Data Driven Decision Making in Digital Entrepreneurship

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/361778286
data-driven-decision-making-in-digital-entrepreneurship
Article in International Journal of Industrial and Systems Engineering · July 2022
CITATIONS READS
0 95
1 author:
Abeba N. Turi
University Canada West
9 PUBLICATIONS 15 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Digital Footprints View project
All content following this page was uploaded by Abeba N. Turi on 05 July 2022.
The user has requested enhancement of the downloaded file.

World Academy of Science, Engineering and Technology
International Journal of Industrial and Systems Engineering
Vol:16, No:4, 2022
Data-Driven Decision-Making in Digital

Entrepreneurship
Abeba Nigussie Turi, Xiangming Samuel Li
1
to market penetration, and seeking customer validation in the
Abstract—Data-driven business models are more typical for short-run [3].
established businesses than early-stage startups that strive to penetrate In this paper, by focusing on digital entrepreneurial practices,
a market. This paper provided an extensive discussion on the principles we will flesh out strategies to monetize the primary or
of data analytics for early-stage digital entrepreneurial businesses.
secondary chain of data that businesses access in a way that
Here, we developed data-driven decision-making (DDDM) framework
that applies to startups prone to multifaceted barriers in the form of boosts their potential growth opportunities. By identifying the
poor data access, technical and financial constraints, to state some. The value of data and examining the potential to exploit such values
startup DDDM framework proposed in this paper is novel in its form through smart business models, the paper presents a framework
Open Science Index, Industrial and Systems Engineering Vol:16, No:4, 2022 publications.waset.org/10012523/pdf
encompassing startup data analytics enablers and metrics aligning with to unlock such potential in one of the digital economy's
startups' business models ranging from customer-centric product underutilized resources. Besides, by taking use cases and
development to servitization which is the future of modern digital
highlighting the respective on-demand business intelligence
entrepreneurship.
tools, the paper will flesh out the standard practices and
Keywords—Startup data analytics, data-driven decision-making, strategies to accelerate growth through DDDM. Further, it will
data acquisition, data generation, digital entrepreneurship. provide an extensive discussion on the potentials which open
data holds for startup businesses and strategies to utilize such
I. INTRODUCTION an open resource to boost business growth at a scale. Digital
blueprints, data valuation, smart business models, on-demand
A S a powertrain of the digital economy, data have proved to
be one of the competitive advantages that businesses
established on such resources excel. Accordingly, data science
business intelligence tools (services and software for processing
digital footprints), DDDM, and data-centric value creation are
and practices have evolved, incorporating the dynamic business the main topics covered in this paper.
needs over time. With the increasing stock of big data from the The rest of this paper is organized as follows. Section II
digital footprints of users, integrating and exploiting the presents related works and highlights the research gap we
benefits of such information in decision making is proven to intend to fill in this work; Section III briefly presents some data
facilitate decision making and unlock business potentials [4]. analytics tools, Section IV presents the practice of data-
Moreover, open data initiatives by the data hoarding tech powered decision making in the startups’ context, Section V
companies, like Microsoft, can help businesses utilize the stock presents data source, challenges, and opportunities in relation
of digital footprints they hoard in their data empire [10]. With to startup data analytics, Section VI highlights the core data
the fast pace of digitization, reinforced with COVID-19 and analytics application areas for startups; and, Section VII
multiple other factors, by 2024, global data created, captured, presents data analytics strategy and a conceptual framework to
copied, and consumed are predicted to reach 149 zettabytes DDDM. Lastly, Section VIII concludes our key findings with
from as low as 12.5 zettabytes a decade ago in 2014 [11]. In a direction for future works.
world where such data can be used at a near-zero marginal cost,
entrepreneurial DDDM strategies and robust analytics are vital. II. RELATED WORKS
There is evidence that using data in decision-making has a Several studies have been conducted to unlock
significant cost reduction effect, specifically in relation to entrepreneurial business potential and identify key challenges
operational efficiency [1]. such businesses observe [3], [5]. With this is the potential of
Startups face multilayer problems in today's dynamic digital footprints generating big data for predictive and visual
business environment. Accordingly, informed decisions of this analytics [6], [13]. Yet, there is evidence that early-stage
form in an entrepreneurial course of actions rather than mere startups face the challenges of skills, capital, market
reliance on intuition can improve business performance. uncertainty, technological uncertainty, time management, and
However, in most cases, startups fail to incorporate DDDM in privacy issues to generate values that contribute to poor startup
their core business lines. Some of the factors identified to limit success rates [3], [5], [12]. Hartmann et al. [6] provided an
startups from using data analytics include a mere focus on extensive discussion on the taxonomy of data-driven business
product development, team expertise in data analytics, amount models used by startups, encompassing seven key business
