Journal of Business Research 123 (2021) 423–437

Contents lists available at ScienceDirect

Journal of Business Research

journal homepage: [Link]/locate/jbusres

There’s an app for that! understanding the drivers of mobile

application downloads
Zeynep Aydin Gokgoz a, *, M. Berk Ataman b, Gerrit H. van Bruggen a
a
Rotterdam School of Management, Erasmus University, Burgemeester Oudlaan 50, 3062 PA Rotterdam, the Netherlands
b
College of Administrative Sciences and Economics, Koc University, Rumelifeneri Yolu 34450, Sarıyer, Istanbul, Turkey

A R T I C L E I N F O A B S T R A C T

Keywords: Since its emergence, mobile applications market has been attracting the attention of all kinds of businesses due to
Mobile marketing the lucrative opportunities apps offer and the market’s low barriers to entry. Yet, in this crowded space, only a
Mobile apps small portion of apps can survive. Using a unique data set of 979 newly released applications, acquired from a
Apple
leading mobile analytics company and enriched with publicly available data, the authors shed light on the factors
Downloads
Online word of mouth
associated with app downloads during an app’s first year of existence. Results from time-varying-parameter
Updates models estimated separately for free and paid apps reveal that gaining traction with users shortly after release
seems critical and that app platform owners can be very influential in these early days. However, as apps mature,
affecting the number of downloads becomes increasingly more difficult. The findings add new insights to the
growing literature on apps and provide practical implications for their developers.

1. Introduction Apps offer businesses, big and small, an opportunity to connect with on-
the-go consumers in various ways. First, apps can become an integral –
Since its emergence in July 2008, the mobile application1 (app or even core – part of a firm’s business model and operate as an addi­
hereafter) space has been growing at an astonishing rate. The Apple App tional channel or a platform generating most of the traffic and (ad)
Store has gone from just 500 apps in July 2008, the month of its launch, revenues (e.g., Facebook, Twitter, Amazon). Some firms start mobile
to 1 million apps in the fall of 2013 and reportedly reached 2 million first and operate almost entirely through their apps (e.g., Uber, Insta­
apps by 2018 (Apple Insider, 2018). Its competitor, the Google Play gram). Second, with the increasing importance of apps in consumers’
Store, offered 2.5 million different apps in 2018 (AppBrain, 2019). daily lives, apps provide firms a new medium for advertising and a
Global downloads from app stores exceeded 194 billion in 2018. Total platform to create and maintain brand engagement (e.g., Ruffles AmiGo,
app revenues, including revenues from paid downloads, in-app pur­ IKEA Place).
chases and in-app subscriptions, hit $101 billion in 2018, up 75% from This emerging market with low entry barriers and lucrative business
its level in 2016 (AppAnnie, 2019).2 As apps become increasingly more opportunities continues to attract businesses diverging from individual
popular among consumers, worldwide app store revenues are forecasted developers to well-established brands and the app market has become
to reach $156.5 billion in 2022 (AppAnnie, 2018). increasingly crowded over the years. However, only a small portion of
The growth of the app market is not surprising, because continuing apps can gain traction with mobile phone users. In 2018, 74% of all apps
advances in wireless technologies and the growing smartphone pene­ were downloaded less than 1000 times, up from 70% in 2014. In
tration have provided businesses with a new channel with unique fea­ contrast, 80% of the downloads in 2019, up from 76% in 2014, are
tures to approach customers (e.g., accessibility at anytime and generated by the top 1% of publishers in the Apple App Store or the
anywhere, customization at a granular level, and location sensitivity). Google Play Store (SensorTower, 2019). Moreover, the average app

* Corresponding author.
E-mail addresses: zaydin@[Link] (Z. Aydin Gokgoz), bataman@[Link] (M.B. Ataman), gbruggen@[Link] (G.H. van Bruggen).
1
Apps are dedicated software applications that run on small, handheld devices such as smart phones, tablets and notebooks.
2
The primary source of revenue depends on the business model apps utilize: free, paid, or freemium. Free apps are downloaded at zero cost and revenues are
generated through in-app advertisements and purchases. Paid apps are downloaded at a cost and may also offer additional features for a fee (mostly < $5). Lastly,
developers can use the freemium model and launch both free and paid versions of the app. The free app encourages trial and promotes the paid version, which comes
with extended functionalities.

[Link]
Received 30 December 2019; Received in revised form 29 September 2020; Accepted 2 October 2020
Available online 15 October 2020
0148-2963/© 2021 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license ([Link]
Z. Aydin Gokgoz et al. Journal of Business Research 123 (2021) 423–437

loses 77% of its daily active users within the first three days after Pedersen, & Thorbjørnsen, 2005), mobile commerce (Shankar, Ven­
download, whereas top apps have significantly higher retention rates katesh, Hofacker, & Naik, 2010), usability of and user experience with
(Chen, 2018). Consequently, the majority of apps don’t generate the mobile devices (Zhang & Adipat, 2005), mobile usage behavior (Ghose,
anticipated revenues, and some are even withdrawn from the market Goldfarb, & Han, 2013), and mobile marketing (Shankar & Balasu­
after a while. bramanian, 2009), research related to app markets is still in its infancy.
Apps such as Everpix (a high quality app that sorts, organizes, and Research on apps can be discussed under two main headings: the ante­
cloud-stores photos) and Vine (an app for sharing short videos) that cedents of app adoption and the consequences of app introduction.
satisfy unique customer needs in the first place, have learnt the impor­ Starting with the latter, the effectiveness of this medium has
tance of having the right business model, adapting to the changing particularly been of interest to researchers. Current research shows the
landscape of the market place and to evolving consumer needs, and the positive effects of app introduction/adoption on brand attitudes and
power of marketing the hard way. Both apps made a promising entrance purchase intentions (Bellman, Potter, Treleaven-Hassard, Robinson, &
to the market and were shutdown later with great disappointment and Varan, 2011; Mclean, Osei-Frimpong, Khalid, & Marriott, 2020),
despair (Smart Insights, 2019). Despite these challenges, firms continue cognitive and affective brand responses (Van Noort and Van Reijmersdal
developing new apps. In order to increase the success rates of apps, it is (2019)), subsequent purchases (Liu, Lobschat, Verhoef, & Zhao, 2019;
important to understand the factors that are associated with app Van Heerde, Dinner, & Neslin, 2019), and even firm value (Boyd,
downloads, especially in the early stages of an app’s lifecycle. In this Kannan, & Slotegraaf, 2019; Cao, Liu, & Cao, 2018; Gill, Sridhar, &
paper, we aim to contribute to developing this understanding by Grewal, 2017).
studying factors that are related to app adoption. As to the antecedents of app adoption, previous research has
Based on a review of related studies in the literature and a detailed advanced our understanding of the impact of user characteristics (Kim,
analysis of a user’s decision journey in the path to app adoption, we Kim, Choi, & Trivedi, 2017), app characteristics (Schulze, Schöler, &
identify a set of factors potentially associated with downloads. The Skiera, 2014), app pricing (Arora, Hofstede, & Mahajan, 2017; Carrare,
literature review revealed a set of variables under the developer’s con­ 2012; Ghose & Han, 2014; Kübler, Pauwels, Yildirim, & Fandrich,
trol and variables reflecting current users’ views, while the analysis of a 2018), app updates (Ghose & Han, 2014; Kübler et al., 2018), other
user’s decision journey revealed additional factors under the platform users’ experiences (Ghose & Han, 2014; Kübler et al., 2018), and, in the
owner’s control, whose effects are yet unexplored. Also unexplored in broader mobile eco-system, integration, ownership, and novelty of the
the literature are whether and how this comprehensive set of variables’ apps (Van den Ende, Jaspers, & Rijsdijk, 2013). Our research is in line
effects differ across app types (i.e., free and paid apps) and, more with the empirical studies focusing on the antecedents of app adoption
importantly, vary over time in the first year following an app’s release. and differs from them in the following respects (see Table 1).
To investigate how all variables jointly affect downloads over the first First, our study differs from other empirical studies in terms of the set
year following an app’s release, we assemble a unique data set by of drivers influencing downloads. Decisions and actions of three players in
combining daily-level transactional data for 979 apps obtained from one the app market, as suggested by Hao, Li, Tan, and Xu (2011), have the
of the foremost mobile analytics companies with publicly available data potential to drive downloads. These are app developers, app users, and
from the Apple App Store. We observe each app since its release and app platform owners. So far, research sheds light on the important roles
study the evolution of the impact of the factors on downloads over time developers and users play in app performance. We add to this knowledge
separately for free and paid apps.3 By doing so, we can offer insights by considering the unexplored role of app platform owners. Platform-
customized to an app’s business type and time on market. We also controlled variables impact the visibility and discoverability of apps
explore the sensitivity of our findings in other pertinent sub-samples of and have the potential to increase downloads to a great extent (see
apps. Section 3 for more details). Specifically, we study the impacts of three
Our findings show that the mere appearance in top apps charts have types of updates, price, and discounting decisions by developers, word-
the largest effect on downloads of free and paid apps alike, followed by of-mouth activity (valence and volume) by users, and appearance on
appearing on a top featured list, especially for paid apps. These results featured lists and position in top apps charts by platform-owners on
highlight the influence that platform owners have on users. Updates downloads. Not having to use ranking as a proxy for success, because we
released by developers have positive effects on downloads and their have access to download numbers, enables us to quantify the additional
effects increase proportional to the amount of improvement. Further effect of merely appearing in top apps charts on downloads. In sum, our
investigation of these effects’ evolutions reveals that a majority of the selection of variables is more comprehensive than up to date research,
factors matter especially early-on following an app’s release. yet limited by the availability and extractability of the data.
This paper continues as follows. In Section 2, we review the literature Second, our study differs from existing research in terms of the nature
on apps and discuss how this study adds to the current knowledge. In and composition of apps under investigation. Previous studies almost
Section 3, we develop the conceptual framework for our study and a exclusively use ‘being ranked in a Top Apps Chart’ as one of the sam­
series of expectations for the explored relationships. Since the field of pling criteria. The use of such a sampling criterion may introduce suc­
mobile marketing is in its early stages and theory development so far cess bias, as it takes quite a high number of downloads to enter these
seems non-existent or at least scarce, we refrain from developing formal charts.4 Though these studies have advanced our understanding of the
hypotheses. In Section 4, we describe the unique data set we have drivers of relatively mature apps’ downloads, the problem of generating
compiled for the sake of this study, the specification of the model, and downloads is more acute for newly released apps. As our results show,
the operationalization of our variables of interest. Sections 5 and 6, download performance early-on is critical for overall downloads. Our
respectively, present the results of our analyses and conclusions with the access to download figures allows us to study factors associated with a
ensuing implications. new app’s performance from its release date onwards independent of the
app’s ranking status. Moreover, we believe that including low-download
2. Research background generating apps as well as high-download generating apps in our sample
helps us develop a broader understanding of the app market dynamics.
Although research in marketing and human computer interaction
has advanced our knowledge of the mobile consumer (Nysveen,
4
For a few statistics on this, please see [Link]
mment-and-opinion/67142 (last accessed on 12/14/2019) and [Link]
3
We do not treat freemium as a third category because there are very few [Link]/aso-blog/infographic-number-of-downloads-to-reach-top
apps in our sample offering both free and paid versions. -rankings (last accessed on 12/14/2019).

