Professional Documents
Culture Documents
Singh 2019
Singh 2019
www.emeraldinsight.com/0972-7981.htm
JAMR
17,2 Analyzing the startup
ecosystem of India: a Twitter
analytics perspective
262 Shiwangi Singh, Akshay Chauhan and Sanjay Dhir
IIT Delhi, New Delhi, India
Received 22 August 2019
Revised 14 October 2019
Accepted 14 October 2019
Abstract
Purpose – The purpose of this paper is to use Twitter analytics for analyzing the startup ecosystem of India.
Design/methodology/approach – The paper uses descriptive analysis and content analytics techniques of
social media analytics to examine 53,115 tweets from 15 Indian startups across different industries. The study
also employs techniques such as Naïve Bayes Algorithm for sentiment analysis and Latent Dirichlet allocation
algorithm for topic modeling of Twitter feeds to generate insights for the startup ecosystem in India.
Findings – The Indian startup ecosystem is inclined toward digital technologies, concerned with people,
planet and profit, with resource availability and information as the key to success. The study categorizes the
emotions of tweets as positive, neutral and negative. It was found that the Indian startup ecosystem has more
positive sentiments than negative sentiments. Topic modeling enables the categorization of the identified
keywords into clusters. Also, the study concludes on the note that the future of the Indian startup ecosystem
is Digital India.
Research limitations/implications – The analysis provides a methodology that future researchers can
use to extract relevant information from Twitter to investigate any issue.
Originality/value – Any attempt to analyze the startup ecosystem of India through social media analysis is
limited. This research aims to bridge such a gap and tries to analyze the startup ecosystem of India from the
lens of social media platforms like Twitter.
Keywords Twitter, Social media, Content analysis, Startup ecosystem, Descriptive analysis
Paper type Research paper
1. Introduction
In recent times, the social media platform has been receiving increasing attention from
entrepreneurs across the world (Xiang et al., 2015; Almotairy et al., 2019). It helps in
information diffusion through platforms like Twitter, which results in strengthening the
relationship and improving the brand image (Sindhani et al., 2019). Social media platforms
attract more interest than alternative sources of information among users (Alalwan, 2018;
Simon, Goldberg, and Adini, 2015; Nisar et al., 2018). In particular, the impact of social media
platform on the startup ecosystem is important because it connects various stakeholders of
the ecosystem as well as improves the business performance (Almotairy et al., 2019). It also
impacts how a startup operates and communicates in the community. It acts as an electronic
word-of-mouth system (Bruns and Burgess, 2012; Park et al., 2016).
Given the immense rise in users of social media platforms and subsequent rise in
user-generated content, social media analytics analyzes communication pattern and
behavior in relation to external phenomenon (Hidayat et al., 2019; Garg et al., 2019; Sobti,
2019; Cao et al., 2018; Dlamini and Johnston, 2018; Kaur et al., 2018; Stieglitz et al., 2018;
Gandomi and Haider, 2015), thus investigating trends and patterns (Boyd and Ellison, 2007).
In particular, Twitter is considered to be a gold mine of data which provides a rich source of
information about the business performance and serves as a low-cost marketing medium
(Malhotra et al., 2012). Twitter is a popular medium of communication not only among
Journal of Advances in
Management Research people, but governments are also using Twitter to connect with the people (Khan et al., 2014).
Vol. 17 No. 2, 2020
pp. 262-281
Twitter has been used extensively to understand public behaviors and sentiments
© Emerald Publishing Limited
0972-7981
(Lakhiwal and Kar, 2016). Recent studies have applied Twitter analytics in contexts like new
DOI 10.1108/JAMR-08-2019-0164 product development (Rathore and Ilavarasan, 2020), neural network (Arora and Kansal, 2019),
political disclosure (Grimaldi, 2019) and global climate change (Dahal et al., 2019). However, in Analyzing the
the present start-up ecosystem, attempts to understand the strategies and moves of startups startup
through social media analysis are scant. This research aims to bridge such a gap and tries to ecosystem of
analyze the startup ecosystem in India through the use of Twitter analytics.
The objective of this study is as follows: first, to examine the Twitter content of the India
founder or co-founder of India’s top startup; and second, to analyze the sentiments of the
tweets and derive key insights about the start-up ecosystem of India. 263
The study contributes to the existing literature in three ways. First, the study extends
the literature of startup ecosystem by analyzing positive and negative sentiments on
Twitter. Second, the study also makes a methodological contribution to start-up ecosystem
research. Prior studies have analyzed various contexts of start-up ecosystem through the
use of various traditional methodologies (Singh, Sinha, Mukunda Das and Sharma 2019;
Motoyama and Knowlton, 2017; Berger and Kuckertz, 2016; Salamzadeh and Kawamorita
Kesim, 2017; Subrahmanya, 2015; Fraiberg, 2017), particularly, case studies and small-scale
surveys. Third, the study aims to identify the factors/events impacting the startup
ecosystem. Fourth, applying twitter analytics, on a large number of tweets will provide
insights on how the startup ecosystem is perceived across geographies.
The structure of this paper is as follows. Section 2 deals with literature review of the
startup ecosystem and employment of Twitter analytics in various fields. Section 3 explains
various methods and analytics techniques that are employed in research and steps taken by
the researcher in data extraction, cleansing, preparation and analysis. Section 4 highlights
the various insights and results divided into descriptive and content analysis sections that
come out as part of the research. Finally, the researcher concludes by discussing and
summarizing findings from data analysis and specifying the limitations of his research.
2. Literature review
Past research has shown the importance of social media analytics to understand the
perception and to identify the events/factors influencing a brand or product performance
(Park et al., 2016; Kaplan and Haenlein, 2010). However, most research are focused toward
the customer’s perception toward the brand or product (Michaelidou and Micevski, 2019;
Rose et al., 2019). Very few studies on social media analytics are focused toward the startup
ecosystem. To fill the gap, this study reviews the literature startup ecosystem and recent
Twitter analytics application in various domains.
Funding
7,000
6,000 6,237
5,000
4,351
4,000
3,000 3,269 3,046
2,790
2,000
1,000 1,262
Figure 1. 902 785
315 583
Year-on-year 0
funding trends of 2014 2015 2016 2017 2018
startups in Bangalore Bangalore ($ Million) Gurgaon ($ Million)
and Gurgaon
Source: Times of India
shows the adoption of social media across different geographical regions revealing the Analyzing the
promotional preference in different regions. It can help in performing market research and can startup
thus help in identifying favorable regions for marketing and high ROI (Patino et al., 2012). ecosystem of
Followers’ metrics indicates relative popularity. More the followers, higher is the popularity.
