave and how they anticipate organizations to interact with them². The substantial changes and developments in technology in the recent years coupled with the rise in the use of mobile devices and applications and the latest advancement in AI and machine learning has increased the researchers’ interest in the chatbots³. From providing weather reports to automating basic customer support functions, chatbots are capable of handling it all. Bots allow user to get personalized and focused interactions without putting much pressure on the limited human resources. Since the introduction of smart phones and mobile applications chatbots have been mostly used for messenger apps rather than on computers. According to Business Insider, about 3 billion people worldwide use mobile messaging applications such as Facebook Messenger, We Chat, Skype, Telegram, Slack, Viber, and Kik. For chatbots. Chatbot is an umbrella term including similar AI based software’s like chatterbots, digital marketers and conversational retailers.⁸^,³¹. Tidio Analysis in²⁷ the month of January 2020 found that 43% consumers favor to message an online Chabot rather than phone customer care service. This is one of fastest digital marketing trend. Applying chatbots for customer contact processes has many advantages. The firm and the customers are able to contact with one another easily and quickly because of the embedding of chatbots in messenger services. Customers want the Chabot to provide quick information¹⁸. However, any Innovative or new system that a business applies is beneficial only if its customers accept such a system¹⁴. Research also suggests that the value of a service is based on the gap between their expectations and the reality²⁰. However, AI enabled chatbot implementation comes with its own challenges. There’s a striking distinction between completely different generations of consumers and the manner in which they perceive technology Corporations that fail to acknowledge these differences fail to successfully execute the language-based chatbot application. One should be able to identify the communication channels that the target market is using and accommodate their needs and preferences in this AI application. It is being observed that ‘Millennials’ are ready to use text-based channels of communication, while older generations still prefer phone and email conversations²⁶.

CONCEPTUAL FRAMEWORK:

The research conceptualizes the link between consumer interface with AI enabled chatbots and the behavior which consumers exhibit through different stages of buying process, namely, pre-purchase, purchase and post-purchaser.

AI Enabled Chabot Interface:

The construct is defined through the level of respondent’s interaction with and the nature of his experience with the AI enabled chatbots.

The Antecedents of AI Enabled Chabot Interface:

Antecedents are defined by the respondent demographics such as age, gender, education and income level and the respondent familiarity with technology in terms of how frequently they use electronic devices to shop through electronic devices.

The Consequences of AI Enabled Chabot Interface:

After establishing the relationships among constructs and antecedents, consequences are identified based on the constructs in consumer buying behavior. Kotler and Armstrong¹² outline shopper behavior as ―the shopping behavior of ultimate consumers— people and households that purchase product and services for private consumption. This can be classified into three stages as follows:

1.

---

Pre-purchase behavior. This was measured with the help of a scale that determined the level of engagement and positive feeling that the respondent experienced during

His interaction with AI enabled chatbot while making a digital/ online purchase. This leads the respondent to develop an attitude towards the digital purchases.

2. Purchase behavior: This was measured with the help of a scale that determined whether the respondent would seek assistance form AI enable chatbot during his digital purchases and perceived benefits thereof Respondent’s reactions to the interface provided by AI enabled chatbots was also included in this scale to determine his behavior.

3. Post–purchase behavior: This was measured with the help of a scale that determined whether the respondent would be willing to repeatedly purchase from a digital vendor who employs AI enabled chatbots to interact with consumers. This measure gave an insight into the respondent’s intention to re-purchase digitally.

In order to examine the research problems, we propose the following hypothesis:

^H1^-^{Young
consumers experience a more satisfying interface with AI enabled chatbots.}

^H2^-^{Male consumers experience more satisfying interface
with AI enabled chatbots}

^H3^-^{More educated consumers experience more
satisfying interface with AI enabled chatbots}

^H4^-^{Consumers
with more level of income experience more satisfying interface with AI enabled}chatbots

Past studies have observed that young and educated consumers are more likely to accept the digital changes to marketing (15) describe how youth who are interested in creative production join social websites, forums, and websites geared towards specialized creation activities and how such communities always had “mechanisms in place for creators to learn from one another”. Past studies have shown how male consumers are more comfortable in adapting to the digital innovations in marketing²⁹. Only 17% women in rural Asian countries falling the category of connected users. Even in urban India, Internet use tends to be male-dominated, with 79% of users being male data by Boston Consulting Group survey. More affluent consumers have also shown to be more acceptable of digital innovations in marketing.

^H5^-^{Consumers, who often use electronic devices to shop
online, have more satisfying interface}with AI enabled chatbots

Technology familiarity of the respondents has been found to be positively related with online purchase frequency. The implications are once again that people having high knowledge of technology are more likely to shop online³⁰

^H6^-^{The more positive the consumer interface with AI
enabled chatbot; more favorable will}be his attitude towards purchasing the products online.

