International Centre for Innovation
Technology and Education
PLATFORM PLAY AMONG INCUMBENT FIRMS:
THE WRONG FOCUS?
Authors
Nicolas van Zeebroeck1 & Jacques Bughin2
1
iCite – Solvay Brussels School of Economics and Management, Université libre de Bruxelles. 50,
avenue F.D. Roosevelt – CP114/05 – 1050 Brussels (Belgium)
[email protected]
2
McKinsey & Company – McKinsey Global Institute, Avenue Louise 480 – B23 – 1050 Brussels
(Belgium)
[email protected]
iCite Working Paper 2017 - 023
iCite - International Centre for Innovation, Technology and Education Studies
Université Libre de Bruxelles – CP114/05
50, avenue F.D. Roosevelt – B-1050 Bruxelles – Belgium
1
Platform play among incumbent firms: the wrong focus?
Nicolas van Zeebroeck
iCITE
Solvay Brussels School of Economics and Management
Université libre de Bruxelles
50, avenue F.D. Roosevelt - CP114/05
1050 Brussels (Belgium)
[email protected]
Jacques Bughin
McKinsey & Company
McKinsey Global Institute
Avenue Louise 480 - B 23
1050 Brussels (Belgium)
[email protected]
WORKING PAPER FOR DISCUSSION
This version: April 2017
Abstract
Scholars have been investigating the quick emergence of digital platforms, essentially through the
lens of digital native firms. If there is limited occurrence of platforms launched by incumbent firms
such as Daimler (Car2Go, Moovel) or Johnson Controls (Panopix), the economics of their strategic
responses through platforms by incumbents merit attention. This paper provides the first largescale empirical evidence on the incidence, nature and profitability of incumbent platforms. Our
contribution is threefold: (1) only 10 to 30% of incumbent firms have already engaged in some
platform play, with substantial heterogeneity across industries, however; (2) incumbent firms
engage in such strategies with a strong focus on their supply chain, suggesting that their platform
initiatives are biased toward supply-side economies of scale, rather than demand-side; and (3)
platform strategies contribute to revenue and/or profit growth only when they are combined with
a priority attached to the demand-side of the business (i.e. customers, instead of the supply chain),
typically through the unbundling or rebundling of products or services, and are part of an offensive
digital strategy at scale. These results are robust to a Heckman-selection equation and a number of
changes to the specification and key measures.
Keywords: Digital transformation, Digital strategy, Platforms, Incumbent capabilities, Demand-side
economies of scale
Acknowledgements
The authors wish to thank McKinsey/Digital for supplying the data. All errors and omissions remain ours alone.
2
1. Introduction
Platform businesses have recently been penetrating a large set of industries beyond the software
and media industries they originate from. This includes retail (Amazon, Alibaba), transport (Uber),
hospitality (Airbnb), payment services (Paypal), or telecom operators (Skype, WhatsApp, WeChat).
In parallel, scholars have taken note, not the least because the economics of platform often involve
a radically new configuration of the value chain and a disruption of established business models.
Following Shapiro and Varian (1999)’s landmark book, several scholars published seminal work on
the economics of multi-sided business in general and platforms in particular, such as Gawer and
Cusumano (2002), Rochet and Tirole (2003), Eisenmann et al. (2006) and Evans and Schmalensee
(2007). Annabella Gawer edited an important book a couple of years later (Gawer, 2009), collating
several important papers on the economics and management of platform businesses. Since then,
several publications on platforms have made their way into the top-selling business books. In 2016
alone, such works included Parker et al. (2016) and Evans and Schmalensee (2016).
A defining element of platforms is the presence of network effects, by which the participation of
users (on one side of the platform business) makes the platform more valuable and attractive to
other users (on the same or another side). Shapiro and Varian (1999) referred to these network
effects as “demand-side economies of scale”. These effects induce typical strategies of platforms
such as cross-subsidizing a certain category of users or a certain side of the platform markets. Once
they reach a certain tipping point at the cross-subsidized side, network effects further become
captive and considerably raise the switching costs, making entry extremely difficult with winnertakes-all battles (Evans and Schmalensee, 2001).