Number of tweets per Twitter handle indicates activeness in the social media space. India
Content analysis focuses on identifying and classifying actual text into various
categories and thus identifies various topics or major themes emerging from tweets. 265
Techniques like sentiment analysis, word cloud analysis, hashtag and frequency analysis,
and word frequency analysis are performed. Sentiment analysis helps to classify the
sentiments as positive, neutral or negative. Word analysis helps to understand the actual
content. Word semantic relationship is established with the help of a hashtag association
(Wang et al., 2014).
Twitter analytics has also been used for evaluation of communication and networking
success using Twitter real-time data feeds across the researcher community (Goodier, 2018).
Though the reliability of content analysis is subject to various questions (Krippendorff,
2004), there are proven statistical techniques to mine Twitter big data (Miller et al., 2014).
Several studies have shown that social media analytics does not reveal the complete picture
(Cresci et al., 2014), but nevertheless, it is an indispensable means of gaining an overview of
underlying insights. Thus, it has been seen that Twitter analytics can be employed in a
number of fields (Boyd and Ellison, 2007).
Various tweet-related studies, like use of geo-tagged microblogs for monitoring social
events, identification of spatial correlation between Twitter networks and network of airline
flights (Takhteyev et al., 2012), GIS-based time-geographic analysis of interactions and
individual activities (Shaw and Yu 2009), and crisis management and social events using
web-enabled geo visual analytics are already being conducted (Crooks et al., 2013), to name a
few. Another famous case is the research on the Internet of Things using the keywords
search on Twitter ( Joseph et al., 2017).
Twitter analytics has also been used for the evaluation of communication and
networking success using Twitter real-time data feeds across the researcher community
(Goodier, 2018). Though the reliability of content analysis is subject to various questions
(Krippendorff, 2004), there are proven statistical techniques to mine Twitter big data (Miller
et al., 2014). Several studies have shown that social media analytics does not reveal the
complete picture (Cresci et al., 2014), but nevertheless, it is an indispensable means of
gaining an overview of underlying insights. Thus, we can see that Twitter analytics has
been employed in a number of fields (Boyd and Ellison, 2007). The top five cited studies on
the application of Twitter analytics in the field of management as published in 2019 is
mentioned in Table I.
1 Grover, P; Kar, AK; Dwivedi, YK; Voting preferences Technological Forecasting and
Janssen, M Social Change
2 Arora, A; Bansal, S; Kandpal, C; Social influencer index Journal of Retailing and
Aswani, R; Dwivedi, Y Consumer Services
3 Li, B; Dittmore, S.W; Scott, OKM; Lo, Motivation differences Sport Management Review
WJ; Stokowski, S Table I.
4 Xiong, Y; Cho, M; Boatwright, B Hashtags in social movement Public Relations Review Top cited paper on
organizations the application of
5 Li, X; Xie, Q; Jiang, J; Zhou, Y; Huang, L Emerging technologies trends Technological Forecasting and Twitter analytics
Social Change (published in 2019)
JAMR 3. Methodology
17,2 The study uses Twitter analytics for analyzing the startup ecosystem of India. Twitter is a
stronger source of social media metrics than Facebook, blogs, Google+ and mainstream
media as less than 5 percent research papers are mentioned on these sources in comparison
with Twitter (Díaz-Faes et al., 2019). Twitter data analytics can help identifying new insights
and techniques for superior business performance (Kaplan and Haenlein, 2010).
266
3.1 Extraction of tweets
The tweets were extracted using a data-streaming Application Programming Interface (API)
called Tweepy. To collect and archive data, Tweepy library, tool was used. It is a platform
used to collect, analyze and manage data (Wisdom and Gupta, 2016). It helps to fetch all
tweets across the timeline by considering the restrictions posed by Twitter. It captures
tweets sample of 15 Indian startups from different industries (Table II) for analyzing the
startup ecosystem. It also analyzes the tweets posted by their Twitter handlers or their
CEO’s handle from September 2008 to February 2019. It draws conclusions upon tweets of
the 53,000+ tweets posted on Twitter. After extraction of raw tweets, data cleansing is done
using OpenRefine tool by Google. It changes the encoding of tweets to UTF-8 and performs
simple cleansing steps of removal of punctuation/emoji/foreign characters from raw data.
3.3 Process
The process of twitter analytics is shown in Figure 2. After performing the extraction of
tweets using Tweepy API and sentiment analysis, such tweets are passed into the LDA
algorithm to measure and identify common themes coming from tweets. The stop word list
is obtained from nltk.corpus library to remove words from the bag of words. After data
cleansing, term dictionary of corpus followed by Document Term Matrix is prepared.
Genism library of Python helps to prepare the LDA Model where the top 10 topics are
identified. Textblob.sentiments library contains NaiveBayesAnalyzer class, which allows us
to perform text mining of tweets and determine the common pattern of Tweets.
4. Findings
4.1 Descriptive analysis
Descriptive analysis provides higher level insights through various metrics calculations like
tweets population providing preliminary information about the sample size of tweets,
Twitter handle analysis indicating Twitter user attributes, retweets counts – providing
insights about retweets made, number of URL shared in tweets indicating web links shared
as part of promotional/informational marketing and geographical-level analysis identifying
focus on the particular geographies ( Joseph et al., 2017).
4.1.1 Tweet metrics. Among 53,115 tweets, the study identifies 3,669 unique hashtags;
12,122 tweets contain hashtags indicating that tweets intersect multiple areas of interests.
The prominent # tags in the tweets are #AIForIndia, #AI, #BigData, #Bitcoin,
#DigitalIndia, #Aadhar, #Brexit, #Budget and #AbHarWishHogiPoori showing that
startups are talking mostly about new technologies like artificial intelligence, big data and
social issues like Brexit, Digital India and Aadhaar related issues and verdicts.
Twitter is also actively being used by ecommerce startups for promotional campaigns in
the form of various hashtags like #AbHarWishHogiPoori, #BigAppShoppingDays,
#BigBillionDays and #BigOffersDay!, created to promote services and business of the
startups. This promotional technique is followed not only by ecommerce startups but also
by startups across different domains like hotel and logistic sector – #AfterhoursAtOYO and
Document Figure 2.