424
Z. Aydin Gokgoz et al. Journal of Business Research 123 (2021) 423–437

Table 1
Comparison of empirical studies on App performance.
Study App Number Sampling Criteria Sampling App Performance Developer- User-side Platform- Contingency
Store of Apps Time Frame (Operationalization) controlled Variables controlled Factors
Variables Variables

Carrare Apple 912 Top 100 Apps 1/1/2009 – 6/ Sales Price – – –

(2012) (Free and Paid) 16/2009 (Inferred from rank and Update
(166 daily market shares)
obs. per app)
Ghose Google 2624 Top 400 Apps 9/5/2012 – 1/ Demand Price Valence – Consumer
and Apple 4706 (Free and Paid) 10/2013 (Estimated sales Update Volume Characteristics
Han (daily obs. for quantities)
(2014) 4 months)
Lee and Apple 7579 Top 300 Apps 12/2010 – Appearance, duration, Update Valence – –
Raghu (Free, Paid, Grossing) 09/2011 and number of apps by Discount Volume
(2014) (39 weekly developer in Top Charts
obs. per app)
Kübler Apple 20 Ranked in Top 100 paid 6/5/2011 – 3/ Popularity Price Valence – Cultural,
et al. apps in at least 80% of 27/2012 (Sales rank data) Update Volume Economic,
(2018) 60 countries and (276 daily Dispersion Structural Factors
remains in this ranking obs. per app) and Category
during the observation
period
This Apple 979 Stratified random Release dates Downloads (Estimated by Update Type Valence Feature Lists Time and Business
Study sample from 40,000 ranging from a leading mobile Price Volume Top Apps Model
new free of paid apps 1/1/2012 to analytics company) Discount Charts
released between 1/1/ 5/31/2012
2012 and 5/31/2012 (365 daily
obs. per app)

Notes: The studies listed herein also control for app characteristics (e.g., app size, description length, etc.) and developer characteristics (e.g., number of previous
successful apps, number of categories in which the developer offers apps, etc.) among other things. We do not list these for brevity.

Finally, our study differs from extant literature in terms of its main relatively less certain about the need, and at times, more certain because
focus. Whereas past research mainly deals with the problem of esti­ s/he has heard about an app through offline word-of-mouth or other
mating app demand (from rankings), the impact of price and its varia­ channels. Depending on how certain the user is, s/he pursues either a
tion across cultural, economic, and structural factors, our goal is to browse path (i.e., browsing the app store navigated by the user inter­
develop an understanding of a comprehensive set of variables that are face) or a search path (i.e., searching for an app by typing in the search
related to downloads for different app types (i.e., free and paid) and, box). The search path is further divided into two inherently different
more importantly, whether and how the effects of these factors vary over types in terms of the specificity of the queries, indicating more refined
time in the first year following an app’s release on the market.5 To our variation around need uncertainty. These are navigational search (i.e.,
knowledge, our paper is the first to link all these variables to downloads searching with a specific app name, such as ‘Angry Birds’) and cate­
and investigate the evolution of their effects. gorical search (i.e., searching with generic phrases, such as ‘free
games’).6 Through the browse and the categorical search paths, the user
3. Conceptual framework and expectations arrives at pages listing several apps. We refer to this milestone in the
journey as app discovery (i.e., the user becomes aware of apps that may
To identify the drivers of app downloads, it is vital to consider the satisfy her/his need).
user’s decision journey that leads to app adoption (i.e., the decision to For users following the browse path and, to a lesser extent, the cat­
download) and factors that facilitate/hinder progression through egorical search path,7 the design and information display of the store’s
journey stages. In what follows, we first outline the user’s decision landing page as well as those of category landing pages will play a
journey, which is built upon the classic demand chain or purchase funnel critical role in app discovery. Prominently displayed on these pages are,
(Lavidge & Steiner, 1961) and is further modified with the specifics of featured lists and top apps charts. Although an app’s appearance in a
the app market, the design of the store, and the behaviors of users featured list or its position on a top apps chart is determined by un­
therein. We then discuss the sources of information users rely on while derlying app characteristics (e.g., design, uniqueness, business model
sequencing through the decision journey and identify the drivers of type, media coverage) and app performance (e.g., past revenues,
downloads. We conclude this section with a discussion on the evolution downloads, engagement, retention (Engström & Forsell, 2018)), these
of an app through its lifecycle and the unique challenges imposed by its lists are created by platform owners. For that reason, we refer to them as
business model to arrive at differential predictions for variables associ­ platform-controlled variables associated with downloads.
ated with downloads. Following app discovery, the user chooses which app(s) to evaluate
in detail (i.e., the decision to click on one of the apps in the list). We refer
3.1. The path to app adoption and variables affecting users’ decisions to this milestone as app consideration. Those conducting navigational
search are likely to transition in and out of the consideration stage
A user’s decision journey starts with the recognition of a need for an
app, which triggers an app search in the store. At times, the user is
6
51% of smartphone users in the U.S. learn about apps because their friends/
family are using them and 48% discover apps by browsing the app store
5
The set of factors associated with downloads change across app business (Google, 2016). According to Apple (2020), the search path drives majority of
models. Whereas boosting downloads by adjusting prices and offering discounts downloads with 65% and most of the search queries are branded (i.e.,
is possible for paid app developers, developers of free apps only have control navigational).
7
over the value proposition of the app. Given this structural difference, we The more generic the categorical search query is, the closer the resulting list
choose to explore the relationships separately for free and paid apps. becomes to lists from browsing top apps charts.

425
Z. Aydin Gokgoz et al. Journal of Business Research 123 (2021) 423–437

swiftly. However, as app stores list similar items alongside the app charts can attract greater attention to the app and boost downloads.
searched for, these users may also discover the mere presence of rival Studies trying to uncover the ranking algorithms of app stores and the
apps. relationship between rankings and downloads reveal interesting insights
Users decide which apps they would like to evaluate in detail based pertaining to appearances and positions of apps in these charts.
on the information available to them on the app list page. In addition to Comparing the effects of WOM metrics and app rankings for a data set of
the app’s icon, name, and position on the list, the only other pieces of 42 days in the Google Play Store, Engström and Forsell (2018) find that a
information available on these pages are the app’s average rating scores, 10-percentile increase in displayed rankings increases downloads by
the number of reviews, and price – determined by the developer. Ratings 20%. Carrare (2012), investigating the effect of current rank on future
and number of reviews reflect previous adopters’ views about the app demand based on a data set of 166 days of top 100 free and paid apps in
and correspond to the online word-of-mouth measures of valence and Apple App Store, finds that consumers’ willingness to pay is $4.50
volume (Dellarocas, 2003). Therefore, we refer to them as user-side var­ higher for a top ranked app compared to an unranked app and declines
iables associated with downloads. The findings in Colicev, Malshe, steeply as the ranking of an app drops. Carrare (2012) also discovers
Pauwels, and O’Connor (2018) support the notion that WoM volume natural breakpoints in rankings corresponding to top 5, top 25 and top
and valence are effective in the transition to the consideration stage. 50. Findings of Garg and Telang (2013) complement those of Carrare
App evaluation takes place on the app description page. These pages (2012): a top ranked app for iPhone (iPad) earns 95 (110) times more
show the app’s price, average rating score, the number of times it has revenue compared to a top 200 ranked app. Accordingly, we expect
been reviewed with an option to access individual ratings and reviews, appearing in the top ranks of these charts to speed up adoption, with
static or dynamic visual and verbal descriptions of the app, and infor­ more prominent positions being more strongly associated with
mation on what’s new in the most up-to-date version of the app with an downloads.
option to review update history. Using these pieces of information, the
user evaluates whether the app can satisfy her/his need and whether the 3.2.2. User-side variables
price s/he needs to pay, if any, for gaining access to the app is accept­ The impact of word-of-mouth on consumer decisions has increased
able. The user, then, decides whether or not to download the app. The with the emergence of online feedback mechanisms. Word-of-mouth has
decision to download terminates the journey, whereas the decision to been shown to be an important factor in determining the success of
not download may lead the user to return to earlier stages. experience goods (De Vany & Walls, 1996) as well as goods in other
In addition to the user-side variables (i.e., WoM valence and volume), industries (e.g., Anderson & Magruder, 2012; Chevalier & Mayzlin,
all other factors that facilitate the user’s progression to the journey’s end 2006; Dhar & Chang, 2009; Duan, Gu, & Whinston, 2011). As to the
stage are directly under the control of the developer. Accordingly, we impact of WoM on app performance, both valence and volume have
refer to them as developer-controlled variables associated with downloads been shown to have a positive impact on app demand (Ghose & Han,
and group them under the app’s value proposition, which the developer 2014; Hao et al., 2011; Kübler et al., 2018). Accordingly, we expect to
seeks to improve by means of updates, and its price (including dis­ find a relationship in the same direction.
counting, if any).8
In sum, variables under different app-market-players’ controls – 3.2.3. Developer-controlled variables
platform owners, users, and developers – influence a potential adopter’s Developer-controlled variables, especially the app’s value proposi­
decision to download an app. Platform-controlled variables are pri­ tion, are effective in sealing the deal for potential adopters. Though an
marily operational on early transitions in the journey, user-side vari­ app’s value proposition is determined prior to launch, updates serve as a
ables on mid- and to late-stage transitions, and developer-controlled tool for further development of the app. In fact, the dynamics of the app
variables largely on late-stage transitions. The user’s journey to app market puts pressure on developers to update apps often and on a reg­
adoption along with our conceptual framework is presented in Fig. 1. ular basis. Fortunately, the continuous feedback from app users provides
developers with the opportunity to offer customized and swift responses
and enjoy favorable response as a result (Aydin Gokgoz, Ataman, & Van
3.2. Expectations Bruggen, 2020).
Previous research agrees on the positive impact of updates on app
3.2.1. Platform-controlled variables performance: the demand is higher for apps that are regularly updated
Platform owners can create attention for apps through the featured (Carrare, 2012; Ghose & Han, 2014; Kübler et al., 2018). Accordingly,
lists they publish on the landing pages. Being featured helps more users we expect a positive relationship between updates and downloads.
discover an app in a crowded environment through its impact on visi­ Though developers may generate additional downloads by means of
bility. Holding all else constant, discovery by a larger group of users updates, we expect the nature of the update to matter. In some updates,
should boost download numbers. Though empirical research on the the developers add new features and functionalities to their apps –
antecedents of app adoption or the drivers of app performance has not referred to as major updates hereafter – with the goal of improving their
studied the effect of being featured, research in other domains shows a app’s value propositions. In others, they improve the existing features –
substantial effect on sales of highlighting a product in its category and referred to as intermediate updates hereafter – or implement develop­
featuring/displaying it in a prominent position (e.g., Blattberg, Briesch, ment tweaks and bug fixes – referred to as minor updates hereafter – to
& Fox, 1995). ensure effective and efficient delivery of the value proposition, respec­
Likewise, as browsing through top apps charts is a prominent way of tively. We expect major updates to have a greater impact on downloads
app discovery, appearance and the position of an app in one of these than minor updates.
Finally, developers of paid apps can influence downloads with price
changes and the discounts they offer. Unlike traditional markets where
8
When asked about how important various factors are when making a de­ regular price changes are relatively infrequent, experimenting with
cision about which app to download, smartphone users in the U.S. rank price
first with 85% (Top 2 Box) followed by privacy or security of information
(84%), how much they’ll use the app (71%), description (71%), memory used
(66%), reviews (61%), and ratings (60%) (Google, 2016). The factors listed
between price and WoM variables are directly related to the efficient and
effective delivery of the app’s value proposition. Colicev et al. (2018) finds that
social media metrics corresponding to WoM valence are strongly associated
with customer satisfaction.