of data at stock, a mere focus on core product delivery, attention activities: free data collection and aggregation, analytics-as-a-
Abeba Nigussie Turi and Xiangming Samuel Li are with University Canada
West, Canada (e-mail: [email protected], [email protected]).
International Scholarly and Scientific Research & Innovation 16(4) 2022 122 ISNI:0000000091950263
Vol:16, No:4, 2022
service, data generation and analysis, free data knowledge the startup to trace users' behavioral patterns of the app usage
discovery, data-aggregation as-a-service, and multi-source data by tracking user data. Such an informed product decision
mash-up and analysis (see also [7]). Kandel et al. [7] identified allowed the company to increase its customer retention by about
an enterprise's organizational feature as one of the key factors 50%. See also [12] for the feedback loops in the data-driven
defining enterprise data analytics' efficiency. economy applied to the prosumers’ blueprint towards data
The literature in the field is limited in providing theories and monetization in the supply chain.
practices for early-stage startup data analytics. This paper will
flesh out the principles of DDDM in the startup context. IV. DATA SOURCE FOR STARTUPS: CHALLENGES AND
Besides, we extend the work on startup data analytics by OPPORTUNITIES
providing a general conceptual framework of DDDM that Data analytics applies to the notion of enabling technologies
applies to digital entrepreneurs. that capture relevant information for economic and non-
economic value generation purposes. Such information can be
III. THE PRACTICE OF ENTREPRENEURIAL DATA-POWERED revealed from massive volumes of data, big data, coming from
DECISION-MAKING the information society's digital blueprints. The massive
Data analytics can play a crucial role at different stages and blueprint of human-machine interaction comes from
scales of business operation, starting from the inception stage transactional mobile and internet records, user-generated
of a product life cycle. Before launching a new business or content across social media, blogs, marketplaces, and various
introducing a new product, big data can be used to make digital platforms, ad-hoc content collected through sensor
informed decisions. This will allow product testing with an networks, IoT, and other channels.
accurate picture before getting into the actual market. Relying Many digital footprints left by users lay the basis for
on the core bubbles or a small circle of people to test a product structured and unstructured data, which businesses can harness
might create some bias when bringing the product to general in their best interest. In most cases, specific data that apply to a
users. The digital economy has brought the world together into given business come from the data businesses capture from
a virtual planet and this has made it easier for startups to have their online store visitors across different sessions, platforms,
insights and tests to their new products from the crowd. For and domains. This treasure contains potential customers of the
example, location analytics can help identify an appropriate product, which businesses pull for their marketing strategies.
business location that fits your business lines. Another example However, several barriers limit entrepreneurs from using these
is property valuations of the real estate industry. A stock of data potentials [3].
enables the analysis of markets and identifies non-economic One of the main challenges early-stage digital entrepreneurs
and economic signals and trends of the real estate industry. face is the limit in the amount of data they have in stock. With
To help us look into the practice of data-powered businesses this limit in place, it will be difficult for startups to rely on
by startups, we picked the clients from the data analytics internal data to make business decisions. In this regard, it is
company, Mixpanel. The company tracks web and mobile ideal for such startups to use relevant external data and tools
usage that allows targeted product development. Here, we will that apply to their business needs. To mention one, leveraging
pick and look at two sample companies that rely on its tool to social data mining of third parties can help in targeted ads. For
make informed decisions using data. example, data exchange platforms, vendors, and tools like
One of companies is the transaction management and Ocean Protocol, Gnip, Immersive Labs, Data Sift, BDEX,
electronic Signature tech company, Docusign, which uses data- Quantcast, and iGrant.io facilitate and enable such real-time
analytic tools from Mixpanel to boost its global reach. The data for customer-centric businesses. Similar to these are the
system allows Docusign to enhance its basic and custom open-source big data analysis platforms and tools like Hadoop,
product metrics like customer retention, the number of MapReduce, GridGain, HPCC Systems, and Storm that have a
documents customers successfully complete, and upgrades. For significant innovation diffusion. Another way to excel on this
example, Mixpanel behavioral analytics enhanced Docusign’s is to make proper use of open-source data available on the web.
paid upgrades by exposing some premium features that An equally important aspect to discuss here is data optimization
incentivize users for the conversion. The analytical tool practice. Optimization of the data businesses capture from
increased upgrade conversions by 5%, bringing about 130,000 online store visitors across different sessions, platforms and
new users per day [14]. domains can allow optimal business performance through
Kast is another company using Mixpanel's data analytics tool improved business operations and accurate decision making.
in its service provision of a "virtual living room". The company
uses content sharing technology to allow the traditional V. DATA ANALYTICS APPLICATION AREAS FOR DIGITAL