Data Data Descriptive Sentiment Process of
Term
Extraction Cleanising Analysis Analysis Twitter analytics
Matrix
JAMR #AmazingOla. Startups are using all the possible opportunities for social events like Kumbh
17,2 with hashtag #ChaloKumbhWithixigo to promote their business.
Startups are also concerned about environmental issues and are helping the world to
reduce environmental pollution with hashtag #BeatPlasticPollution. Digital payment
solutions are being vehemently supported by startups through #BharatQR. Trending
hashtag #customerservice shows that the startups are keeping customers at the heart of
268 their operations. This analysis of Tweets metrics shows that Indian startups are not only
keen on the adoption of latest technology trends but also are quite responsive on the latest
developments in technology space. It also reveals that startups are influenced by local
government policies like budgetary and tax measures.
4.1.2 User metrics. Among 53,515 tweets, there were 15 unique Twitter handles with
average 3,500+ tweets per startup and 12,737 replies, i.e. approx. 850 replies per startup on
average. This clearly shows that Twitter is an active medium of communication for the
startups. Visibility of each startup can be inferred using the number of retweets received by
each startup. In terms of number of retweets, we can clearly see from Figure 3 that
faisalMouthshut (Online review website startup) is the most retweeted Twitter handle
followed by maheshmurthy (Chief executive officer of digital agency Pinstorm and a veteran
seed fund investor), educartdotcom (an online educational startup) and 1kunalbahl (CEO of
snapdeal – an online market place).
User analysis shows that FaisalMouthshut (CEO of mouthshut.com) is the most popular
startup CEO among Twitter users who retweet most of his tweets as compared to other
startups. Edukartdotcom, 1kunalbahl, maheshmurthy and vijayshekher are other
prominent Twitter handles which receive a large number of retweets. This is not
surprising as mouthshut.com is an online review website founded in 2,000, and is most
popular among Twitter users. Maheshmurthy (CEO of Indian digital agency Pinstorm) is
the next Twitter user having largest number of retweets. Remaining startups like ixigo,
meseshoapp, cleartrip, chargebee, cleartax, Razorpay, etc. from a sample more or less
receive equal number of retweets, which is significantly less in number in comparison to
other startups receiving a large number of retweets.
Another insight revealed from this analysis is that ecommerce, e-payments and online
review-based startups receive the greatest number of retweets in comparison with startups
of other categories. This is easily explainable as the business of these startups heavily
depends upon online subscribers: more retweets are expected for these startups due to their
main focus on online subscribers.
4.1.3 Number of tweets by Twitter handle. The count of tweets per Twitter handle shows
which startup is most active on Twitter. Analysis of Tweets frequency per startup
1,400,000
1,200,000
1,000,000
800,000
600,000
400,000
200,000
0
Figure 3.
o
pp
ip
a
h
a
in
s
y
ay
ar
ut
l
l
ah
sa
m
ig
di
ec
be
ab
th
r tr
x_
co
sh
kh
oa
p
ix
ln
oo
ur
ST
an
ge
eb
ea
or
rta
th
he
ar
al
sh
hm
or
az
nb
td
kG
ar
ic
cl
ou
un
ea
C
ys
ee
oy
Pr
ar
ch
om
es
hi
cl
ja
m
uk
ac
ah
vi
sa
Zo
ed
_s
m
i
fa
(Figure 4) shows that Faisal (CEO of mouthshut.com, an online review website), Vijay Analyzing the
Shekhar Sharma (CEO of Paytm) and Kunal Bahl (CEO of Snapdeal.com, an ecommerce startup
company) are most active Twitter users. Thus, number of tweets per Twitter handle helps ecosystem of
us to compare and infer relative level activeness of startups among them. Average number
of tweets per Twitter handle is 2,310 which shows that Twitter being an active medium is India
the preferred medium of communication with customers. Most of the tweets have customer
experience and issue at their hearts; this pattern is evident across the industries. 269
4.1.4 Followers metrics. The popularity of a startup is generally assessed in terms of
numerical strength of its followers as shown in Figure 5. It is generally believed that more
the followers, more popular is the startup. Twitter data analysis shows that Vijay Shekhar
Sharma (CEO of Indian startup Paytm) is followed by a majority of users on Twitters
followed by Kunal Bahl of Snapdeal (an Indian ecommerce startup) and Sachin Bansal of
Flipkart (an Indian ecommerce startup). Mahesh Murty (CEO of Pinrest) also has
comparable number of followers similar to Sachin Bansal. The analysis reveals that
ecommerce-based startups are now utilizing Twitter as the major medium of reaching users
as compared to the other startups.
4.1.5 URL metrics. URL metrics analyses tweets for the presence of hyperlinks to the other
webpages. This study shows that out of more than 53,000 tweets, only 2,063 tweets contain
web links to different tweets. It further shows that less than 1 percent of the tweets contain
the URL. It indicates that the URL is generally not tweeted by the startups in their tweets.
4,500
4,000
3,500
3,000
2,500
2,000
1,500
1,000
500
0
oy app
in
R aba
p
hm o
h
ed cle in
tin
su h
lu h
ea ee
M com
a
s
Figure 4.
ee hy
m 1
y
ty
ar r
ut
un al
kG hl
es nd
a
m
tri
es ixig
ch Tec
as
Sa iris
di
ja rpa
1
ja
x_
ur
1k ans
sh
as lba
ni
om kh
b
m urt
lb
ln
oo
re nk
ar
m
bh
o
eb
ge
rta
ja
rg
hM
is tdot
na
th
S
o
Zo he
sh
a
ba sha
or
az
ic
ar
ou
ku
ys
Pr
ar
hi
Am ha
cl
twitter handle
uk
ac
ah
al
vi
s
_s
m
fa
350,000
300,000
250,000
200,000
150,000
100,000
50,000
0
hm o
ea ch
lu h
ar ia
az in
a
ip
s
C pp
ku ain
ee om
e
ba ha n
su h
oy urty
y
cl ay
he t
ar
m 11
ou hl
sa nal l
es nd
ys hu
a
es ixig
Sa iris
ab
as ebe
ti
as
Pr rth
ch lnd
x_
r tr
lM ba
cl Te
1k ns
kh
ni
om oa
p
lb
uk oo
j
m otc
k
vi ths
m
bh
eb
or
ea
u
rta
rg
hM
ja
Figure 5.