426
Z. Aydin Gokgoz et al. Journal of Business Research 123 (2021) 423–437

Fig. 1. Conceptual Framework.

Notes: The grey circles show the journey stages of an individual decision maker on the path to app adoption, curved grey dashed arrows the transitions, straight grey
dashed arrows the factors most effective on the transitions, and thin light-grey dashed arrows the factors effective to a lesser extent on the transitions. The black solid
boxes and the variables listed therein show the relationships (black solid arrows) explored in this study.

different price points to arrive at the right one, especially in the early life
Table 2
cycle of the apps, is a common practice in this market.9 The app store
Expected effects of download drivers.
provides developers with the opportunity to move smoothly between
price points by allowing them to schedule price changes. Developers can Prediction Evolution

alter price points permanently as soon as they realize that they have Variable Free Paid
chosen a price point that is too high for their potential user base or they Apps Apps
can temporarily offer discounts to expand the user base. Platform-controlled Variables
The effects of prices and discounts on downloads have been inves­ Appearance on Featured ++ + Decrease over time
tigated in several studies. For instance, Kübler et al. (2018) find that the Lists
Position in Top App Charts Decrease over time
demand for apps is sensitive to prices and price sensitivity varies across
++ +
User-side Variables
countries with different economic and cultural backgrounds as well as Valence of WOM + ++ Decrease over time
app categories (e.g., games and non-games). Ghose and Han (2014) Volume of WOM + ++ Decrease over time
investigate the competition between Apple and Google stores and find Developer-controlled
Variables
that discounting increases app demand more in Google Play Store than
Updates + ++ Increase over time
Apple App Store. Accordingly, we expect to find a negative (positive) Price NA – Decrease over time
relationship between price (discounts) and downloads. Discount NA + Increase/Constant over
We summarize our expectations for the signs of the effects of all time
variables on downloads and how we expect these effects to vary over an Notes: ++ indicates that the association between a variable and downloads is
app’s lifecycle and across business models, discussed subsequently, in stronger for a specific app type compared to the other. As for the over-time effect
Table 2. of price on downloads, we expect a reduction in the magnitude of price elas­
ticity. Accordingly, a decrease over time means that price elasticity, which is
negative, moves closer to zero (i.e., demand becomes less sensitive to prices).

3.3. Changes over an App’s lifecycle and differences between business

models
9
See [Link] for more de­
tails. In our sample, we observe sufficient variation in regular price and dis­ Apps and their potential adopters experience changes over the life­
counting variables. Specifically, 50.4% of paid applications undergo at least one cycle and different conditions across app business models on three
regular price change over the first year of the app’s existence and more apps fronts: (1) the source and the amount of information available, (2) the
change regular prices in the first six months of the data. Moreover, 45.9% of all nature and the extent of risks perceived, and (3) the perceptions and
apps in the sample offer at least one discount.

427
Z. Aydin Gokgoz et al. Journal of Business Research 123 (2021) 423–437

expectations of the untapped potential. As a result, the strength of the greater importance on the value the app offers compared to free apps.
association between the variables and downloads can evolve over the Accordingly, we expect the effect of updates to increase over time, as
phases of an app’s lifecycle and be different for free vs. paid apps. they serve as a tool to expand the potential user base, and to be stronger
First, the source and the amount of information available vary over for paid apps than free apps, as users have higher expectations of the
time and across apps. Because the number of users who have down­ app.10 Moreover, consistent with the results in Simon (1979) and Bij­
loaded the app is likely to be lower in the early days of an app’s lifecycle molt, van Heerde, and Pieters (2005), we expect the magnitude of price
(i.e., low observability), the likelihood of discovering the app through elasticity to be larger in the early phases of an app’s lifecycle. In the later
channels other than the app store itself will be lower early on (Rogers, phases, as fewer and more attentive users with higher willingness to pay
2003). As an app matures and increasingly more users download it, would be on the market to find the app that satisfies their unique need,
potential adopters will gain access to more information and from various price may lose its importance. As for the over-time effects of discounts,
sources (e.g., offline WoM, press coverage, publicity, etc.). Moreover, we expect this positive association to start high and either increase over
the amount of information that potential adopters need to process on the time or at least stay high, as discounts may serve as an encouragement
platform varies substantially across free and paid apps. As 90% of Apple for less enthusiastic adopters throughout an app’s lifecycle.
App Store apps are free, the adoption decision is more taxing for users
looking to download a free app. The relative complexity of the free apps 4. Methodology
sub-market means greater information overload and higher search costs
(Payne, Bettman, & Johnson, 1993). To reduce search costs and deal 4.1. Data
with the undesirable consequences of complexity, potential adopters of
free apps may engage in selective information processing and utilize For the purpose of studying the drivers of app downloads, we
heuristics on the path to choice more than those of paid apps (Bettman, assembled a unique data set. The data set consists of a comprehensive
Luce, & Payne, 1998). One readily accessible source of information that list of variables acquired from one of the most prominent mobile ana­
may ease the burden of app discovery and app consideration is the lytics companies. The variables in this data set are downloads, revenues,
prominence of the app in the store (Ghose et al., 2013). Accordingly, we updates, appearance in featured lists and position in top apps charts. We
expect the effects of platform-controlled variables to be highest early on augment our transactional data set with publicly available data on app
and decrease over time, and to be greater for free apps than paid apps. ratings and reviews. To that end, we developed a web crawler to collect
Second, perceived risks associated with the acquisition of an app ratings and reviews from the web page of the iTunes app store.
evolve over the lifecycle and vary across free and paid apps. Among the To be able to answer our research question, we needed to observe
various types, two strongly correlated risks are relevant for the purposes each app from its initial release date in the app store. Therefore, we took
of our study: functional risk and financial risk (Jacoby & Kaplan, 1972). a stratified random sample of 1011 apps from 40,000 apps released in
Perceived functional risk is higher early on in a new product’s lifecycle the Apple App Store during the first five months of 2012 (between
because it is quite difficult to anticipate product performance in the January 1 and May 31) and obtained daily observations for all variables
early days. However, uncertainty about the product’s performance re­ over a one-year time frame starting on the day each app was released.
duces as it matures and potential adopters who are on the market in the Thirty-two apps had less than 365 usable observations and were later
later phases of the product’s lifecycle perceive lower functional risk discarded from the sample.11 The stratification ensures that the distri­
(Babić Rosario, Sotgiu, De Valck, & Bijmolt, 2016). Moreover, a crucial bution of the twenty-two app categories and business model types (i.e.,
difference between the two business models is the monetary risk asso­ free vs. paid app) in the app store is accurately represented in the
ciated with the purchase. While potential users of paid apps face this sample. Because some categories, such as Food & Drinks or Education,
risk, those of free apps don’t. The mere presence of a monetary risk did not have enough new apps launched in the sampling period, they are
implies paid apps score lower on trialability compared to free apps underrepresented. Moreover, we did not have any new Newsstand apps
(Rogers, 2003). As consumers rely on WoM taking place on the app store launched in the sampling period.12 These differences are compensated
to deal with perceived risks (Babić Rosario et al., 2016; Shen, 2015), we with a slight overrepresentation in some other categories, such as
expect the effectiveness of user-side drivers of downloads to decline over Games. Yet, we believe our sample provides a sufficiently accurate
time. We also expect the relationship to be weaker for free apps than representation of the situation in the app store at the time of data
paid apps, as users can readily try free apps without any transactional collection and helps us avoid the risk of producing results driven by
costs. For paid apps, on the other hand, potential users perceive greater category idiosyncrasies. Next, we present the model specification and
risk and the reviews of past users can provide them with useful addi­ detail the definition and operationalization of each variable in the
tional information. model.
Third, the composition of potential adopters and, consequently, the
variety of needs the developer should satisfy evolve over an app’s life­
cycle and across business models. Assuming away app discovery bot­ 4.2. Model specification and estimation strategy
tlenecks on the path to app adoption, those who download the app early
on are either innovators with high willingness to try new ideas or those As our goal is to explain which factors are related to downloads and
who value what the app’s initial version(s) has to offer (Rogers, 2003). how these relationships evolve over the first year of an app’s life cycle,
What separates the remaining users on the market who have not yet
downloaded the app from those who have, are their evaluations of the
app’s value proposition and their willingness to pay for that value. 10
When asked about why users have chosen to pay for apps over free alter­
Converting these remaining users to potential adopters requires ad­ natives, smartphone owners in the U.S. list app’s content as the top reason
justments to the value proposition and, if the app is paid, the price. (45%) and app’s features/functionality as the second reason (35%) for paying
Introducing new and improved versions of the app by means of updates for apps (Google, 2016).
11
and lower regular prices can stimulate demand and speed up growth for Eight of these applications were withdrawn before reaching the 1-year mark
and 24 either had a name change or were withdrawn after the first year
new offerings (Ataman, Van Heerde, & Mela, 2008). Temporary price
rendering access to publicly available data impossible.
reductions can further encourage app adoption by lowering the 12
The twenty-two application categories are Books, Education, Lifestyle,
perceived risk of making the wrong purchase. Moreover, as perceived Magazines/Papers, News, Reference, Entertainment, Music, Photo/Video, So­
monetary risk is most strongly associated with functional risk among all cial Networking, Games, Food/Drink, Health/Fitness, Medical, Sports, Business,
perceived risk types (Jacoby & Kaplan, 1972), potential users of paid Finance, Navigation, Productivity, Travel, Utilities, and Weather (Source:
apps are likely to have higher expectations from the developer and place Apple, 2018). Newsstand was later removed by Apple.