physical human interaction experience (e.g., watching movies ENTREPRENEURSHIP
or playing games in a shared space). The startup uses Mixpanel Data analytics, including data visualization and machine
for its DDDM to understand its users. For example, the data learning, can empower digital entrepreneurs to acquire new
analytics took user feedback to identify "Power Users" (active customers, create new value chains, and improve operational
members in the last 15 days and a minimum of 30 time- efficiency. Operational efficiency is one of the forefront value-
frequency of partying through the app). This feedback is used add coming to businesses with data analytics. The following
for product and feature developments. The analytics allowed
Vol:16, No:4, 2022
attempts illustrate some of the application areas of data or Amazon recommends related products based on transaction
analytics for startups. records and other users' digital footprints.
A. New Product Design F. Customer Relationship Management through Chatbots
Creating personalized new products for each customer can be One of the powerful data analytics applications is online
challenging. However, the digital footprint of users has enabled chatbots powered by artificial neural networks (ANN) as a deep
this through aggregated data analytics. For example, Hellofresh learning technique. The chatbot is a conventional software
(a meal-kit company) consumer data streams are massive, agent to communicate with human users through a natural
aggregated from social media, emails, and user impressions on language based on ANN. Chatbots have the advantage of being
their websites, allowing customized fresh ingredients and self-service, including 24x7 full online care, global languages,
recipes for individual customers. Here, a special case of digital cost-saving for call centers, as well as interactive and intelligent
entrepreneurship takes the form of prosumers in which user experiences. Chatbots are expected to help enterprises to
consumers of digitally enabled products and services are the save about $8 billion annually by 2022 in customer-support
producers themselves, thus adding to new product development costs, compared to only the $20 million estimated saving in
and designs. 2017 [13]. One of the common chatbots is Amelia by IPSoft, as
the market-leading digital employee and conversational AI
B. Online Platform Building
application. Chatbots can offer high-volume and high-impact
Data analytics tools can help digital entrepreneurs develop a customer services such as Frequently Asked Questions (FAQs),
professional website without HTML, Web 3.0, CSS, and service ticket management, account management, and customer
JavaScript. For example, AI-powered website builders like care.
Wit.ai, Wix, and Dialogflow. Content generation tools such as
Articoolo, Wordsmith, and Quil can analyze and learn from G. Fraud Prevention
other web data and generate human-like content to help startups Data analytics has been extensively applied for business
engage online customers more. Besides, the Google Analytics fraud detection and prevention. Many fraud detections use UPL
tool allows to monitor, analyze and improve startup web classification models, such as decision tree, logistics regression,
business performances. SVM, and even ANN deep learning. In business practice,
various unsupervised learning methods like profiling,
C. Reputation Management
clustering, anomaly detection, and co-occurrence association
Trust, as repeatedly coined, is the currency of the digital are also used for different fraud detection and prevention.
economic system. There are diverse rating and reputation
schemes in the P2P business models. For example, Li [9] H. Data Analytics for IoT Machine-to-Machine (M2M)
applied text analytics and machine learning approaches, Applications
including Naive Bayes, logistic regression, support vector M2M-based IoT data differ from traditional big data in terms
machine (SVM), and long short-term memory (LSTM-AI), to of the growing "ocean" volume, heterogeneous formats,
more than three million online text reviews collected from the embedded noises, and real-time attention needed like medical
Airbnb platform. A large knowledge-based label (CORPUS) is care and animal surveillance. The wide-spreading 5G networks
built, and an innovative text review score (TRS) is constructed and IoTs applications will trigger a new round of Industry 4.0
as a new online trust measure. Digital entrepreneurs can adopt innovations, including smart homes, smart cities, smart
a similar approach or take advantage of this sizeable online agriculture, and smart globe, or called Smart Xs (Smart
review CORPUS to rapidly build up new online reputation Everything). Meanwhile, this brings a huge potential for new
systems. entrepreneurs to create various M2M-based IoTs applications
using advanced data analytics tools.
D. New Product Market Launch
Digital startups can use data analytics to find an optimal VI. DATA ANALYTICS STRATEGY AND CONCEPTUAL
marketing strategy to launch their new products successfully. FRAMEWORK FOR DIGITAL ENTREPRENEURSHIP
For example, various relevant data points from previous market
Based on the startup metrics of Kemell et al. [8] and the
launches can be collected from open data sources for similar
Pirate Metrics, or “AARRR” of McClure [10] and Hartmann et
products in a specific market; and data visualization and
al. [6] taxonomy of data-driven business models for startups,
analytics tools can be employed to find the effectiveness of each
we developed a conceptual framework for the entrepreneurial
marketing promotion and product pricing.
DDDM process.
E. Digital Marketing and Customer Profiling Here, we aggregate the foundation for the entrepreneurial
Data analytics can empower digital startups to know more DDDM under two main pillars: (1) The enablers which allow