In today’s Web-enabled world, thanks to convenience and accessibility of the Internet, consumers often utilize the World Wide Web to obtain pre-purchase product information

(8). Here researchers propose to investigate whether or not chatbots results in a favorable consumer attitude towards digital purchases.

^H7^-^{The more satisfying the interface with AI enabled
chatbot is for consumer, the more}favorable will be his behavior towards purchasing the products online.

The behavior of online shoppers plays an unfathomable role to the success of companies and this has an over arching influence in online selling. (25).

^H8^-^{The more satisfying the interface with AI enabled
chatbot is for consumer, the higher will}be his intention to repeatedly purchase products online.

Bearden have suggested that the propensity of customers to complain depends on the degree of satisfaction or dissatisfaction, purchase importance, perceived benefits/costs of complaining, personal characteristics and situational influences. Consumer is said to be satisfied if he has intention to purchase the products again. Here we propose to find out whether Chabot creates desire to purchase the products again.

RESEARCH METHODOLOGY:

The survey of 200 respondents residing in Delhi-NCR region was conducted in the month of March 2021 to investigate the suggested hypotheses. The technique used for selecting the sample for the study was convenience sampling. The tool used for data collection was a structured, non-disguised questionnaire. Due to the second wave of pandemic, the questionnaire was administered through emails only (as Google Form). The response rate was 76% despite repeated requests and reminders to fill the questionnaire.

The questionnaire was divided into two sections: section was used to collect demographic information about the respondents and section 2 was used to record responses on respondents’ interface with AI enabled chatbots and how this interface impacts the digital purchase behavior. The survey consisted of eleven multiple-choice questions and seven 5-point Likert scale type questions adapted from past studies. The questionnaire was pre-tested with 50respondents and suggestions and modifications were duly incorporated. Reliability of each of the scales was assessed through Cronbach’s alpha coefficient.

Descriptive Statistics:

1) Demographics factors:

a) Gender

Table1: Gender-wise Distribution of Respondents

	N	%
Male	81	53.3%
Female	71	46.7%

The respondents were mainly dominated by male-53.29% whereas females were 46.71% of the total sample size.

b) Age

More than 80% respondents belonged to category of 18-35 years followed by more than 10% of the respondents in the category of 15-18 years.

Table 2: Age-wise Distribution of Respondents

	N	%
15-18	16	10.5%
18-35	122	80.3%
35-50	9	5.9%
Above 50	5	3.3%

c) Education

Table 3: Education-wise Distribution of Respondents

	N	%
Secondary education	5	3.3%
Higher secondary education	14	9.2%
Under graduation	83	54.6%
Post-graduation	38	25.0%
Doctorate /Phil	1	0.7%
Professional	11	7.2%

More than 80% of the respondents belonged to higher education categories.

d) Occupation:

Table 4: Occupation-wise Distribution of Respondents

	N	%
Student	101	66.4%
Unemployed	3	2.0%
Homemaker	9	5.9%
Private job	24	15.8%
Self employed	8	5.3%
Government job	7	4.6%

Majority of the respondents were students66.45%.

e) Income Level:

Table 5: Income-wise Distribution of Respondents

	N	%
25000-50000	38	25.0%
50000-75000	49	32.2%
75000-150000	44	28.9%
Above150000	21	13.8%

The majority of the respondents belonged to higher income groups.

f) Technology Familiarity:

Table 6: Distribution of Respondents’ Technology Familiarity

	N	%
Never	3	2.0%
Rarely	22	14.5%
Neutral	16	10.5%
Often	59	38.8%
Always	52	34.2%

Majority of the respondents were using electronic devices for shopping from often to always.

Reliability was tested with the help of Cronbach alpha coefficient. The results are shown in the table below:

Table 7: Reliability Analysis of Scales Used

Variable

Source

Scale items

Cronbach’salpha value

Perceived

Benefits

State of Chatbots Report 2018, USA

0.601

Consumer interface with AI enabled chatbot

Paraskevi Tzani, (2019).

0.646

ANALYSIS:

We analyzed the data using the descriptive statistics and the statistical tools of correlation and regression. SPSS Version2.1 was used to analyze the data. In order to examine the relationship between the study variables Karl-Pearson correlation was computed. Among the antecedent variables only technology familiarity had significant correlation with consumer interface with AI enabled chatbot. The construct had significant correlations with the consequent variables-purchase attitude, purchase behavior and purchase intention.