428
Z. Aydin Gokgoz et al. Journal of Business Research 123 (2021) 423–437

we specify a download response model with time varying parameters. industry. In terms of iOS downloads in particular, the majority of apps
The model explains downloads as a function of variables under the are claimed to be estimated with a margin of error below 3% and 95% of
control of app platform owners, app users, app developers, and several apps with a margin of error below 10%.
control variables: The platform-controlled variables include appearance on featured lists
and position in top app charts. Our data set contains information on (1)
∑
M ∑
N ∑
P
ln(Dit ) = αi + ln(Dit− 1 ) + βPLT βUSR βDEV whether an app was on a “Featured List” and which list it was featured
mt Ximt + nt Xint + pt Xipt
m=1 n=1 p=1 on and (2) the position of an app in a top apps chart and which chart it
∑
K was in. We classify the “Featured List”s into two main categories, top
+ γk Zikt + uit (1) featured lists and other featured lists, and code appearance on top (other)
k=1
featured lists as a dummy variable.15 We operationalized appearance in
top apps chart considering only the “Top Free” and “Top Paid” charts, as
where ln(Dit) is the natural logarithm of the number of times app i was
they are the most important ones with the highest traffic. Inspired by the
downloaded on day t. Because there are a few days with no downloads
findings of Carrare (2012) and the design of the app store at the time of
(0.55%), we add 1 to all observations before taking the logarithm. αi is
data collection, we acknowledge the natural break points in these charts
an app-specific constant.13 Ximt , Xint , Xipt and Zikt are, respectively,
and code appearance in a top apps chart using three dummy variables:
platform-controlled (m = 1, ⋯, M), user-side (n = 1, ⋯, N), developer-
above-the-fold (i.e., if an app was among the first five apps in the chart),
controlled (p = 1, ⋯, P), and control (k = 1, ⋯, K) variables that
below-the-fold (i.e., if an app was among the second five apps in the
explain daily downloads.
chart), and below-the-2nd-fold (i.e., if an app was among the apps listed
Following Chevalier and Mayzlin (2006), we specify a log–log model,
between the 11th and 25th positions).16
as there are scale effects emerging from higher views of popular apps
The user-side factors associated with downloads include valence and
compared to that of less popular apps. Because all our continuous in­
volume of WoM. Valence of WoM is operationalized as the average rating
dependent variables are log-transformed, their coefficients can be
score of the app’s most recent version. Using the ratings and reviews
interpreted as elasticities. The coefficients of the dummy variables, on
data we crawled from the official web-page of iTunes, we calculated the
the other hand, are semi-elasticities.
average rating score for an app’s most recent version by dividing the
To give the model the flexibility to capture the changing relationship
sum of all user ratings up to day t to the cumulative number of reviews
between the explanatory variables and downloads during the first year
up to day t, which is our measure for volume of WoM.
of the app’s life cycle, we specify the (semi-)elasticities as a function of
The developer-controlled variables associated with downloads include
linear and quadratic time trend:
updates and regular price and discount depth (only for paid apps). We
βPLT PLT PLT * PLT * 2
(2a) operationalize updates using the information in the three-digit number
mt = βm0 + βm1 t + βm2 t
known as the version number (e.g., Version 2.3.1). We infer the nature of
the changes made to the app from the digit changes between two
USR * USR *
(2b)
2
βUSR = βUSR
n0 + βn1 t + βn2 t
nt
consecutive versions: a change in the first digit indicates a major
improvement; a change in the second digit indicates an intermediate
DEV * DEV *
(2c)
2
βDEV = βDEV
p0 + βp1 t + βp2 t
pt improvement, while a change in the third digit indicates a minor
The quadratic time trend allows us to capture possible curvilinear improvement.17 We code each update as a step dummy that is switched
relationships over time. Following Liechty, Fong, and DeSarbo (2005), on for five days following an update.
we apply a transformation to time trend in the quadratic model for In addition to daily downloads, our data set contains information on
interpretation purposes: t* = (t/365 – 1/2). total revenues from downloads. We use these data to calculate the actual
Since the challenges faced by free and paid apps and the set of var­ price of a paid app on a daily basis (in cents) and determine the regular
iables associated with downloads for these apps are different, we esti­
mate the model separately for free and paid apps.14 Moreover, as
sensitivity checks, we explore whether there are any discrepancies in the 15
‘Featured List’ is a general term for all curated lists published by the plat­
results for different subsets of apps with respect to app categories (games form. We observe 189 apps (out of 979) featured in 180 different lists. Given the
vs. non-games), brands (new apps vs. apps by existing brands), and an scattered nature of these lists and the low number of featured apps, we decided
app’s ranking status (all apps vs. apps ranked at least 120 days) by to classify these lists under top featured lists and other featured lists. The
estimating the model separately for these sub-samples. reasoning behind this distinction is that top featured lists are the main lists that
are the easiest for users to notice, while others are not. Users are exposed to the
top featured lists on the landing page and need to actively search for the other
4.3. Variable definitions and operationalization lists. ‘Top Overall’, ‘New and Noteworthy’, or ‘Editor’s Choice’ are examples of
top featured lists. Other featured lists include category specific or curated lists
The dependent variable in Eq. (1) is the natural logarithm of the daily for special days (e.g., Mother’s Day Gift Guide, Apps for Graduates).
16
number of unique downloads of app i obtained from our data provider. At the time of data collection, Apple App Store top apps charts rolled on a
They are first time downloads that are unique to the user and do not continuous scrolling basis where each screen contained five apps. Therefore, we
contain updates. Download numbers are estimated using transactional separated the effect of being visible on the first page (referred to as ‘above-the-
fold’) from that of the second page (referred to as ‘below-the-fold’) and the
download data available to the mobile analytic company through their
following pages (referred to as ‘below-the-2nd-fold’) We think 5-page-views-by-
clients and public ranking charts. The mobile analytic company’s access
scrolling corresponding to the natural breakpoint at 25 provides us with a
to transactional data from over 100,000 apps with over 1.5 billion comprehensive list of top apps.
downloads leverages a level of accuracy that is unmatched in the 17
To illustrate the association between changes in version number digits and
the nature of the updates consider an app with the following history: Version
2.3 “Added History option”. Version 2.4 “Added a screenshot option. Now can
13
We assessed whether fixed-effects or random-effects correction would be save the picture in your iPad gallery any time you want. Find this option in
appropriate to control for time invariant differences across applications using game menu”. Version 2.4.1 “Updated ABOUT and HISTORY views”. Version 3
the Hausmann test. The results of this test suggested that the fixed-effects model “Clear option for removing the packages and images, UI changes, New packages
is appropriate in our case. at the top of the list in selector, Ability to share packages to your friends (email,
14
With the help of a Chow test, we assessed whether we can pool the co­ FB, twitter), Ability to create own packages”. The update from Version 2.4.1 to
efficients. The test result suggests estimating separate coefficients for free and Version 3 is a major update, from Version 2.3 to Version 2.4 is an intermediate
paid apps (F77,357181 = 14.699, p < .01). update, and from Version 2.4 to Version 2.4.1 is a minor update.

429
Z. Aydin Gokgoz et al. Journal of Business Research 123 (2021) 423–437

price dynamically by checking the mode of actual prices in a fixed time Table 3
window. Specifically, after setting the actual price on the first day equal Definition and operationalization of variables.
to the regular price, we calculate the difference between the actual price Variable Definition Operationalization Type/ Source
on a given day and the regular price of the previous day. If this difference Transformation
is zero (i.e., no price change), we set the regular price equal to the actual (Range before
trans.)
price. Otherwise, we look forward 30 days, calculate the mode of actual
prices in this time window, and set the regular price to the mode if the Downloads Daily Number of times Continuous/ Data
mode is equal to the current price, if they are not equal the regular price downloads app i was Log Provider
of an app downloaded on (0–354,395)
is set to previous day’s price.18 This procedure allows us to separate day t.
temporary changes in prices from permanent changes. We define dis­ Platform-controlled Variables
count depth as the ratio of cents-off to the regular price of the app. Appearance Whether an Divided into two Dummy/N.A. Data
Finally, as control variables we include (1) previous day’s downloads, on app has categories: Top and (N.A.) Provider
Featured been Other. “1” if the
which helps us to account for the unobserved effects of offline WOM, ads
Lists featured by app exists on one of
and other forms of publicity, (2) the number of days passed since an app the platform the featured lists
has been updated, to capture the effect of the frequency of updates, (3) under each
dummy variables for days of the week, special dates such as holidays category and “0”
(Christmas, New Year’s Eve etc.) and special occasions (Mother’s/Fa­ otherwise.
Appearance Whether Divided into three Dummy/N.A. Data
ther’s Day, Valentine’s Day etc.), and (4) several step dummies to con­ in Top and where categories: above- (N.A.) Provider
trol for the introduction of new devices and new iOS software updates. App an app has the-fold, below-
Table 3 summarizes the definition and operationalization of the Charts been placed the-fold, and
variables in the model and Table 4 presents summary statistics per in the top below-the-2nd-
app charts fold. “1” if the app
business model type.
exists in one of
these positions and
5. Results “0” otherwise.
User-side Variables
Valence of Average Average rating Continuous/ iTunes
Table 5 displays the coefficient estimates of our main models (i.e., for
WOM Rating Score score of the current Log Web
free and paid apps) and of models we estimated for sensitivity checks. version of app i on (1–5) page
Because our model includes interactions among all regressors and first- day t calculated
and second-order time trend, discussing the results coefficient-by- from the ratings of
coefficient is not fruitful. Instead, we calculated the marginal effect of users who also
wrote a review for
each variable over time – starting on the day of the release and reaching the app up to day t
365 with increments of two weeks – and the 95% confidence interval Volume of Cumulative Total number of Continuous/ iTunes
around this estimate using the Delta method. Fig. 2 displays the effects of WOM number of reviews of app i up Log Web
platform-controlled variables, Fig. 3 the effects of user-side variables, reviews to day t. (0–160,285) page
Developer-controlled Variables
and Fig. 4 the effects of developer-controlled variables.
Updates Whether an Divided into 3 Dummy/N.A. Data
In what follows, we first discuss the impact of each variable on app has categories: minor, (0–1) Provider
downloads using the average of the marginal effects over time and, if been intermediate, and
available, compare the (semi-)elasticities to earlier findings. We then updated major. Dummy
present our findings as to how these effects vary over the course of an variable for each
update category for
app’s first year of existence. To facilitate comparison with our expec­ five days following
tations, we summarize the key findings in Table 6. We conclude this the release of a new
section by discussing whether and how our main findings change under version.
different sub-samples of apps. Price Regular Inferred from a Continuous/ Data
price of an dynamic search Log Provider
app in cents over daily actual (0–49.99)
5.1. Platform-controlled variables prices.
Discount % cents-off (Regular Price – Continuous/ Data
Actual Price)/ None Provider
Being featured in a top list increases downloads of a free app by 3.93% Regular Price (0–100%)
and a paid app by 12.73% on average (Fig. 2 Panel A). Contrary to our Control Variables
expectations, it benefits a paid app about three times as much. This Day of the Control for Monday is chosen Categorical/ NA
week day of the as the baseline. NA
result may suggest that appearance in the top curated lists (e.g., Top
week (1–7)
Overall, New and Noteworthy, or Editor’s Choice) improves app dis­ Days since Counts days Number of days Continuous Data
covery rates more in less-crowded app categories than it does in more- last since last since last either of (0–365) Provider
crowded categories. Alternatively, it may indicate that, potential update update the update
adopters of paid apps consider these curated lists as a reliable source for categories.

a quality signal in their search for confirmation and uncertainty reduc­ Notes: Before applying the log transformation, we add 1 to all downloads as we
tion before they commit to a transaction. As to the temporal variation of have a few days with no downloads (0.55% of all observations). Exploratory
analysis of average download numbers centered on each update and the
observation that users give most feedback in the first few days after a new
18
In a highly dynamic market where 5-day price drops have been claimed to version release (Pagano & Maalej, 2013) supports our choice of 5-day time
have considerable effects on downloads, we choose 30 days as a long enough window. We check the sensitivity of our findings by considering a 4-day time
time window to outrun temporary price discounts and identify a new regular window, the second likely candidate, and find that the results are robust.
price level. Moreover, we checked the sensitivity of our findings by considering
15- and 45-day time windows and find that our results are robust. ([Link]
[Link]/2013/01/31/app-sales-work-five-day-iphone-app-price-drops-
boost-downloads-by-1665-on-ipad-by-871-revenue-growth-by-day-3/, last
accessed on 27/12/2019).