In theory, new entrants would play a platform game more often than an incumbent more interested
in preserving its own base (Evans and Gawer, 2016). However, the negative consequence of being
disrupted by a platform entrant is likely higher for an incumbent, that may play the game as a way to
pre-empt entry and reinvent itself. Johnson et al. (2008) makes the case that maintaining a business
requires recognizing when it needs a fundamental change. Platform-based business models, enabled
by digital technology, may help internalize the benefit of this fundamental change to an incumbent’s
business model. Eisenmann et al. (2011) developed the concept and theoretical foundation of a
typical strategy for incumbent firms, which they call “platform envelopment”. It consists for an
incumbent firm in re-bundling existing functionalities to offer them across several markets or
platforms.
Yet, beyond some anecdotal stories,1 empirical evidence is lacking about the incidence of and the
returns to platforms developed by incumbent firms. This paper provides a first attempt at filling this
gap. First, we seek to provide some survey-based evidence on the incidence and nature of the
adoption of platform strategies by incumbent firms. Second, we explore the conditions under which
such strategies are adopted in general, and the ability of established firms to challenge their supplyside focus. And third, we provide some evidence on the conditions under which platform strategies
contribute to firm performance.
Our empirical analysis is based on a McKinsey-designed executive survey of over 2000 incumbent
companies across a wide range of geographies and industries (from a panel of 12000 firms). In a
nutshell, our results suggest that (1) only 10 to 30% of incumbent firms have already engaged in
1
Daimler and BMW for instance have gained some success with their respective car sharing platforms Car2Go
and DriveNow, whereas Johnson Control’s Panopix platform had to be closed to external developers after 2
years of existence (Van Alstyne et al., 2016)
3
platform play, with substantial heterogeneity across industries, however; (2) incumbent firms
engage in such strategies with a strong focus on their supply chain, suggesting that their platform
initiatives are biased toward supply-side economies of scale, rather than demand-side; and (3)
platform strategies contribute to revenue and/or profit growth only when they are combined with a
priority attached to the demand-side of the business (i.e. customers, instead of the supply chain).
These results point at the need for incumbent firms to build the right demand-side capabilities (not
necessarily digital) and to reduce their risk aversion in order to play such strategies boldly and at
large, rather than on a small scale. The paper therefore calls for more research on the dynamic
capabilities required to engage in platform strategies and how to develop them, as initiated recently
by Leijon et al. (2017).
The paper is organized in 4 sections. Section 2 presents the contextual framework and summarizes
the state of the art on platform businesses. Section 3 describes the data, which is empirically
analyzed in Section 4. Section 5 concludes.
2. Theoretical background
Different definitions of platform businesses have been proposed in the literature, by scholars
including Gawer and Cusumano (2002), Rochet and Tirole (2003), Eisenmann et al. (2006) and Evans
and Schmalensee (2007). They are often based on the notion of a core, stable set of functions,
serving as a foundation for the development of a wide range of complementary products or services,
often by third-parties. More recently, Parker et al. (2016) define a platform as “a business based on
enabling value-creating interactions between external producers and consumers. It [the platform]
provides an open, participative infrastructure for these interactions and sets governance conditions
for them.”
Evans and Gawer (2016) propose a typology of platforms, organizing them in 4 categories:
transaction platforms (facilitating transactions between different users, such Uber or Amazon
Marketplace), innovation platforms (a technology, product or service acting as a foundation on top
of which other firms develop complementary technologies, such as Apple’s iOS or Microsoft
Windows), integrated platforms (a combination of the former 2 categories, supporting content
creation and collaboration, such as Facebook and Youtube), and investment platforms (companies
running a platform portfolio strategy and acting as a holding company, active platform investor, or
both, such as Priceline Group).
Gawer (2009) organized platforms along a different axis based on the context in which they appear:
internal platforms (used within a given firm), supply-chain platforms and industry platforms. Supplychain platforms are defined (Gawer, 2009) as “a set of subsystems and interfaces that forms a
common structure from which a stream of derivative products can be efficiently developed and
produced by partners along a supply chain.” They are frequent in the automotive industry for
instance. Industry platforms are “products, services or technologies that are developed by one or
several firms, and that serve as foundations upon which other firms can build complementary
products, services or technologies.” Prototypical examples include Microsoft Windows and Apple’s
iPhone.