na
S
ba
ar
Zo sh
g
or
td
kG
ic
in
ar
R
h
Am ha
ja
re
Follower’s metrics
ac
ah
s
ed
_s
m
i
fa
JAMR One of the reasons for this could be the limit of 140 characters for tweets. This URL analysis
17,2 also reveals that tweets by startups contain shorter URL considering the word limit on Twitter
as opposed to full URLs. By analyzing the tweets with links, the tweets URL can be categorized
into an informative URL and promotional URL. The informative URL redirects the user to an
informational page like enabling small-scale entrepreneurs to sell online, five infrastructure
projects that will transform realty in MMR, and five Simple Tips to Save Tax by Investing in
270 Property, to name a few. Promotional URL provides a link to promote the startup on the social
media like Snapdeal handloom day.
4.1.6 Geographical-level analysis of tweets. Generally, in the Indian startup context,
we find location-based tweets analysis. Figure 6 illustrates that New Delhi based
startups and CEOs are mostly active on Twitter while Chennai-based startups are less
active on Twitter. Bengaluru is fast catching up with Delhi in terms of Twitter adoption.
This could be attributed to the huge multilingual population of Delhi, Bengaluru and
Mumbai with English as common language compared to Chennai tweeting mainly in Tamil.
More than 50 percent of total tweets are coming from New Delhi and Bengaluru based
startups, whereas the contribution of Mumbai and Chennai based startups is 30 and 8
percent, respectively. This clearly shows the preference for Twitter adoption by startups
across different locations.
It is observed that New Delhi has received approximately 15,000 tweets from
startups, whereas Chennai has less than 4,000. This shows a stark difference in terms of
startups numbers and Twitter adoption between the capital Delhi and southernmost state
of India. It shows the need to pay attention toward startup ecosystem in southern India
by the policymakers.
Chennai
Mumbai
Gurugram
New Delhi
Figure 6. Bengaluru
Location-based tweets
0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000
Word Frequency Word Frequency
Analyzing the
startup
Please 4,556 Booking 1,450 ecosystem of
India 3,104 GST 1,160
GOI 1,591 Assist 1,001 India
Sorry 1,350 Concern 665
Thank 1,316 Launched 611
Internet 255 Complaint 265 271
India 3,104 Assistance 337
Apologize 196 Digital 262 Table III.
Refund 1,090 Inconvenience 575 Word frequency
Dm 2,072 Budget 337 analysis
0.069 hashtags per tweet. Analysis of hashtag frequency shows that startups are mostly
concerned about #GST implementation as evident in the following word cloud and
particularly about tax and #GSTR1 form. It is also evident in the Information Technology (IT)
Act #66 A, which is related to the penalized sending of “offensive messages.” Other
trending topics on Twitter are promotional events of various products and services like
#oneplus5, #OLAPAY, #XIOMI and latest technologies like #BITCOIN, #CES2017 event,
etc. Few social and geopolitical events are also evident in the tweets of startups like
#BREXIT, #HIRING, #JOBS, etc.
4.2.3 Sentiment analysis. The main aim of sentiment analysis is the classification of text
into sentiment categories of positive, negative and neutral. In Figure 7, sentiment analysis of
tweets of the startups shows that 45 percent of total 53,000+ tweets are positive in nature,
while neutral sentiments are close to 44 and 11 percent of the total tweets are negative in
nature. This analysis shows that the Indian startups have more positive than negative
sentiments, which shows that the Indian startup ecosystem is optimistic. However, a lot of
work has to be done by the startups through ecosystem functioning to convert neutral
sentiments to positive ones. A snapshot of the classification of text is shown in Figure 10.
Location-based analysis of sentiments shows that New Delhi is leading in terms of
positive sentiments (Figure 8) because of the availability of infrastructure, connectivity and
huge market base of consumers from different areas of the country. Bangalore is slightly
lagging in terms of infrastructure and connectivity; hence, startups are finding it a bit
difficult to do business there as compared to New Delhi. The following example of tweets
shows the positive sentiments:
• RT @BeingPractical: 120 days. 0 Spent on marketing. 1 Million Downloads for
@PaytmMoney. and by GST Tech.
23,474, 23,780,
44% 45%
Figure 7.
5,883, Sentiment breakdown
11% of tweets
Total positive Total negative Total Neutral
JAMR 8,000
17,2 7,000
6,000
5,000
272 4,000
3,000
2,000
Figure 8. 1,000
Total positive tweets
by user location 0
Bengaluru Chennai Gurugram Mumbai New Delhi
• The webinar on Filing of GSTR 10 Final Return on GST portal in Hindi is now
available on GSTN’s Youtube channel.
• RT @gfulgoni: Stunning growth in the ad business.
In terms of negative tweets (Figure 9), Bangalore is leading with tweets mostly related to
the policy measures and infrastructure availability. The following example of tweets
shows negative sentiments, e.g., a few tweets by Sachin_bansal show common issues faced
by startups:
• @rehanyarkhan There is a long period of uncertainty coming before the world settles
down into a new rhythm. Expect turbulence for some time.
• Driving on the road why do some people overtake and then slow down in front of
you. #firstworldtrafficproblem.
• @snapdeal team supports #CarFreeDay in Gurgaon. About time we all fixed this
problem together.
Chennai, on the other hand, is at bottom of the negative as well as positive sentiments
compared to the other cities, which is attributed to the presence of a smaller number of
startups, due to language and culture barriers along with infrastructure availability.
In terms of neutral sentiments by user location, it is evident from the following analysis
that New Delhi is leading, followed by Mumbai and Bengaluru. Chennai has least number of
2,500
2,000
1,500
1,000
Figure 9. 500
Total negative tweets
by user location 0
Bengaluru Chennai Gurugram Mumbai New Delhi
neutral tweets which shows that Chennai-based startups are not using Twitter much as Analyzing the
compared to other locations. Figure 10 shows the distribution of neutral sentiments of startup
tweets by startups from various locations like Bengaluru, Chennai, Gurugram, Mumbai and ecosystem of
New Delhi.