430
Z. Aydin Gokgoz et al. Journal of Business Research 123 (2021) 423–437

Table 4 fold increases downloads by 19.55% and below-the-2nd-fold by

Summary statistics. 27.35%. The effects are notably larger for free apps, as expected, and
Free Apps Paid Apps change sharply with each fold. Moreover, appearing in top apps charts
has a much larger impact than (top) featured lists.
Number of Apps 602 377
Downloads 1350.337 322.697 As to how the effects of appearing in top apps charts evolve over
(5857.519) (2348.376) time, we observe that appearing above and below the fold has a rela­
Platform-controlled Variables tively stable effect on free app downloads and a diminishing effect on
Appearance on Top Featured List 0.004 0.009 paid app downloads. The effectiveness of appearing below-the-2nd-fold
(0.065) (0.094)
Appearance on Other Featured List 0.010 0.015
declines following the release of an app, for free and paid alike, and
(0.100) (0.121) increases back to the initial level of effectiveness towards the end of the
Appearance Above-the-fold 0.001 0.002 year. Collectively, these results suggest that appearing in top apps
(0.034) (0.047) charts, anywhere above the 2nd fold, increases the speed with which
Appearance Below-the-fold 0.001 0.002
paid app downloads reach their market potential and gradually lose
(0.032) (0.045)
Appearance Below-the-2nd-fold 0.003 0.005 their ability to bring in new users.
(0.055) (0.072)
User-side Variables
5.2. User-side variables
Valence of WOM 2.963 3.453
(1.684) (1.503)
Volume of WOM 143.052 126.289 Panel A and Panel B in Fig. 3 display, respectively, WoM valence and
(1231.176) (545.603) WoM volume elasticities for free and paid apps. In line with Babić Rosario
Developer-controlled Variables
et al. (2016), we find that not all WOM metrics are positively associated
Minor Update 0.028 0.024
(0.165) (0.153)
with performance. Specifically, we find that a 10% increase in average
Intermediate Update 0.032 0.033 rating score decreases free app downloads by 0.13% and increases paid
(0.177) (0.179) app downloads by 0.23% on average. As expected, WoM valence has a
Major Update 0.001 0.004 higher impact on downloads in high risk situations (i.e., paid apps).
(0.036) (0.066)
More interesting patterns emerge when the evolution of valence
Price NA 2.621
(3.742) elasticities is considered. App download’s sensitivity to changes in WoM
Discount NA 0.009 valence early on is quite different for free and paid apps: an increase in
(0.083) average rating scores lowers the demand for free apps (by 0.42%, on
Control Variables average, in the first six months) but boosts download numbers for paid
Days since last update 81.615 92.594
(83.650) (89.565)
apps (by 0.22%, on average, in the first six months). The difference
disappears as apps mature and valence elasticities of free and paid apps
Notes: Cell entries are means and standard deviations, in parentheses, across all converge towards the end of the first year – approximately 0.02% and
apps and time periods.
0.05% increase for free and paid apps, respectively.
These findings raise concerns about the credibility of reviews for free
this factor’s effectiveness, free apps enjoy a similar lift, in terms of apps written early on, where users may be less involved or the barrier to
magnitude, throughout the year. Appearance on a top featured list starts leave a review may be quite low – a particularly interesting issue
to boost downloads significantly only later in a free app’s first year of considering the growing literature on fake reviews and their effects on
existence. For paid apps, being featured in top lists has a substantial sales (Dellarocas, 2006; Hu, Bose, Koh, & Liu, 2012; Mayzlin, Dover, &
effect on downloads early on. The effectiveness of this tool gradually Chevalier, 2014; Streitfeld, 2011). Our finding suggests that users take
decreases over time and reaches a similar level of effectiveness observed the reviews of free apps written early on less seriously and even step
for free apps. back from downloading the app. However, as time passes and the
In contrast, being featured in other lists fails to increase downloads: average rating score of a free app stabilizes around a certain value,
averaged over the entire year, downloads of free apps decline by 6.37% potential adopters start taking this information more seriously and into
and paid apps by 4.24% (Fig. 2, Panel B). Although contrary to our account.
expectations, this result is not very surprising due to the very narrow and As for paid apps, the results support the notion that potential
scattered nature of other featured lists. An obvious distinction between adopters want to reduce perceived risks when purchasing apps by pro­
the top featured lists and others leaps out. Considering there are about cessing the information provided by current users. The experiences
180 different lists, one may suggest that, instead of boosting downloads, encoded in these reviews matter more in potential adopters’ decisions in
appearing in these lists limits the general interest in the app and may the first half of the year and increasingly less from then on.
even prevent users who normally would have downloaded the app to shy WoM volume elasticities of downloads and their behaviors over time
away. The magnitude of the deleterious effect declines over time but are quite similar across free and paid apps. On average, a 10% increase
never completely disappears for free apps. Interestingly, being featured in WoM volume increases free app downloads by 0.13% and paid apps
in other lists becomes effective for paid apps towards the end of the year. by 0.16%. This effect increases towards the mid-year of the app’s release
In line with our expectations, merely appearing in top apps charts has a and declines at an increasing rate as apps mature.
positive effect on downloads except for paid apps appearing above-the-
fold later in their first year of existence (see Fig. 2, Panels C-D).19 On
5.3. Developer-controlled variables
average, getting into the list of apps presented above-the-fold increases
downloads of free apps by 80.28%, below-the-fold by 60.95%, and
Panels A-C in Fig. 4 display the relationships between minor, inter­
below-the-2nd-fold by 57.36%. For paid apps, appearing above-the-fold
mediate, and major updates and downloads of free and paid apps. As
has a negligibly small effect on average, whereas appearing below-the-
expected, updates benefit app demand in general. On average, down­
loads increase by 1.07% (minor), 0.83% (intermediate), and 22.30%
19
This unexpected result is due to a data peculiarity. We observe very few
paid apps appearing above-the-fold in this sub-section of the time window.
Accordingly, we are cautious about drawing strong conclusions about that
particular data partition.

431
Z. Aydin Gokgoz et al. Journal of Business Research 123 (2021) 423–437

Table 5
Parameter estimates per App type.
Main Models Models for Sensitivity Checks

Variable Free Apps Paid Apps Games Non-Games New Apps Existing Apps All Apps Ranked Apps

Constant 1.346*** 1.503*** 1.632*** 1.348*** 1.429*** 1.372*** 1.409*** 1.910***

Time − 0.324*** − 0.466*** − 0.349*** − 0.267*** − 0.283*** − 0.305*** − 0.289*** − 0.287***
Time2 0.775*** 1.034*** 0.948*** 0.784*** 0.761*** 0.877*** 0.800*** 2.036***
Top Featured Lists 0.023 0.117*** 0.063*** 0.116*** 0.093*** 0.097*** 0.087*** 0.110***
Top Featured Lists × Time 0.038 − 0.138*** − 0.011 − 0.024 − 0.032 0.018 − 0.017 0.034
Top Featured Lists × Time2 0.172 0.024 0.208 − 0.253* − 0.107 − 0.048 − 0.036 − 0.198
Other Featured Lists − 0.045** − 0.088*** − 0.045 − 0.067*** − 0.052*** − 0.077*** − 0.064*** − 0.034
Other Featured Lists × Time 0.088*** 0.035 0.321*** 0.090*** − 0.030 0.158*** 0.027 − 0.017
Other Featured Lists × Time2 − 0.240* 0.491*** 0.909*** − 0.004 0.165 0.051 0.148 − 0.383**
Above-the-fold 0.662*** − 0.076 − 0.073 0.536*** − 0.000 0.617*** 0.296*** 0.578***
Above-the-fold × Time − 0.217 − 1.857*** − 1.546*** − 0.073 − 1.016*** − 0.100 − 0.220* − 0.191
Above-the-fold × Time2 − 0.859 − 0.764 − 0.104 − 0.054 0.540 − 0.323 0.874** − 0.593
Below-the-fold 0.424*** 0.102* − 0.142** 0.512*** 0.178*** 0.368*** 0.247*** 0.499***
Below-the-fold × Time − 0.046 − 0.553*** 0.094 − 0.036 − 0.101 − 0.104 − 0.121 − 0.111
Below-the-fold × Time2 0.570 0.690 3.512*** − 0.315 1.448*** 0.502 1.083*** − 0.627*
Below-the-2nd-fold 0.393*** 0.141*** 0.206*** 0.236*** 0.218*** 0.247*** 0.230*** 0.320***
Below-the-2nd-fold × Time 0.054 0.021 0.133** 0.148*** 0.128*** − 0.043 0.091** 0.069
Below-the-2nd-fold × Time2 0.658*** 1.086*** 1.522*** 1.066*** 1.337*** 0.783*** 1.175*** 0.030
ln(Valence) − 0.003 0.035*** − 0.034*** 0.018*** 0.007** 0.011** 0.009*** − 0.007
ln(Valence) × Time 0.069*** 0.004 − 0.017 0.059*** 0.042*** 0.045*** 0.045*** − 0.151***
ln(Valence) × Time2 − 0.119*** − 0.130*** − 0.089** − 0.131*** − 0.126*** − 0.077** − 0.119*** − 0.099
ln(Volume) 0.018*** 0.021*** 0.014*** 0.024*** 0.020*** 0.012*** 0.018*** 0.038***
ln(Volume) × Time − 0.004** − 0.005** 0.005* − 0.006*** − 0.004** − 0.013*** − 0.007*** 0.017***
ln(Volume) × Time2 − 0.051*** − 0.059*** − 0.073*** − 0.056*** − 0.046*** − 0.060*** − 0.049*** − 0.222***
Minor Update 0.038*** 0.067*** 0.067*** 0.046*** 0.049*** 0.045*** 0.048*** 0.089***
Minor Update × Time 0.046*** 0.077*** 0.035 0.073*** 0.059*** 0.026 0.049*** 0.117***
Minor Update × Time2 − 0.312*** − 0.343*** − 0.398*** − 0.328*** − 0.308*** − 0.390*** − 0.331*** − 0.868***
Intermediate Update 0.052*** 0.070*** 0.010 0.074*** 0.055*** 0.061*** 0.058*** 0.105***
Intermediate Update × Time 0.107*** 0.048** 0.070** 0.083*** 0.081*** 0.050* 0.074*** 0.188***
Intermediate Update × Time2 − 0.506*** − 0.313*** − 0.083 − 0.520*** − 0.420*** − 0.375*** − 0.417*** − 0.795***
Major Update 0.066 0.132 0.015 0.067 − 0.000 0.072 0.021 0.096
Major Update × Time 0.736 0.206 1.265** 0.217 0.326 0.460 0.334* 0.756
Major Update × Time2 1.161 0.270 2.820*** 0.499 0.921** 1.109 0.905** 0.753
ln(Price) – − 0.140*** − 0.185*** − 0.123*** − 0.144*** − 0.129*** − 0.140*** − 0.126***
ln(Price) × Time – 0.029*** − 0.026*** − 0.016*** − 0.021*** − 0.012*** − 0.019*** 0.035***
ln(Price) × Time2 – − 0.022** 0.050*** − 0.015*** 0.007* 0.007 0.007* − 0.069***
Discount Depth – 1.394*** 1.483*** 1.291*** 1.467*** 0.997*** 1.355*** 1.145***
Discount Depth × Time – − 0.389*** − 0.626*** − 0.177*** − 0.350*** − 0.482*** − 0.409*** − 0.348***
Discount Depth × Time2 – − 1.694*** − 2.272*** − 1.124*** − 1.938*** − 0.626* − 1.660*** − 2.157***
ln(Downloadst-1) 0.737*** 0.723*** 0.720*** 0.736*** 0.735*** 0.728*** 0.734*** 0.723***
Tuesday 0.038*** 0.030*** 0.043*** 0.031*** 0.035*** 0.034*** 0.035*** 0.049***
Wednesday 0.034*** 0.030*** 0.051*** 0.024*** 0.031*** 0.037*** 0.032*** 0.053***
Thursday 0.053*** 0.064*** 0.100*** 0.038*** 0.056*** 0.060*** 0.057*** 0.077***
Friday 0.066*** 0.079*** 0.153*** 0.036*** 0.072*** 0.069*** 0.071*** 0.102***
Saturday 0.142*** 0.152*** 0.257*** 0.099*** 0.144*** 0.150*** 0.146*** 0.186***
Sunday 0.126*** 0.122*** 0.186*** 0.099*** 0.122*** 0.129*** 0.124*** 0.158***
Days Since Last Update − 0.000*** − 0.001*** − 0.000*** − 0.000*** − 0.000*** − 0.000*** − 0.000*** − 0.001***
Number of apps 602 377 285 694 707 272 979 99
Number of observations 219,730 137,605 104,025 253,310 258,055 99,280 357,335 36,135
R-Square 0.717 0.752 0.769 0.718 0.738 0.723 0.734 0.781
Average VIF 8.76 8.74 9.26 7.13 7.68 7.71 7.40 11.65