The singularity and strength of platform models primarily stem from network effects. Platforms are,
by definition, multisided businesses, which mean that they serve as an infrastructure for interactions
among different sets of stakeholders or users. As a consequence, the value of a platform derives not
just from its own set of idiosyncratic features, but from the presence of potential interaction
counterparts for its users. Users on one side tend to make the platform more valuable to users on
4
the other side, and vice-versa. Jeff Bezos refers to this principle as “Amazon flywheel” (Stone, 2013).
Varian and Shapiro (1999) coined the term “demand-side economies of scale”.
The key to platform growth and success is indeed to build economies of scale on the demand-side of
the profit equation: technological improvements on the user side of platforms can raise efficiencies
in social interactions and demand aggregation to make bigger networks more valuable to their users
(Parker et al., 2016). Because network effects can give the largest company in a platform market an
advantage that is extremely difficult for competitors to overcome (Parker et al., 2016), the typical
entry strategy into a platform market requires revolutionary functionality to be offered (Henderson
and Clark, 1990; Bresnahan, 1999), which underlies the sequential winner-take-all battles typical of
platform markets, as observed by Evans and Schmalensee (2001), with superior new platforms
replacing old ones (Eisenmann et al., 2011).
Given the wide scope of applicability of platform businesses and the solid barriers to entry which
network effects can build, incumbent firms have reason to be concerned and to react. Examples of
platform plays include Accor opening a marketplace to independent hotels to compete against the
aggressive of OTAs in the travel industry; or Still Schibsted, building and consolidated classified
marketplaces globally. In their survey of platform businesses across the world, Evans and Gawer
(2016) also provide several examples of platform businesses developed by incumbent firms. Such
examples include Johnson Control’s Panopix, and automotive manufacturers’ initiatives such as
Daimler’s Car2Go and BMW’s DriveNow, Commonwealth Bank of Australia (CBA)’s payment
platform “Pi” and Android-based point of sales device “Albert”, open to third-party developers, or
the Swiss company LEGIC’s shift from security products to a technology platform. Leijon et al. (2017)
further investigate the case of 4 incumbent firms’ incursion into the area of digital platforms.
Eisenmann et al. (2011) highlight an alternative strategy, which they call “platform envelopment”. It
consists in bundling functionalities from different platforms into one, thereby extending the user
base of one platform to an adjacent area. This is typically the case of Daimler with its multi-modal
mobility platform Moovel, offering centralized access to various mobility platform providers such as
its own Car2Go, but also Flinkster or Deutsche Bahn, which can be booked and paid via the Moovel
app.
Yet, demand-side economies of scale (i.e. network effects) involve a significant departure from
traditional business models, which often excel at building and maximizing supply-side economies of
scale. Such a shift of focus may require new capabilities for incumbents to develop, as pointed out
by Leijon et al. (2017). Specifically, they discuss the difficulty of opening up to external value creation
and the resistance it may trigger. To handle the tensions between internal and external, their 4 case
companies ended up with specialized ecosystem setups. In short, their argument is that the shift
from products to platforms requires opening up the value creation process and giving away some
degree of control and planning over innovation and value creation.
Traditional firms will therefore face a tough challenge in their attempts to develop a platform
business: acquiring the capabilities needed to build demand-side economies of scale and grow in size
rapidly enough to build protective network effects. We derive our first set of hypotheses directly
from this assumption:
Hypothesis 1a: traditional firms are more likely to engage in platform strategies when they perceive
a larger risk of disruption by digital innovators.
Hypothesis 1b: traditional firms engaging in platform strategies are biased toward supply-side
economies of scale.
5
By the arguments developed in the literature about platform business models, their profit equation
rests on demand-side economies of scale. We therefore expect the contribution of a platform
strategy to the performance of incumbent firms to depend upon the successful development of such
network effects on the demand side. Given that such network effects are costly to build (consider
the cash burn of successful platforms such as Amazon and Facebook in their early years) and become
protective only beyond a certain tipping point (Gawer, 2009), their profitability should depend on
their growth past such a tipping point. Eisenmann et al. (2011)’s theory of platform envelopment
further suggests that developing a platform through the bundling of existing functionalities may be a
successful strategy for incumbents facing the threat of platform disruption. From these
observations, we infer the two following hypotheses:
Hypothesis 2a: platform strategies contribute to revenues and profits only to the extent that they
focus on attracting the demand-side.