about their customers regarding product features and pricing. the digital entrepreneur to make data-driven decisions, and (2)
Personalization is vital to delivering a satisfactory customer the business metrics which need to be identified to fit the
experience. Data analytics enables an efficient, personalized decision-making process with the core business needs and
user experience enhancing customer conversion and retention. goals. Kemell et al. [8] provided key startup metrics for tech
For example, Netflix traces users' movie or TV show preference entrepreneurs, including engagement metrics that measure how
often people engage with a product. For example, for an
Vol:16, No:4, 2022
uberfication business model, Daily Active Users (DAU) and platform (including analytical tools, cloud storage technology,
Monthly Active Users (MAU) metrics measure customer security and privacy-preserving solution) and data (acquisition-
engagement with the app. Another metric is the churn rate external data-capture from data exchange platforms, data
which measures customer stickiness to a product, like the rate analysis platforms, open-source like social data for targeted ads
of cancelation or non-renewal in a subscription-based business and real-time data for customer-centric business, keyword
model. As is typical for startups to get customer endorsement performance and ranking, search behavior, page visits, tweets,
and validation, customer engagement, product reviews, and likes, share comments, or generation- internal data capture from
customer feedback are essential metrics startups focus on for tracking (potential) customer or user data, transaction records,
product development and enhanced services. As startups automated user tracking, customer feedback, customer
consider scaling their business, equally important are the engagement, business operations).
growth metrics. The second pillar of business metrics is the performance
Under the first pillar, by extending the key big data analytics metrics under (i) Business and financial metrics, (ii) User/
enablers indicated by Behl et al. [2] and Kandel et al. [5], we customer metrics that capture behavioral patterns of (potential)
identified four main enablers which we identified them as customers allowing for customer-centric business, (iii) Service
people (encompassing the technical skills and entrepreneurial metrics that capture the process and product metrics, and (iv)
leadership a startup owns), startup's organizational features and Social media metrics as a measure of business visibility for
process (referring to the startup's internal business process, demand-pull.
relationships and diversity of data sources), infrastructure and
Fig. 1 Entrepreneurial DDDM Process: A Conceptual Framework
A data pool from diverse sources allows the startup to draw Such courses of action include operational efficiency,
business insights through analytics and visualization using augmentation of value proposition and differentiation for
analytical tools in the decision process. This will feed into the competitive advantages, and ultimately attaining business
entrepreneurial business model based on the revenue models sustainability through robust data-supported measures
defining the startups' potential and actual business lines. Here, prescriptive analytics. Fig. 1 depicts a conceptual framework
the analytics can support customer-centric value-add; for entrepreneurial DDDM. Summing up, based on the
servitization to build new revenue streams through service literature and best practices in data analytics, the main steps for
business lines (demand-pull), business reengineering; product entrepreneurial data analytics are discussed as follows:
development; process-oriented value-add (tech-push and
A. Business Understanding
automation); or an overall business transformation and
reengineering. Business analytics and visualization capture Digital entrepreneurs should start with compelling business
value from data and integrate data-powered decisions into the problems in mind and look for robust data analytics tools to
business strategy. Hence, we identified the major courses of solve.
action in the DDDM process.
Vol:16, No:4, 2022
B. Data Collection maximize the return on per customer. A new fraud detection
At this stage, we identify internal or external data sources that model can be trained and integrated into a customer relationship
will give an insight into the business and fuel growth for startup. management (CRM) system to monitor accounts and identify
In line with the business rules, another important thing to potential transactional fraud to minimize business loss.
consider is to plan what you intend to acquire from the data. It
should be noted that new goals can arise as one observes the VII. CONCLUSION
trends and patterns in the data. Digital entrepreneurs often fail to incorporate DDDM in their
businesses during their startup incubation, contributing to a
C. Data Understanding
very high startup business failure rate. This work fleshed out
It is crucial to understand the strengths and weaknesses of the the core state-of-the-art data analytics tools and application
data; since there is no exact match between the data and targeted areas for early-stage startups, including new product design,
problems to be solved. Because a historical dataset is often online platform building, digital marketing, customer care,
composed of unnecessary features which do not serve the operational efficiency, and IoT-based machine-to-machine
purpose, it is vital that the analyst goes through the data and automation. Further, we offer practical strategies for
filter the required information in order to draw insight out of it. entrepreneurial data-powered decision-making, such as data
For example, a client database, product marketing response acquisition in the face of limited internal data generation. The
data, and business transactional records cover different
conceptual framework for the entrepreneurial DDDM process