Notes: *indicates p < .1, ** indicates p < .05 and *** indicates p < .01. We use Monday as the baseline while dummy-coding days of the week variable. All models
include other controls, which are not shown here to conserve space.

(major) in response to free-app updates and by 3.78% (minor), 4.35% development tweaks) or an intermediate update (i.e., improvements to
(intermediate), and 17.18% (major) for paid apps.20 existing features of an app) shortly after an app’s release signals low app
The evolution patterns of the update semi-elasticities are similar quality (i.e., not ready for the market). However, approximately three
across update types and app business models, and the order of magni­ months into an app’s existence, the effects are reversed, and updates
tude is mostly preserved. As expected, the effect of an update increases start to boost downloads as expected.
moving from minor to intermediate updates and this increase is larger Panel D in Fig. 4 displays the evolution of price elasticity over time.
for paid apps. Interestingly, minor updates released shortly after the Consistent with the low-price elasticities reported for US Apple App
launch of free/paid apps lower the demand (Fig. 4, Panel A). We observe Store (e.g., Ghose & Han, 2014; Kübler et al., 2018), we find that a 10%
a similar pattern for intermediate updates of free apps. This result may increase in price lowers downloads by 1.41% on average. The magnitude
suggest that having to offer a minor update (i.e., bug fixes and of price elasticity declines with the passage of time: downloads become
less sensitive to price changes as the app matures. As to the effect of
discounting on downloads, displayed in Fig. 4 Panel E, we find a 13.20%
20
Though major updates findings are consistent with expectations direction­ increase in app demand in response to a 10% temporary reduction in
ally and magnitude wise, there is substantial uncertainty around the estimates – price. The increase in app demand in response to a discount is more than
due to the scarcity of major updates released in the last half of the data. Hence, double what has been reported in other studies (e.g., Ghose & Han,
we refrain from drawing strong conclusions about their effects.

432
Z. Aydin Gokgoz et al. Journal of Business Research 123 (2021) 423–437

0.30 0.30
Free Apps Paid Apps Free Apps Paid Apps

0.20 0.20

0.10 0.10

0.00 0.00

-0.10 -0.10

-0.20 -0.20

-0.30 -0.30
1 15 29 43 57 71 85 99 113 127 141 155 169 183 197 211 225 239 253 267 281 295 309 323 337 351 365 1 15 29 43 57 71 85 99 113 127 141 155 169 183 197 211 225 239 253 267 281 295 309 323 337 351 365
Days Days

(A) Appearance on Top Featured Lists (B) Appearance on Other Featured Lists

1.00 1.00 1.00

Free Apps Paid Apps Free Apps Paid Apps Free Apps Paid Apps

0.50 0.50 0.50

0.00 0.00 0.00

-0.50 -0.50 -0.50

1 15 29 43 57 71 85 99 113 127 141 155 169 183 197 211 225 239 253 267 281 295 309 323 337 351 365 1 15 29 43 57 71 85 99 113 127 141 155 169 183 197 211 225 239 253 267 281 295 309 323 337 351 365 1 15 29 43 57 71 85 99 113 127 141 155 169 183 197 211 225 239 253 267 281 295 309 323 337 351 365
Days Days Days

(C) Appearance Above-the-fold (D) Appearance Below-the-fold (E) Appearance Below-the-2nd-fold

Fig. 2. Effectiveness of Platform-controlled Variables over Time. Notes: Blue represents free apps and red represents paid apps. The solid lines are average marginal
effects and the shaded areas show 95% confidence interval. (For interpretation of the references to colour in this figure legend, the reader is referred to the web
version of this article.)

0.05 0.05
Free Apps Paid Apps Free Apps Paid Apps

0.00 0.00

-0.05 -0.05

-0.10 -0.10
1 15 29 43 57 71 85 99 113 127 141 155 169 183 197 211 225 239 253 267 281 295 309 323 337 351 365 1 15 29 43 57 71 85 99 113 127 141 155 169 183 197 211 225 239 253 267 281 295 309 323 337 351 365
Days Days

(A) WoM Valence (B) WoM Volume

Fig. 3. Effectiveness of User-side Variables over Time. Notes: Blue represents free apps and red represents paid apps. The solid lines are average marginal effects and
the shaded areas show 95% confidence interval. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of
this article.)

2014), indicating that users are more discount sensitive shortly after the ranked at least 120 days (mimicking previous studies’ settings). As the
launch of an app and discounts lose their power to boost downloads with data exhibits a tendency to crumble, we refrain from further breaking
maturity. Supporting this conclusion, we observe that offering discounts these sub-samples down to free and paid apps. The coefficient estimates
increases downloads more and more as time passes, reaching peak obtained from these sub-samples are presented in Table 5.
effectiveness mid-year, declines from then on to a level lower than its Noting that the over-time behavior of a given variable associated
initial effectiveness. with downloads is fairly similar across different sub-samples within a
grouping (e.g., games vs. non-games), unless otherwise mentioned, we
observe the following structural differences. First, appearing on other
5.4. Sensitivity checks featured lists, which had a deleterious effect on downloads, starts to
work for games but only later in the first year of a game’s existence.
To explore the sensitivity of our findings, we re-estimate the models Second, the effectiveness of appearance in top app charts varies sub­
with different sub-groups of apps in our data. Specifically, we compared stantially across sub-samples. Appearing above (below) the fold is much
the results from (1) games (fun-oriented and hedonic) to non-games more (more) strongly associated with downloads of games (vs. non-
(task-oriented and utilitarian), (2) new apps (no history or customer games), apps by existing businesses (vs. new apps), and apps ranked
base to rely on, mobile first) to apps by existing businesses (with a at least 120 days (vs. all apps).
customer base that can readily be activated, not necessarily mobile Third, ratings and, more importantly, the number of reviews appear
first), and (3) all apps in the sample (current study’s setting) to apps

433
Z. Aydin Gokgoz et al. Journal of Business Research 123 (2021) 423–437

0.20 0.20 2.00

Free Apps Paid Apps Free Apps Paid Apps Free Apps Paid Apps

0.15 0.15 1.50

0.10 0.10 1.00

0.05 0.05 0.50

0.00 0.00 0.00

-0.05 -0.05 -0.50

-0.10 -0.10 -1.00

-0.15 -0.15 -1.50

-0.20 -0.20 -2.00

1 15 29 43 57 71 85 99 113 127 141 155 169 183 197 211 225 239 253 267 281 295 309 323 337 351 365 1 15 29 43 57 71 85 99 113 127 141 155 169 183 197 211 225 239 253 267 281 295 309 323 337 351 365 1 15 29 43 57 71 85 99 113 127 141 155 169 183 197 211 225 239 253 267 281 295 309 323 337 351 365
Days Days Days

(A) Minor Update (B) Intermediate Update (C) Major Update

-0.10 1.60
Free Apps Paid Apps Free Apps Paid Apps

-0.11

1.40
-0.12

-0.13
1.20

-0.14

-0.15 1.00

-0.16

0.80
-0.17

-0.18
0.60
-0.19

-0.20 0.40
1 15 29 43 57 71 85 99 113 127 141 155 169 183 197 211 225 239 253 267 281 295 309 323 337 351 365 1 15 29 43 57 71 85 99 113 127 141 155 169 183 197 211 225 239 253 267 281 295 309 323 337 351 365
Days Days

(D) Price Elasticity (E) Discount Depth

Fig. 4. Effectiveness of Developer-controlled Variables over Time. Notes: Blue represents free apps and red represents paid apps. The solid lines are average marginal
effects and the shaded areas show 95% confidence interval. (For interpretation of the references to colour in this figure legend, the reader is referred to the web
version of this article.)