Results are represented as follows-

Table 8: Correlation Analysis Summary

Variables	Mean	Standard deviation	Correlation with AI enabled chatbots interface
Age	2.02	0.55	-0.123
Gender	1.47	0.50	-0.134
Income	2.32	0.99	0.069
Education	3.32	1.03	-0.055
Technological Familiarity	3.89	1.09	0.274*
Attitude	2.54	0.69	0.259*
Behavior	3.04	0.51	0.211*
Intention	3.09	0.95	0.292*

Note: *Correlation is significant at0.01 level (1-tailed).

We thus find that except H1, H2, H3 and H4 (pertaining to relationship between demographic factors and AI enabled chatbot interface), all other hypotheses, viz., H5, H6, H7 and H8 are supported by the survey results.

In the above table we observed significant correlation between technological familiarity, attitude, behavior, intention and our construct i.e., interface with AI enabled chatbot. Simple Linear regression analysis was used to investigate the cause-and-effect relationship between the construct and these variables. Results of the above analysis are as follows-

We observed that, all relations came out to be significant. From the above table we observed That 6.9% variation in AI enabled chatbot interface were caused by technological familiarity of The respondents. Also, 6.1% variation in respondents’ purchase attitude towards online Purchases, 8.5% variation in respondents’ purchase behavior towards online purchases and 4.5% variation in respondents’ purchase intention towards online purchases was caused by AI Enabled chatbot interface.

CONCLUSIONS AND IMPLICATIONS:

The study proposed to investigate the relationship between AI enabled chatbot interface and online buying behavior of consumers classified across three different stages namely- pre-purchase behavior (attitude formation), purchase behavior and post- purchase behavior(intention for repeat purchases). The results of the study have serious implications of the marketers.

It was observed that the consumers with high technology familiarity experience the more positive interaction with the AI enabled chatbots. The marketers need to acknowledge this observation and ensure adequate education of the consumers in the usage of the new technological innovations so that more and more consumers can appreciate and benefit from such innovations.

The study results show that the impact of AI enabled Chabot interface was the most prominent at purchase stage of buying behavior (β = 0.292) and least significant at the post-purchase stage (β = 0.211). This implies that the marketers must aim to invest resources to enhance the quality of interface at the actual purchase stage. This may call for developing software’s with more robust algorithms that should have the cognitive capabilities to address any purchase related issue, query or consideration of the prospects with human-like skills, empathy and promptness. The resultant consumer interface will have far-reaching positive consequences for the marketers, who are tirelessly eyeing for customer loyalty.

With Alexas and Siris and the likes already ruling the consumer markets with their ability of mass customization, consumers can get familiarized to other digital innovations in the area of marketing too. What is required is that marketers mould these innovations to become more sensitive to their consumers’ needs, preferences and circumstances.

STUDY LIMITATIONS AND DIRECTIONS FOR FUTURE RESEARCH:

The study suffers from certain limitations which also posit as prospective research considerations.

Firstly, the generalizability of the study may be enhanced by employing more scientific ways of sample selection and by surveying a larger sample. The current pandemic restricted our respondent base to our mailing lists and immediate contacts only...

Then, fewer antecedents to AI enabled chatbot interface were examined in the study. The demographic factors may be studied for their moderating impact on the relationship between the construct and the consequent variables.

The scale used in the study may be conceptualized at a more abstract level and validated to measure AI enabled Chabot interface in general for better operationalization.

Future research in the area may posit to investigate the impact of AI enabled chatbot interface on online shopping behaviors of consumers with respect to specific product categories.

REFERENCES:

1. AbuShawar, B. and Atwell, E., (2015). ALICE chatbot: Trials and Outputs. Computaciony Sistemas, 19(4), pp.625-632.

2. Belleghem,S.(2016).How the digital world is shaping the modern consumer. Strategic Direction, 32(5),1–2.https://doi.org/10.1108/sd-02-2016-0026

3. Brandtzaeg, P.B. and Folstad,A., (2017). Why people use chatbots. s.l., Springer, pp. 377-392

4. Charlton,G.(2013).Consumers prefer live chat for customer service :stats.

5. Goasduff, L. (2019, July 31). Chatbots Will Appeal to Modern Workers –

7. Great Learning Team. (2020, December 22). Basics of building an Artificial Intelligence Chatbot. Great Learning Blog: Free Resources What Matters to Shape Your Career! https://www.mygreatlearning.com/blog/basics-of-building-an-artificial-intelligence-chatbot/

8. Häubl, G., & Trifts,V.(2000). Consumer Decision Making in Online Shopping Environments: The Effects of Interactive Decision Aids. Marketing Science, 19(1), 4–21.https://doi.org/10.1287/mksc.19.1.4.15178

9. Hernández, Blanca & Jiménez-Martínez, Julio & Martín, M. (2011). Age, gender and income:Do they really moderate online shopping behaviour? Online Information Review.35.113-133.10.1108/14684521111113614.