4.2.4 Word cloud. Word cloud explains the visualization of tweet texts (Figure 11). It India
highlights the word with more frequency (Mcnaught and Lam 2010). The highlighted words
are GST, India, Fail, customers, Tax, UPI, Oneplus and others. The word cloud analysis 273
(Figures 12 and 13) of hashtags reveals that falling indices, GST implementation, Brexit
uncertainty, regulatory decisions like 66A, infrastructure issues, etc. are major issues for
startups, whereas they are optimistic about tax-related policies like E-way bill, GST Bill,
Digital payments, Aadhar verdict, etc. Figure 12 shows the negative sentiment hashtags,
whereas Figure 13 shows positive hashtags tagged by startups in their tweets.
Thus, Hashtag frequency and word cloud of frequent words provide deeper level
insights of tweets posted by the startups and their CEO. Sentiment-level analysis further
enriched the analysis through clear identification of negative and positive topics.
4.2.5 Topic modeling. The tweets were clustered into ten prevalent themes of startup
ecosystem, using LDA. The identified themes are technology and ecommerce platform,
travel industry and digital transaction, customer experience, digital connectivity and mobile
phone launch, digital payment initiatives, social issues, customer’s issue resolution, startup
issues and difficulties, government initiatives and political issues, and entrepreneurial story
of India. Each of these clusters have different keywords. The clusters and their keywords
are shown in Table IV.
8,000
7,000
6,000
5,000
4,000
3,000
2,000
Figure 10.
1,000 Total neutral tweets
0 by user location
Bengaluru Chennai Gurugram Mumbai New Delhi
Figure 11.
Hashtag word cloud
for all tweets
JAMR
17,2
274
Figure 12.
Hashtag word
cloud for negative
sentiment tweets
Figure 13.
Hashtag word
cloud for positive
sentiment tweets
Technology and ecommerce platforms are extensively using latest technological means in
their business and are persuading sellers and customers to come together at their online
marketplace. A group of researchers (Misopoulos et al., 2014) mentioned how an airline
industry is using Twitter data analysis to gauge and improve customer experience. Startups
are appreciating digital initiatives and are more focused on better customer experience.
Startups are also concerned about the various social issues like education, water crisis,
elections, online education mediums and technology. It shows that the Indian startups are
not only profit driven but also cognizant of social issues. Another import insight is that
Twitter is now being considered as the main source of resolving customer issues
immediately. Twitter allows real-time escalation of issues and may harm the reputation of
the startup; hence, real-time resolution is being provided by startups on Twitter.
Various other topics emerging from topic modeling are issues and difficulties faced by
them in their entrepreneurial journeys like infrastructure-related issues, budget concerns,
taxation difficulties and tough competition in the Indian startup ecosystem providing
Cluster
Analyzing the
number Cluster name Keywords startup
ecosystem of
1 Technology and Snapdeal, online, India, app, products, design, awesome, new, sell, now,
e-commerce platform sellers, launches, great, mobile, offline, freecharge, platform and market place India
2 Travel industry and Refund, amount, processed, airline, booking, fee, cancellation, per, day, failed,
digital transaction cleartrip, however, charged, working, any, trip, fare and transaction
3 Customer experience Thanks, good, delivered, awesome, morning, work, made, experience, like, 275
food, guys, ordered, great, glad, order, nice and quite
4 Digital connectivity India, launch, android, Samsung, galaxy, RAM, China, now, official, new,
and mobile phone announce, oneplus, phone, note, asus and power
launch
5 Digital payment Payment, online, mobile, UPI, how, using, payment, new, digital, here, app,
initiatives user, razorpay, future, tech and pay
6 Social issues Degree, role, technology, customers, water, love, service, online education,
election, education, learning, program, MBA, google, campaign, free, career
and read
7 Customer’s issue Sorry, please, soon, feedback, earliest, connect, escalated, ensure, service,
resolution concerned, discussed, surely, provider, thank, asap and detail
8 Startup issues and Old, cost, set, risks, harsh, diverse, resilience, competitors, dominating,
difficulties behavior, deceptively, regulatory, harder, ceo, founder, economic and our
9 Government Gstbill, budget, budgetology, budgetclearhoga, airways, infrastructure,
initiatives and digitalindia, govt, logical, interim, deficit, parliament, centre, invoicing,
political issues protests, economic and government Table IV.
10 Entrepreneurial Unicorns, month, results, first, movement, till, now, what, were, compliance, Clusters and
story of India day, from, destinations, implemented and countries their keywords
inspiration and preparing young startups to adopt a cautious path. Twitter can help in
interactive and reactive marketing for business and thus help startups in promoting brands
and solving customer issues (Burton and Soboleva, 2011). Therefore, startups are not only
engaging with customers for resolution of their issues, but are also discussing social issues
and sharing their experiences on the social media. Startups are also responding to the
government initiatives, political issues and difficulties; for example, GST bill, budget
announcements, Digital India campaign, Section 66A of the Indian IT Act, risks and
regulatory challenges. Thus, topic modeling generates deeper insights and reveals thoughts
of the Indian startup ecosystem.
6. Conclusion
Based on the Twitter analytics data of 53,115 tweets, the paper shares key insights
emerging from the analysis related to the startup ecosystem. The user analysis, URL
analysis, sentiment analysis and identification of themes provide a broad picture on the
focus of the Indian startup ecosystem and factors/events which boost this ecosystem. It also
provides the key cluster which forms the base of the Indian startup ecosystem and
keywords from each cluster provide insight on these identified clusters. Positive emotions
show the brighter picture of the startup ecosystem of India, whereas the negative emotions
are the ones which need to be addressed.
This research is a small step toward using data analytics techniques in the context of the
Indian startup ecosystem. However, the author acknowledges the scope for further research
in this space. First, the data collection period was very short. For a broader view, future Analyzing the
studies could collect data over various time periods. Second, this study is limited to hashtags startup
of various startups. However, the other search words which may refer to the Indian startup ecosystem of
ecosystem could be included. To investigate a meaningful pattern in social media
communication, other techniques like network mapping can be used to study the behavior of India
the components of the startup ecosystem.
277
References
Acs, Z.J., Autio, E. and Szerb, L. (2014), “National systems of entrepreneurship: measurement issues
and policy implications”, Research Policy, Vol. 43 No. 3, pp. 449-476.