to have a stronger impact on downloads of non-games (vs. games) and on users’ adoption decisions. Among these platform-controlled vari­
new apps (vs. apps of existing businesses). Comparison of the user-side ables, the impact of appearance in top apps charts stands out, especially
variables’ effects of all apps vs. apps ranked at least 120 days shows early on in a paid app’s lifecycle and throughout a free app’s. Featuring
that both over-time behavior and strength of associations exhibit sub­ works when and where it benefits the platform owner the most: early on
stantial differences. Fourth, minor updates bring in slightly more users in the case of paid apps, which are featured almost twice as much, and
to games (vs. non-games). Intermediate updates, on the other hand, are only after gaining some traction with the users in the case of free apps,
more strongly associated with the downloads of non-games (vs. games). which experience a twofold increase in top feature placements over
Though they follow a similar over-time pattern, the relationships be­ time. For an app platform that generates its revenues from paid app
tween minor and intermediate updates and downloads observed in the downloads and in-app purchases, most of which comes from free apps,
apps ranked at least 120 days sub-sample (vs. all apps) are accentuated. our results have face validity.
Finally, potential adopters of games (vs. non-games), new apps (vs. The sheer size of platform-controlled variables’ effects begs the
apps by existing businesses), and apps ranked at least 120 days (vs. all question is this merely an awareness effect? Though it is without a doubt
apps) are more sensitive to price changes independent of the direction of that the platform’s decision to give apps a prominent position in the
change. We also observe a change in the over-time behavior of price store aids app discovery, it is likely that potential adopters use this to
elasticity when comparing all apps to ranked apps. infer additional information about apps. This is especially the case early
on in an app’s lifecycle, when relying on others’ opinions is not an op­
6. Conclusions, implications and avenues for future research tion either because the information is limited, or they think it is unre­
liable – evidenced by the deleterious effect of WoM valence on free app
In this paper, we explore factors that are related to app downloads downloads shortly after release. Equipped with the common knowledge
and how these relationships may develop during the first year after an that an app’s position in a top chart is determined by a combination of (i)
app’s release. Our access to a rich database with download numbers a few unknown proprietary factors, (ii) previous updates, ratings, re­
enables us to include newly released apps in the sample independent of views, and downloads of the app – whose effects are controlled for in our
their ranking status and explore the effects of a broader set of variables analyses –, and (iii) user engagement, retention, and revenues from in-
to develop a more generalized understanding. Time-varying results from app purchases, one may argue that what this variable also captures is
separate analyses for free and paid apps add interesting insights to the the effect current users’ repeated interactions have on download de­
existing literature and suggest valuable implications for app developers. cisions. This can be considered as a signal of app’s quality based on
Overall, our results suggest that the decisions and the actions of all others’ actions. Likewise, a “feature” can also be considered as a signal of
three players in the app market – developers, users, and platform owners quality that reflects the platform’s “seal of approval”. Users looking to
– seem critical for improving app performance at least at some point in download an app may infer that the featured app must be of high quality
the year following an app’s release. The effects of platform-controlled because the platform owner has no incentive to feature an app with a
variables, explored for the first time in this stream of the literature, low likelihood success. Independent of whether the actions of the plat­
dwarf those of variables shaping an app’s relevance (i.e., developer- form owner merely aids app discovery or are used as quality signals that
controlled factors) and variables reflecting current users’ views about facilitate app evaluations or downloads, the app platform is powerful.
an app (i.e., user-side factors), highlighting the power app platform has What can developers do given the influence app platforms have on

434
Z. Aydin Gokgoz et al. Journal of Business Research 123 (2021) 423–437

Table 6 The pattern of results observed for paid apps suggests that potential
Summary of results for the effects of external and internal factors on downloads adopters are even more cautious than those of free apps, especially early
for Free vs. Paid Apps. on. They combine signals from multiple sources (i.e., what others before
Free Apps Paid Apps them have said and done, whether the platform owner thinks the app is
Variable Direction Evolution over Direction Evolution over
worthy of a prominent display in the store, how much they have to part
time time with to acquire the app). However, as the paid app matures, the signals
emanating from platform-controlled variables lose their relative power
Platform Controlled
Variables over those from app users and app developers.
Appearance on Top + Flat U shape ++ Decrease Because getting the paid app in a top featured list as early as possible in
Featured Lists the lifecycle helps with the discoverability problem and boosts downloads,
Appearance on – Flat Inverse U – U shape the developer’s actions that affect the curators’ decisions to feature an app
Other Featured shape
Lists
assumes great importance. These decisions are based on, among other
Appearance Above- ++ Flat Inverse U – Decrease factors, the app’s content and functionalities (i.e., the value proposition) as
the-fold shape well as user-experience and user-interface design (i.e., how effectively and
Appearance Below- ++ Flat U shape + Decrease efficiently the value proposition is delivered). Therefore, developers of paid
the-fold
apps should aim to launch a ready-for-the-market app that (better) satisfies
Appearance Below- ++ U shape + U shape
the-2nd-fold a unique need and find ways of encouraging the users to rate and review the
User-side Variables application. With the initial boost they can get from being featured in a top
Valence of WOM – Decrease in + Inverse U list and, possibly sometime later a prominent position in the top apps charts,
magnitude shape they can attract more users who not only engage with the app but also
Volume of WOM Inverse U Inverse U
generate the much needed positive word-of-mouth for even more down­
+ +
shape shape
Developer loads. As updates released later in the paid app’s lifecycle are capable of
Controlled generating additional downloads, the developer can mine the reviews
Variables accumulated over time and decide in what direction to improve the app in
Minor Update Inverse U Inverse U
+ ++
the next version, ensuring longevity and a steady stream of revenues for the
shape shape
Intermediate Update + Inverse U ++ Inverse U developer as well as the platform owner. Considering the recent changes in
shape shape Apple’s revenue sharing model with developers of paid apps (i.e., a lower
Major Update + Flat U shape + Flat U shape cut in the second year), survival has become more important than before.
Price NA NA – Decrease in Another decision variable at the disposal of a paid app developer is
magnitude
Discount NA NA + Inverse U
price. Developers should carefully determine their price points prior to
shape release. As our results indicate that users are more price sensitive in the
early phases of an app introduction and gradually become less con­
Notes: Directions are based on the sign of the average effect over time and does
cerned with price, penetration pricing seems to be the more meaningful
not necessarily imply that the effect stays in that region throughout. ++ in­
dicates that the association between a variable and downloads is stronger for a
choice. Yet, developers have room to set a slightly higher initial price
specific app type compared to the other. and stimulate downloads and expand their user bases via promotions.
Moreover, as discounting reaches its peak effectiveness several months
into the app’s first year of existence, developers also have the flexibility
users’ adoption decisions? The pattern of results observed for free apps
to gradually increase the app’s price and encourage additional down­
suggests that potential adopters are cautious early on. They are mainly
loads occasionally by offering a discount.
swayed by the platform-controlled variable of appearance in top apps
Though our study adds interesting insights to the existing literature,
charts and the number of reviews the app has garnered (i.e., WoM vol­
there are several issues that may provide fruitful avenues for future
ume), which signal how large and engaged the current user base is.
research. First, given the nature of our data, we investigate the drivers of
Subjective evaluations of others (i.e., WoM valence) and developers’
downloads at the aggregate level (i.e., number of downloads on a day).
attempts to fix or improve apps (i.e., minor and intermediate updates)
However, while conceptualizing, we rely on an individual’s decision
are either seen as signals that cannot be trusted or that the app is of
journey on the app platform and identify factors that may facilitate
inferior quality. As free apps mature, potential adopters start to rely
users’ transitions from one journey stage to the next. Future research can
more on what others say in addition to what others do. Yet, others’
enrich the insights by modeling at the individual level, provided these
opinions never exceed the influence that (signals of) their actions have.
data exist and are accessible. Second, our findings on updates are
The only thing that is fully under the free-app developer’s control is
particularly interesting, as this is the only variable that is completely
the app’s value proposition; what unique need it satisfies and how
under the app developer’s control independent of the business model. In
effectively and efficiently it does so. The observation that intermediate
this paper, we inferred the nature of updates from the change in version
updates appearing in the early days of a free app have a stronger dele­
numbers. Future research should consider the nature of these updates by
terious effect on downloads than minor updates, which quickly recover,
analyzing the verbal descriptions of the improvements that accompany
suggests that developers are better off releasing almost-ready-for-the-
new version releases. Third, our analyses do not provide detailed in­
market apps. Though potential adopters may palate inefficiencies (i.e.,
sights into several systematic differences across apps. Although we
bugs) early on, they are unforgiving when it comes to ineffectiveness (i.
provide initial insights on how the effects change across games vs. non-
e., subpar content, functionality). This is especially important consid­
games or branded vs. non-branded developers, future research may
ering most free app developers launch minimum-viable-products and fix
contribute to an even more enriched understanding of these categorical
the issues along the way. As updates released later in a free app’s life­
differences. Finally, another growing area of research includes business
cycle have a larger effect on downloads, especially when the changes
model type as an important driver of app sales and performance. Espe­
between two consecutive versions of the app are more than minor, the
cially for digital platforms, freemium business models are of increasing
developer can aim to increase the relevance of the app in the eyes of a
interest. Though extant work in this domain sheds some light on the
larger pool of users and maintain or improve engagement of the current
effectiveness of freemium/paymium business model, our study could be
user base. This, in turn, may generate positive word-of-mouth on the
extended to investigate the performance of apps following different
platform and increase the chances that the free app gets a prominent
business models.
position on the platform.