10. Investopedia(https://www.investopedia.com/terms/c/chatbot.asp)

11. Jog, Yatin and Dhote, Tripti and Gavade, Nutan& Shrivastava, Gesu. (2015). Effectiveness of Digital Marketing in Education: An Insight into Consumer Perceptions. Indian Journal of Science &Technology. 8.200-205.10.17485/ijst/2015/v8iS4/69267.

12. Joseph Weizenbaum’s ELIZA: Communications of the ACM, January 1966. (1966). JosephWiesenbaum.https://www.csee.umbc.edu/courses/331/papers/eliza.html

13. Kaczorowska -Spychalska, Dominika. (2019). How chatbots influence marketing. Management.23.251-270.10.2478/manment-2019

14. Kannan, P. K., Li, H. “A.”. (2017). Digital marketing: A framework, review and research agenda. International Journal of Research in Marketing, 34(1), 22–45. https://doi.org/10.1016/j.ijresmar.2016.11.006

15. Kowalska, M. (2012). The Internet Impact on Market Behavior of Young Consumers. Journal of international studies, 5(1), 101–106. https://doi.org/10.14254/2071-8330.2012/5-1/13

16. Khanna, Anirudh and Pandey, Bishwajeet and Vashishta, Kushagra and Kalia, Kartik and Bhale, Pradeep kumar and Das, Teerath.(2015). A Study of Today's A.I. through Chatbots and Rediscovery of Machine Intelligence. International Journal of u- and e-Service, Science andTechnology.8.277-284.10.14257/ijunesst.2015.8.7.28.

17. Kumavat,P.P.(2012). ABookReviewOnPrinciplesOfMarketingManagement14ThEditionBy Philip Kotler and Gary Armstrong. International Journal of Scientific Research, 1(4), 83–85.https://doi.org/10.15373/22778179/sep2012/29

18. Maddox, M.(2018). Your customers are changing, your customer communications should too

19. Mhatre, N., Motani, K., Shah, M., & Mali, S. (2016). Donna Interactive Chat-bot acting as a Personal Assistant. International Journal of Computer Applications, 140(10), 6–11.https://doi.org/10.5120/ijca2016909460

20. Parasuraman, A., Zeithaml, V. A., & Berry, L. L. (1985). A Conceptual Model of Service Quality and Its Implications for Future Research. Journal of Marketing,49(4),41–50.https://doi.org/10.1177/002224298504900403

21. Royle,J., and Laing, A.(2014).The digital marketing skills gap: Developing a Digital Marketer Model for the communication industries. International Journal of Information Management, 34(2), 65–73. https://doi.org/10.1016/j.ijinfomgt.2013.11.008

22. Schrier, C. (2018, November 27). How Demographics Affect AI Chatbot Usage.

23. SEITZ, W.D.(1972). Consumer Education as the Means to Attain Efficient Market Performance. Journal of Consumer Affairs, 6(2), 198–209. https://doi.org/10.1111/j.1745-6606.1972.tb00512.x

24. Shawar, Bayan & Atwell, Eric. (2007). Chatbots: Are they Really Useful? LDV Forum. 22.29-49.

25. Sienkiewicz, A. (2021, January 29). 11 Chatbot Statistics and Trends You Need to Know in2021.

26. SmartAction.https://www.smartaction.ai/blog/demographics-affect-chatbot-adoption-use/

27. Tidio. https://www.tidio.com/blog/chatbot-statistics/-0015.

28. Tzani, Paraskevi.(2019). The application of chatbots in Digital marketing communications and their effect on consumer behavior.10.13140/RG.2.2.13566.74563.

29. Voelker, R.(2001). Bridging the Digital Divide. JAMA, 285(2), 156.https://doi.org/10.1001/jama.285.2.156-jwm00011-4-1

30. Willett, SandraL."Preliminary Statement, "A handout prepared for the Panelon Consult Education of the 1974 meeting of the Education Commission of the States, June21,1974.

31. Xiao,B., and Benbasat, I.(2007).Consumer decision support systems for e-commerce change new

Received on 11.09.2021 Modified on 19.11.2021

Asian Journal of Management. 2022;13(1):11-16.

DOI: 10.52711/2321-5763.2022.00003

Independent variable	Dependent variable	_R2	Beta standardized coefficient	T statistics	P-value
Technological familiarity	Chabot interface	0.069	.274**	3.484	<.001
Chabot interface	Attitude	0.061	.259**	3.286	0.001
Chabot interface	Behavior	0.085	.292**	3.740	<0.001
Chabot interface	Intention	0.045	.211*	2.649	0.009