intersectional populations with varying degrees of reliability. developed in this work relies on two key pillars of enablers that
D. Data Organization allow the digital entrepreneur to make data-driven decisions
This phase often is carried out with data understanding, and the entrepreneurial business metrics that fit the decision-
where the data are manipulated and processed into proper forms making process with core business needs and goals. In this
for data visualization and data analytics tools to yield valuable regard, future work direction will be to look into the dynamics
results. At this stage, the data are organized to form a of data analytics techniques and startup business performance
description and/or identify themes and trends in user behavior. under the agile digital entrepreneurship environment. Another
Examples include converting raw data to a tabular format with avenue of research related to this is the empirical evidence and
rows and columns, dealing with missing data, changing data quantitative analysis of data-powered decision-making in the
types, removing excess data, data value normalization, etc. startup business context, including the sectoral analysis.
Besides, data organization includes cleaning the data and
getting an insight with further considerations for data ACKNOWLEDGMENT
compatibility issues in the data organization process. One of the Funding by the University Canada West is gratefully
principal reasons for this is that as the startup grows, the data acknowledged.
analytics will be replicated at a scale as the business and data
analytics systems evolve. REFERENCES
[1] Bean R. (2017) How Companies Say They’re Using Big Data in Harvard
E. Modeling Business Review- Analytics And Data Science. Retrieved February 21,
It is the primary phase where data analytics techniques are 2022, from https://hbr.org/2017/04/how-companies-say-theyre-using-
big-data
applied to the data, including visualization, clustering,
[2] Behl, A., Dutta, P., Lessmann, S., Dwivedi, Y. K., & Kar, S. (2019). A
classification, and deep learning (see Section IV). The model conceptual framework for the adoption of big data analytics by e-
has to specify key instruments and metrics on which the data commerce startups: a case-based approach. Information Systems and e-
analytics relies in order to address the core business problems, Business Management, 17(2), 285-318.
[3] Berg, V., Birkeland, J., Pappas, I. O., & Jaccheri, L. (2018). The Role of
see Pirate Metrics, or "AARRR" [10]. Data Analytics in Startup Companies: Exploring Challenges and Barriers.
In Conference on e-Business, e-Services and e-Society (pp. 205-216).
F. Evaluation Springer, Cham.
Digital entrepreneurs can employ different evaluation [4] Caulkins, J. P., Bao, Y., Davenport, S., Fahli, I., Guo, Y., Kinnard, K., ...
& Kilmer, B. (2018). Big data on a big new market: Insights from
metrics such as the confusion matrix and ROC/AUC curves to Washington State's legal cannabis market. International Journal of Drug
evaluate the model performance in terms of accuracy, precision, Policy, 57, 86-94.
recall, cross-validation, and overfitting capacity. Model [5] Giardino, C., Bajwa, S. S., Wang, X., & Abrahamsson, P. (2015, May).
Key challenges in early-stage software startups. In International
evaluation is a science rather than an "art". It includes a rigorous Conference on Agile Software Development (pp. 52-63). Springer, Cham.
assessment of various models to select a robust model for a [6] Hartmann, P. M., Zaki, M., Feldmann, N., & Neely, A. (2016). Capturing
successful digital business. value from big data–a taxonomy of data-driven business models used by
startup firms. International Journal of Operations & Production
G. Deployment Management.
[7] Kandel, S., Paepcke, A., Hellerstein, J. M., & Heer, J. (2012). Enterprise
Deployment is the final phase of putting data analytics to real data analysis and visualization: An interview study. IEEE Transactions on
use to achieve business results. It involves implementing the Visualization and Computer Graphics, 18(12), 2917-2926.
best-evaluated model from the previous step, such as predicting [8] Kemell, K. K., Wang, X., Nguyen-Duc, A., Grendus, J., Tuunanen, T., &
the likelihood of customer churn and sending a special offer to Abrahamsson, P. (2020). Startup Metrics That Tech Entrepreneurs Need
to Know. In Fundamentals of Software Startups (pp. 111-127). Springer,
highly possible "churn" customers to retain them in business to Cham.
Vol:16, No:4, 2022
[9] Li SX (2021), Short or Long Review? Text Analytics and Machine