Alalwan, A.A. (2018), “Investigating the impact of social media advertising features on customer
purchase intention”, International Journal of Information Management, Vol. 42 No. 1, pp. 65-77.
Al-Daihani, S. and AlAwadhi, S. (2015), “Exploring academic libraries’ use of Twitter: a content
analysis”, The Electronic Library, Vol. 33 No. 6, pp. 1002-1015.
Almotairy, B., Abdullah, M. and Abbasi, R. (2019), “The impact of social media adoption on
entrepreneurial ecosystem”, Bioscience Biotechnology Research Communications, Vol. 12 No. 1,
pp. 60-71.
Arias, M., Arratia, A. and Xuriguera, R. (2013), “Forecasting with Twitter data [TIST]”, ACM
Transactions on Intelligent Systems and Technology, Vol. 5 No. 1, pp. 1-25.
Arora, M. and Kansal, V. (2019), “Character level embedding with deep convolutional neural network
for text normalization of unstructured data for Twitter sentiment analysis”, Social Network
Analysis and Mining, Vol. 9 No. 1, pp. 1-14.
Berger, E.S.C. and Kuckertz, A. (2016), “Female entrepreneurship in startup ecosystems worldwide”,
Journal of Business Research, Vol. 69 No. 11, pp. 5163-5168.
Bocken, N.P.M. (2015), “Sustainable venture capital – catalyst for sustainable start-up success?”,
Journal of Cleaner Production, Vol. 108 No. A, pp. 647-658.
Boyd, D.M. and Ellison, N.B. (2007), “Social network sites: definition, history, and scholarship”, Journal
of Computer‐mediated Communication, Vol. 13 No. 1, pp. 210-230.
Bruns, A. and Burgess, J.E. (2012), “Researching news discussion on Twitter: new methodologies”,
Journalism Studies, Vol. 13 Nos 5/6, pp. 801-814.
Burton, S. and Soboleva, A. (2011), “Interactive or reactive? Marketing with Twitter”, Journal of
Consumer Marketing, Vol. 28 No. 7, pp. 491-499.
Cao, Y., Ajjan, H., Hong, P. and Le, T. (2018), “Using social media for competitive business outcomes: an
empirical study of companies in China”, Journal of Advances in Management Research, Vol. 15
No. 2, pp. 211-235.
Cresci, S., Petrocchi, M., Spognardi, A., Tesconi, M. and Di Pietro, R. (2014), “A criticism to society (as
seen by twitter analytics)”, IEEE 34th International Conference on Distributed Computing
Systems Workshops (ICDCSW), pp. 194-200.
Crooks, A., Croitoru, A., Stefanidis, A. and Radzikowski, J. (2013), “#Earthquake: Twitter as a
distributed sensor system”, Transactions in GIS, Vol. 17 No. 1, pp. 124-147.
Dahal, B., Kumar, S.A.P. and Li, Z. (2019), “Topic modeling and sentiment analysis of global climate
change tweets”, Social Network Analysis and Mining, Vol. 9 No. 24, pp. 1-20.
Dhir, S. and Dhir, S. (2017), “Corporate risk scorecard: a comparative study of US and German firms
risk score”, International Journal of Business Continuity and Risk Management, Vol. 7 No. 4,
pp. 277-291.
Dhir, S., Ongsakul, V., Ahmed, Z.U. and Rajan, R. (2019), “Integration of knowledge and enhancing
competitiveness: a case of acquisition of Zain by Bharti Airtel”, Journal of Business Research,
available at: https://doi.org/10.1016/j.jbusres.2019.02.056
JAMR Díaz-Faes, A.A., Bowman, T.D. and Costas, R. (2019), “Towards a second generation of ‘social media
17,2 metrics’: characterizing Twitter communities of attention around science”, PLoS ONE, Vol. 14
No. 5, pp. 1-18.
Dlamini, N.N. and Johnston, K. (2018), “The use of social media by South African organisations”,
Journal of Advances in Management Research, Vol. 15 No. 2, pp. 198-210.
Fraiberg, S. (2017), “Start-up nation: studying transnational entrepreneurial practices in Israel’s start-
278 up ecosystem”, Journal of Business and Technical Communication, Vol. 31 No. 3, pp. 350-388.
Gandomi, A. and Haider, M. (2015), “Beyond the hype: big data concepts, methods, and analytics”,
International Journal of Information Management, Vol. 35 No. 2, pp. 137-144.
Garg, E., Swami, S. and Malhotra, S.K. (2019), “Branding effectiveness measurement in non-profit
environment”, Journal of Advances in Management Research, Vol. 16 No. 1, pp. 4-22.
Goodier, S. (2018), “Evaluating the network: a workflow for tracking Twitter interactions using social
networking analysis”, Journal of Interactive Media in Education, Vol. 2018 No. 1, pp. 1-13.
Grimaldi, D. (2019), “Can we analyse political discourse using Twitter ? Evidence from Spanish 2019
presidential election”, Social Network Analysis and Mining, Vol. 9 No. 49, pp. 1-9.
Hasan, Z., Dhir, S. and Dhir, S. (2019), “Modified total interpretive structural modelling (TISM) of
asymmetric motives and its drivers in Indian bilateral CBJV”, Benchmarking: An International
Journal, Vol. 26 No. 2, pp. 614-637.
Herrmann, B.L., Gauthier, J.F., Holtschke, D., Berman, R. and Marmer, M. (2015), “The global startup
ecosystem ranking 2015”, available at: http://startup-ecosystem.compass.co/ser2015/ (accessed
February 20, 2019).
Hidayat, S.E., Rafiki, A. and Khalifa, M.H.A. (2019), “The social media adoption of public sector in the
Kingdom of Bahrain”, Journal of Advances in Management Research, Vol. 16 No. 1, pp. 23-37.
Joseph, N., Kar, A.K., Ilavarasan, P.V. and Ganesh, S. (2017), “Review of discussions on Internet of
things (IoT): insights from twitter analytics”, Journal of Global Information Management, Vol. 25
No. 2, pp. 38-51.
Joshi, K. and Satyanarayana, K. (2014), “What ecosystem factors impact the growth of high-tech start-
ups in India?”, Asian Journal of Innovation and Policy, Vol. 3 No. 2, pp. 216-244.
Kang, D. and Park, Y. (2014), “Review-based measurement of customer satisfaction in mobile service:
sentiment analysis and VIKOR approach”, Expert Systems with Applications, Vol. 41 No. 4,
pp. 1041-1050.