435
Z. Aydin Gokgoz et al. Journal of Business Research 123 (2021) 423–437

References Ghose, A., & Han, S. (2014). Estimating demand for mobile applications in the new
economy. Management Science, 60(6), 1470–1488. [Link]
mnsc.2014.1945.
Anderson, M., & Magruder, J. (2012). Learning from the crowd: Regression discontinuity
Google (2016). How People Discover, Use, and Stay Engaged with Apps <[Link]
estimates of the effects of an online review database. The Economic Journal, 122
[Link]/_qs/documents/331/how-users-discover-use-apps-google-res
(563), 957–989. [Link]
[Link]> Downloaded on 9 July 2019.
AppAnnie (2018). “The 2017-2022 App Economy Forecast: 6 Billion Devices, $157
Hao, L., Li, X., Tan, Y., & Xu, J. (2011). The economic value of ratings in App market.
Billion in Spend & More” <[Link]
SSRN Electronic Journal. [Link]
pp-annie-2017-2022-forecast/> Retrieved 9 July 2020.
Hu, N., Bose, I., Koh, N., & Liu, L. (2012). Manipulation of online reviews: An analysis of
AppAnnie (2019). The State of Mobile 2019. Downloaded on 5 December 2019.
ratings, readability, and sentiments. Decision Support Systems, 52(3), 674–684.
AppBrain (2019). Number of Android Apps on Google Play <[Link]
[Link]
/stats/number-of-android-apps> Retrieved 9 July 2020.
Jacoby, J., & Kaplan, L. B. (1972). The components of perceived risk. Proceedings of the
Apple (2020) <[Link] Retrieved 9 July 2020.
Annual Conference of the Association for Consumer Research, 10, 382–393.
Apple Insider (2018). “The Revolution Steve Jobs Resisted: Apple’s App Store Marks 10
Kim, M., Kim, J., Choi, J., & Trivedi, M. (2017). Mobile shopping through applications:
years of Third-party Innovation” <[Link]
Understanding application possession and mobile purchase. Journal of Interactive
-revolution-steve-jobs-resisted-apples-app-store-marks-10-years-of-third-party-inno
Marketing, 39, 55–68. [Link]
vation> Retrieved 9 July 2020.
Kübler, R., Pauwels, K., Yildirim, G., & Fandrich, T. (2018). App popularity: Where in the
Arora, S., Hofstede, F., & Mahajan, V. (2017). The implications of offering free versions
world are consumers most sensitive to price and user ratings? Journal of Marketing,
for the performance of paid Mobile Apps. Journal of Marketing, 81(6), 62–78. https://
82(5), 20–44. [Link]
[Link]/10.1509/jm.15.0205.
Lavidge, R., & Steiner, G. (1961). A model for predictive measurements of advertising
Ataman, M. B., Van Heerde, H., & Mela, C. (2008). Building brands. Marketing Science,
effectiveness. Journal of Marketing, 25(6), 59–62. [Link]
27, 1036–1054. [Link]
Lee, G., & Raghu, T. S. (2014). Determinants of Mobile Apps’ success: Evidence from the
Aydin Gokgoz, Z., Ataman, M. B. & Van Bruggen, G. (2020). If it ain’t broke, don’t fix it:
App Store Market. Journal of Management Information Systems, 31(2), 133–170.
How incorporating user feedback in product development affects mobile application
[Link]
ratings. Working Paper. Rotterdam School of Management, Erasmus University.
Liechty, J. C., Fong, D. K. H., & DeSarbo, W. S. (2005). Dynamic models incorporating
Babić Rosario, A., Sotgiu, F., De Valck, K., & Bijmolt, T. H. A. (2016). The effect of
individual heterogeneity: Utility evolution in conjoint analysis. Marketing Science, 24
electronic word of mouth on sales: A meta-analytic review of platform, product, and
(2), 285–293. [Link]
metric factors. Journal of Marketing Research, 53(3), 297–318. [Link]
Liu, H., Lobschat, L., Verhoef, P. C., & Zhao, H. (2019). App adoption: The effect on
10.1509/jmr.14.0380.
purchasing of customers who have used a mobile website previously. Journal of
Bellman, S., Potter, R. F., Treleaven-Hassard, S., Robinson, J. A., & Varan, D. (2011). The
Interactive Marketing, 47, 16–34. [Link]
effectiveness of branded Mobile Phone Apps. Journal of Interactive Marketing, 25(4),
Mayzlin, D., Dover, Y., & Chevalier, J. (2014). Promotional reviews: An empirical
191–200. [Link]
investigation of online review manipulation. The American Economic Review, 104(8),
Bettman, J., Luce, M. F., & Payne, J. W. (1998). Constructive consumer choice processes.
2421–2455. [Link]
Journal of Consumer Research, 25(3), 187–217. [Link]
Mclean, G., Osei-Frimpong, K., Khalid, A., & Marriott, H. (2020). Examining consumer
Bijmolt, T., van Heerde, H., & Pieters, R. (2005). New empirical generalizations on the
attitudes towards retailers’ M-commerce mobile applications – An initial adoption
determinants of price elasticity. Journal of Marketing Research., 42, 141–156. https://
vs. continuous use perspective. Journal of Business Research, 106, 139–157. https://
[Link]/10.1509/jmkr.42.2.141.62296.
[Link]/10.1016/[Link].2019.08.032.
Blattberg, R. C., Briesch, R., & Fox, E. J. (1995). How promotions work. Marketing
Nysveen, H., Pedersen, P. E., & Thorbjørnsen, H. (2005). Intentions to use mobile
Science, 14(3), 122–132.
services: Antecedents and cross-service comparisons. Journal of the Academy of
Boyd, D. E., Kannan, P. K., & Slotegraaf, R. J. (2019). Branded Apps and their impact on
Marketing Science, 33(3), 330–346. [Link]
firm value: A design perspective. Journal of Marketing Research, 56(1), 76–88.
Pagano, D., & Maalej, W. (2013). User feedback in the Appstore: An empirical study. In
[Link]
21st IEEE international requirements engineering conference (RE), Rio de Janeiro (pp.
Cao, L., Liu, X., & Cao, W. (2018). The effects of search-related and purchase-related
125–134). [Link]
Mobile App additions on retailers’ shareholder wealth: The roles of firm size, product
Payne, J., Bettman, J., & Johnson, E. (1993). The adaptive decision maker. New York:
category, and customer segment. Journal of Retailing, 94(4), 343–351. [Link]
Cambridge University Press.
org/10.1016/[Link].2018.08.003.
Rogers, E. (2003). Diffusion of Innovations (5th ed.). Simon and Schuster.
Carrare, O. (2012). The impact of bestseller rank on demand: Evidence from the App
Schulze, C., Schöler, L., & Skiera, B. (2014). Not all fun and games: Viral marketing for
market. International Economic Review, 53(3), 717–742. [Link]
utilitarian products. Journal of Marketing, 78(1), 1–19. [Link]
j.1468-2354.2012.00698.x.
jm.11.0528.
Chen, A. (2018). “New Data Shows Losing 80% of Mobile Users is Normal, and Why the
Sensortower (2019). “The Top 1% of App Publishers Generate 80% of All New Installs”
Best Apps Do Better” <[Link]
<[Link]
ur-mobile-users-is-normal-and-that-the-best-apps-do-much-better/> Retrieved 9
201%20Percent%20by%20Downloads,total%20of%2029.6%20billion%20down
July 2019.
loads> Retrieved 9 July 2020.
Chevalier, J. A., & Mayzlin, D. (2006). The effect of word of mouth on sales: Online book
Simon, H. (1979). Dynamics of price elasticity and brand life cycles: An empirical study.
reviews. Journal of Marketing Research, 43(3), 345–354. [Link]
Journal of Marketing Research, 16(4), 439–452. [Link]
jmkr.43.3.345.
002224377901600401.
Colicev, A., Malshe, A., Pauwels, K., & O’Connor, P. (2018). Improving consumer
Shankar, V., & Balasubramanian, S. (2009). Mobile marketing: A synthesis and prognosis.
mindset metrics and shareholder value through social media: The different roles of
Journal of Interactive Marketing, 23(2), 118–129. [Link]
owned and earned media. Journal of Marketing, 82(1), 37–56. [Link]
intmar.2009.02.002.
10.1509/jm.16.0055.
Shankar, V., Venkatesh, A., Hofacker, C., & Naik, P. (2010). Mobile marketing in the
De Vany, A., & Walls, W. (1996). Bose-Einstein dynamics and adaptive contracting in the
retailing environment: Current insights and future research avenues. Journal of
motion picture industry. Economic Journal, 106(439), 1493–1514. [Link]
Interactive Marketing, 24(2), 111–120. [Link]
10.2307/2235197.
intmar.2010.02.006.
Dellarocas, C. (2003). The digitization of word of mouth: Promise and challenges of
Shen, G. (2015). Users’ adoption of mobile applications: Product type and message
online feedback mechanisms. Management Science, 49(10), 1407–1424. [Link]
framing’s moderating effect. Journal of Business Research, 68(11), 2317–2321.
org/10.1287/mnsc.49.10.1407.17308.
[Link]
Dellarocas, C. (2006). Strategic manipulation of internet opinion forums: Implications for
Smart Insights (2019). “Top 7 Biggest Flops in the Mobile App Industry” <[Link]
consumers and firms. Management Science, 52(20), 157–1593. [Link]
[Link]/mobile-marketing/top-7-biggest-flops-mobile-app-industry/>
10.1287/mnsc.1060.0567.
Retrieved 9 July 2020.
Duan, W., Gu, B., & Whinston, B. A. (2011). The dynamics of online word-of-mouth and
Streitfeld, D. (2011). In a Race to Out-rave, 5-Star Web Reviews Go for $5. <[Link]
product sales: An empirical investigation of the movie industry. Journal of Retailing,
.[Link]/2011/08/20/technology/[Link]>
84, 233–242. [Link]
Retrieved 8 December 2019.
Dhar, V., & Chang, E. A. (2009). Does chatter matter? The impact of user-generated
Van den Ende, J. C. M., Jaspers, F. P. H., & Rijsdijk, S. A. (2013). Should system firms
content on music sales. Journal of Interactive Marketing, 23(4), 300–307. [Link]
develop complementary products? A DYNAMIC MODEL AND AN EMPIRICAL TEst.
org/10.1016/[Link].2009.07.004.
Journal of Product Innovation Management, 30, 1178–1198. [Link]
Engström, P., & Forsell, E. (2018). Demand effects of consumers’ stated and revealed
jpim.12053.
preferences. Journal of Economic Behavior & Organization, 150, 43–61. [Link]
Van Heerde, H., Dinner, I., & Neslin, S. (2019). Engaging the unengaged customer: The
org/10.1016/[Link].2018.04.009.
value of a Retailer Mobile App. International Journal of Research in Marketing, 36(3),
Garg, R., & Telang, R. (2013). Inferring App demand from publicly available data. MIS
420–438. [Link]
Quarterly, 37(4), 1253–1264. [Link]
Van Noort, G., & Van Reijmersdal, E. A. (2019). Branded Apps: Explaining effects of
Gill, M., Sridhar, S., & Grewal, R. (2017). Return on engagement initiatives: A study of a
brands’ mobile phone applications on brand responses. Journal of Interactive
business-to-business Mobile App. Journal of Marketing, 81(4), 45–66. [Link]
Marketing, 45, 16–26. [Link]
org/10.1509/jm.16.0149.
Zhang, D., & Adipat, B. (2005). Challenges, methodologies, and issues in the usability
Ghose, A., Goldfarb, A., & Han, S. (2013). How is the mobile internet different? Search
testing of mobile applications. International Journal of Human-Computer Interaction,
costs and local activities. Information Systems Research, 24(3), 613–631. [Link]
18(3), 293–308. [Link]
org/10.1287/isre.1120.0453.

436
Z. Aydin Gokgoz et al. Journal of Business Research 123 (2021) 423–437

Zeynep Aydin Gokgoz is a PhD candidate in Rotterdam School of Management, Erasmus Journal of Marketing Research, Marketing Science, Journal of Product and Brand Man­
University. She received her Bachelors’ with honors in Statistics from Middle East Tech­ agement, and Social Indicators Research. Assoc. Prof. Ataman’s research focuses on
nical University and her Master’s degree in Industrial Engineering from Bilkent University. developing quantitative models to support marketing decision making and seeks to help
Her research interests include Bayesian Modeling, Time Series Models and Text Analysis managers better understand how marketing affects performance specifically in the long
with applications on mobile marketing and user generated content. She presents her work run.
at leading international conferences such as European Marketing Academy Conference and
Marketing Science Conference.
Gerrit Van Bruggen Professor van Bruggen’s primary research interest lies in strategic
marketing issues and the impact of information technology and information systems on
M. Berk Ataman is an Associate Professor of Marketing at Koç University Faculty of marketing strategy and decision making. His research has been published in the field’s
Economics and Administrative Sciences. Assoc. Prof. Ataman received his [Link]. and [Link]. leading scientific journals including Marketing Science, the Journal of Marketing, the
degrees in Management Engineering from Istanbul Technical University and his Ph.D. Journal of Marketing Research, Management Science, MIS Quarterly, Information Systems
from Tilburg University. He held research positions at Rotterdam School of Management, Research and Interfaces. He is a regular speaker at international management conferences
(Erasmus University, 2006-2013), Tilburg University (2002-2006), and Istanbul Technical and an experienced teacher at the executive level. He is a former visiting scholar at the
University (1999-2002). Assoc. Prof. Ataman’s work has appeared in journals such as Smeal College of Business Administration at Pennsylvania State University in the US.

437