Learning Approaches to Online Reputation, International Journal of
Business and Management Research (IJBMR), 9 (1), pp 28-40, e-ISSN:
2347-4696.
[10] McClure, D. (2007). Startup metrics for pirates: AARRR. Startup Metrics
for Product Marketing & Product Management.
[11] NCES. (n.d.). Volume of data/information created, captured, copied, and
consumed worldwide from 2010 to 2025(in zettabytes). In Statista -
Telecommunications. Retrieved February 21, 2022, from
https://www.statista.com/statistics/871513/worldwide-data-created/
[12] Oliva, F. L., & Kotabe, M. (2019). Barriers, practices, methods and
knowledge management tools in startups. Journal of knowledge
management.
[13] Turi, A. N. Technologies for Modern Digital Entrepreneurship
Understanding Emerging Tech at the Cutting-Edge of the Web 3.0
Economy.
[14] Mixpanel- service performance and efficiency available at,
https://mixpanel.com/customers/docusign-accelerates-growth/ Retrieved
Apri1 2, 2022

Data Driven Decision Making in Digital Entrepreneurship

Uploaded by

Copyright:

Available Formats

Data Driven Decision Making in Digital Entrepreneurship

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Data Driven Decision Making in Digital Entrepreneurship

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Article in International Journal of Industrial and Systems Engineering · July 2022

Digital Footprints View project

The user has requested enhancement of the downloaded file.

Data-Driven Decision-Making in Digital

relationships and diversity of data sources), infrastructure and

Fig. 1 Entrepreneurial DDDM Process: A Conceptual Framework

conceptual framework for the entrepreneurial DDDM process

[9] Li SX (2021), Short or Long Review? Text Analytics and Machine

You might also like