Kaplan, A.M. and Haenlein, M. (2010), “Users of the world, unite! The challenges and opportunities of
social media”, Business Horizons, Vol. 53 No. 1, pp. 59-68.
Kaur, I., Shri, C. and Mital, K.M. (2018), “Performance management model for teachers based on
emotional intelligence and social media competencies”, Journal of Advances in Management
Research, Vol. 15 No. 4, pp. 414-433.
Khan, G.F., Yoon, H.Y., Kim, J. and Park, H.W. (2014), “From e-government to social government: Twitter
use by Korea’s central government”, Online Information Review, Vol. 38 No. 1, pp. 95-113.
Krippendorff, K. (2004), “Reliability in content analysis”, Human Communication Research, Vol. 30
No. 3, pp. 411-433.
Kwak, H., Lee, C., Park, H. and Moon, S. (2010), “What is Twitter, a social network or a news media?”,
Proceedings of the 19th international conference on World Wide Web, pp. 591-600.
Kwok, L. and Yu, B. (2013), “Spreading social media messages on Facebook: an analysis of restaurant
business-to-consumer communications”, Cornell Hospitality Quarterly, Vol. 54 No. 1, pp. 84-94.
Lakhiwal, A. and Kar, A.K. (2016), “Insights from Twitter analytics: modeling social media personality
dimensions and impact of breakthrough events”, Proceedings of the Conference on e-Business, e-
Services and e-Society, pp. 533-544.
Liu, B. (2012), “Sentiment analysis and opinion mining”, Synthesis Lectures on Human Language
Technologies, Vol. 5 No. 1, pp. 1-167.
Low, M.B. and Abrahamson, E. (1997), “Movements, bandwagons, and clones: industry evolution and Analyzing the
the entrepreneurial process”, Journal of Business Venturing, Vol. 12 No. 6, pp. 435-457. startup
Lu, W. and Stepchenkova, S. (2014), “User-generated content as a research mode in Tourism and ecosystem of
hospitality applications: topics, methods, and software”, Journal of Hospitality Marketing &
Management, Vol. 24 No. 2, pp. 119-154. India
McNaught, C. and Lam, P. (2010), “Using Wordle as a supplementary research tool”, Qualitative Report,
Vol. 15 No. 3, pp. 630-643. 279
Malhotra, A., Malhotra, C.K. and See, A. (2012), “How to get your messages retweeted”, MIT Sloan
Management Review, Vol. 53 No. 2, pp. 61-66.
Mason, C. and Brown, R. (2014), “Entrepreneurial ecosystems and growth oriented entrepreneurship”,
OECD, Vol. 30 No. 1, pp. 77-102.
Michaelidou, N. and Micevski, M. (2019), “Consumers’ ethical perceptions of social media analytics practices:
risks, benefits and potential outcomes”, Journal of Business Research, Vol. 104 No. 1, pp. 576-586.
Miller, Z., Dickinson, B., Deitrick, W., Hu, W. and Wang, A.H. (2014), “Twitter spammer detection using
data stream clustering”, Information Sciences, Vol. 260 No. 1, pp. 64-73.
Misopoulos, F., Mitic, M., Kapoulas, A. and Karapiperis, C. (2014), “Uncovering customer service
experiences with Twitter: the case of airline industry”, Management Decision, Vol. 52 No. 4,
pp. 705-723.
Moniz, A. and de Jong, F. (2014), “Sentiment analysis and the impact of employee satisfaction on firm
earnings”, Advances in Information Retrieval, Vol. 8416 No. 1, pp. 519-527.
Motoyama, Y. and Knowlton, K. (2017), “Examining the connections within the startup ecosystem: a
case study of St. Louis”, Entrepreneurship Research Journal, Vol. 7 No. 1, pp. 1-28.
NASSCOM (2015), Startup India: Momentous Rise of the Indian Startup Ecosystem, Zinnov Consulting,
Bangalore.
Naudé, W. (2010), “Entrepreneurship, developing countries, and development economics: new
approaches and insights”, Small Business Economics, Vol. 34 No. 1, pp. 1-12.
Neck, H.M., Meyer, G.D., Cohen, B. and Corbett, A.C. (2004), “An entrepreneurial system view of new
venture creation”, Journal of Small Business Management, Vol. 42 No. 2, pp. 190-208.
Neuman, W.L. (1997), Social Research Methods: Qualitative and Quantitative Approaches, 3rd ed.,
Allyn and Bacon, Boston, MA.
Nisar, T.M., Prabhakar, G. and Patil, P.P. (2018), “Sports clubs’ use of social media to increase spectator
interest”, International Journal of Information Management, Vol. 43 No. 1, pp. 188-195.
Pang, B. and Lee, L. (2008), “Opinion mining and sentiment analysis”, Foundations and Trends in
Information Retrieval, Vol. 2 Nos 1-2, pp. 1-135.
Pant, S. (2019), “Startups-Bangalore-vs-Gurgaon”, available at: https://timesofindia.indiatimes.com/
india/startups-bangalore-vs-gurgaon/articleshow/68149030.cms (accessed February 25, 2019).
Park, S.B., Jang, J. and Michael Ok, C. (2016), “Analyzing Twitter to explore perceptions of Asian
restaurants”, Journal of Hospitality and Tourism Technology, Vol. 7 No. 4, pp. 405-422.
Patino, A., Pitta, D. and Quinones, R. (2012), “Social media’s emerging importance in market research”,
Journal of Consumer Marketing, Vol. 29 No. 3, pp. 233-237.
Ramyadharshni, S.S. and Prathiba, P. (2018), “Topic categorization on social network using Latent
Dirichlet Allocation”, Bonfring International Journal of Software Engineering and Soft
Computing, Vol. 8 No. 2, pp. 16-20.
Rathore, K.A. and Ilavarasan, P.V. (2020), “International journal of information management pre- and
post-launch emotions in new product development: Insights from Twitter analytics of three
products”, International Journal of Information Management, Vol. 50 No. 1, pp. 111-127.
Rose, S., Sreejith, R. and Senthil, S. (2019), “Social media data analytics to improve the customer
services: the case of fast-food companies”, International Journal of Recent Technology and
Engineering, Vol. 8 No. 2, pp. 6359-6366.
JAMR Roundy, P.T., Brockman, B.K. and Bradshaw, M. (2017), “The resilience of entrepreneurial ecosystems”,
17,2 Journal of Business Venturing Insights, Vol. 8 No. 11, pp. 99-104.
Salamzadeh, A. and Kawamorita Kesim, H. (2017), “The enterprising communities and startup
ecosystem in Iran”, Journal of Enterprising Communities, Vol. 11 No. 4, pp. 456-479.
Saxenian, A.L. (1994), Regional Advantage: Culture and Competition in Silicon Valley and Route 128,
Harvard University Press, Cambridge, MA.
280 Shamali, M.A., Al-Khoury, P. and Subbarao, A.N. (2019), “An empirical research on with bit coin
purchase intentions of Lebanon citizens and its effects on supply chain strategy”, International
Journal of Supply Chain Management, Vol. 8 No. 4, pp. 788-794.
Shaw, S. and Yu, H. (2009), “A GIS-based time-geographic approach of studying individual activities
and interactions in a hybrid physical–virtual space”, Journal of Transport Geography, Vol. 17
No. 2, pp. 141-149.
Simon, T., Goldberg, A. and Adini, B. (2015), “Socializing in emergencies – a review of the use of social
media in emergency situations”, International Journal of Information Management, Vol. 35 No. 5,
pp. 609-619.
Sindhani, M., Parameswar, N., Dhir, S. and Ongsakul, V. (2019), “Twitter analysis of founders of top 25
Indian startups”, Journal for Global Business Advancement, Vol. 12 No. 1, pp. 117-144.
Singh, S. and Dhir, S. (2019), “Structured review using TCCM and bibliometric analysis of international
cause-related marketing, social marketing, and innovation of the firm”, International Review on
Public and Nonprofit Marketing, available at: https://doi.org/10.1007/s12208-019-00233-3
Singh, S., Dhir, S., Das, V.M. and Sharma, A. (2019), “Interrelationships among the institutional enablers
of national innovation system”, in Ahmed, Z.U. (Ed.), Advancements in Global Business Research
across Emerging Countries, McGraw Hills, pp. 421-441.
Singh, S., Sinha, S., Mukunda Das, V. and Sharma, A. (2019), “A framework for linking entrepreneurial
ecosystem with institutional factors: a modified total interpretive structural modelling
approach”, Journal for Global Business Advancement, Vol. 12 No. 3, pp. 382-404.
Sobti, N. (2019), “Impact of demonetization on diffusion of mobile payment service in India: antecedents
of behavioral intention and adoption using extended UTAUT model”, Journal of Advances in
Management Research, Vol. 16 No. 4, pp. 472-497.
Spigel, B. (2017), “The relational organization of entrepreneurial ecosystems”, Entrepreneurship Theory
and Practice, Vol. 41 No. 1, pp. 49-72.
Stieglitz, S., Mirbabaie, M., Ross, B. and Neuberger, C. (2018), “Social media analytics – challenges in
topic discovery, data collection, and data preparation”, International Journal of Information
Management, Vol. 39 No. 1, pp. 156-168.
Subrahmanya, M.H.B. (2015), “New generation start-ups in India: what lessons can we learn from the
past?”, Economic and Political Weekly, Vol. 50 No. 12, pp. 56-63.
Takhteyev, Y., Gruzd, A. and Wellman, B. (2012), “Geography of Twitter networks”, Social Networks,
Vol. 34 No. 1, pp. 73-81.
Tan, K.H., Ji, G., Lim, C.P. and Tseng, M.L. (2017), “Using big data to make better decisions in the digital
economy”, International Journal of Production Research, Vol. 55 No. 17, pp. 4998-5000.
Wang, Y., Liu, J., Qu, J., Huang, Y., Chen, J. and Feng, X. (2014), “Hashtag graph based topic model for tweet
mining”, Proceedings of the 2014 IEEE International Conference on Data Mining, pp. 1025-1030.
Wisdom, V. and Gupta, R. (2016), “An introduction to Twitter data analysis in python”, available at:
www.researchgate.net/publication/308371781 (accessed September 28, 2019).
Wong, P.K., Ho, Y.P. and Autio, E. (2005), “Entrepreneurship, innovation and economic growth:
evidence from GEM data”, Small Business Economics, Vol. 24 No. 3, pp. 335-350.
Xiang, Z., Schwartz, Z., Gerdes, J.H. and Uysal, M. (2015), “What can big data and text analytics tell us
about hotel guest experience and satisfaction?”, International Journal of Hospitality
Management, Vol. 44 No. 1, pp. 120-130.
Further reading Analyzing the
Arora, A., Bansal, S., Kandpal, C., Aswani, R. and Dwivedi, Y. (2019), “Measuring social media startup
influencer index- insights from Facebook, Twitter and Instagram”, Journal of Retailing and ecosystem of
Consumer Services, Vol. 49 No. C, pp. 86-101.
Grover, P., Kar, A.K., Dwivedi, Y.K. and Janssen, M. (2019), “Polarization and acculturation in US
India
Election 2016 outcomes – can twitter analytics predict changes in voting preferences”,
Technological Forecasting and Social Change, Vol. 145 No. C, pp. 438-460.
281
Li, B., Dittmore, S.W., Scott, O.K.M., Lo, W.J. and Stokowski, S. (2019), “Why we follow: examining
motivational differences in following sport organizations on Twitter and Weibo”, Sport
Management Review, Vol. 22 No. 3, pp. 335-347.
Li, X., Xie, Q., Jiang, J., Zhou, Y. and Huang, L. (2019), “Identifying and monitoring the development trends
of emerging technologies using patent analysis and Twitter data mining: the case of Perovskite
solar cell technology”, Technological Forecasting and Social Change, Vol. 146 No. C, pp. 687-705.
Robinson, A.C., Savelyev, A., Pezanowski, S. and MacEachren, A.M. (2013), “Understanding the utility
of geospatial information in social media”, Proceedings of the 10th International ISCRAM
Conference, Baden-Baden, pp. 918-922.
Xiong, Y., Cho, M. and Boatwright, B. (2019), “Hashtag activism and message frames among social
movement organizations: semantic network analysis and thematic analysis of Twitter during
the #MeToo movement”, Public Relations Review, Vol. 45 No. 1, pp. 10-23.
For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com