Hypothesis 2b: platforms are revenue and profit-enhancing only to the extent that they are built at
scale, i.e. part of an offensive strategy with large investments in digital technology.
3. Data
Our empirical analysis relies on survey data. The survey was run by TNS in 2016 on behalf of
McKinsey within a representative panel of 12,000 C-suite executives across industries and regions.
Conducted every year for almost a decade, the survey covers a variety of business issues. Earlier
waves of the survey have been used in different research efforts, including Brynjolfsson et al. (2011)
and Bughin (2016).
The 2016 edition of the survey covers the firms’ perception of digitization in their industry and how
they responded to it. The sample for our analysis, described at length in Bughin et al. (2017) and
Bughin and van Zeebroeck (2017a), is made up of 2135 firms.2 The data further include a number of
firm-level observables, which we control for. These controls include proxies for the region, size,
industry and type of activities (mono/multi products/services). Roughly 43% of companies have their
headquarters in Europe, versus 30% in the US; one third of the firms in the sample have sales in
excess of $1 billion; 62% are privately owned, 35% are primarily B2C firms, and the most
represented sectors are high tech (22%), financial services (20%), and professional services (20%).
3.1. Measuring the adoption of a platform model
Our identification of platform players is based on the responses to the following question: “Turning
to your organization’s goals for digital going forward, which of the following are the most important
objectives of its digital strategy?” The survey offered 6 options, among which up to 3 could be
ranked in order of priority:



Better serving the needs of current customers (e.g., unbundling or rebundling existing
products or services, providing more opportunities for customization)
Introducing new products or services with digital features to meet new demand and/or the
needs of new customers
Tapping a previously inaccessible supply of given products or services in the market (e.g., ride
sharing or room sharing)
2
As mentioned in Bughin and van Zeebroeck (2017a), various strategies have been used to limit and assess
potential common method bias issues, including randomizing the response items, ensuring anonymity, and
using a Harman single factor test, which did not indicate presence of common method bias in our data.
6



Making it easier for customers to access the available supply of products or services through
digital channels (e.g., through new marketplaces)
Scaling down the company’s cost structure by automating, digitizing, or virtualizing
processes
Redefining the industry’s value chain so customers and suppliers can interact more directly
and benefit from network effects (e.g., a platform offering free access to information,
crowdsourcing, disintermediating suppliers)
Among the 6 options, we use the last one as an indicator of platform play: “Redefining the industry’s
value chain so customers and suppliers can interact more directly and benefit from network effects
(e.g., a platform offering free access to information, crowdsourcing, disintermediating suppliers)”.
Firms are considered platform players if they ranked this option 1st.3 Table 1 reports the frequency of
each of these options as ranked by respondents. It shows that incumbent firms focus their top
priority on better serving their current customers through un- or re-bundling, developing new
offerings or serving new customer segments, and distributing through digital channels with equal
frequencies (about 22%). Cost cutting and platform play are almost two times less frequently cited
as top priorities, whereas access to new sources of supply (such as ride sharing or room sharing) are
barely cited by 2% of the responding incumbents.
Frequency of being cited Frequency of being
among top 3 priority
cited as top 1 priority
Better serving existing customers
62%
(through bundling or customization)
55%
New products/services or customers
13%
New sources of supply
57%
Digital distribution channels
45%
Cost cutting or downscaling (e.g. automation)
32%
Reconfigure the value chain (platform play)
Table 1 – Digital strategies pursued by incumbent firms
23%
21%
2%
22%
13%
12%
3.2. Measuring firm performance
Our analysis uses two different measures of performance: revenue growth and profit margin growth.
Both of them are self-reported in the survey, in the form of 12 buckets (5% wide for most of them),
spanning from -50% or more to +50% or more. Both measures exhibit a unique mode at the [5%;9%]
bucket. The buckets are numbered in ascending order from 1 to 12. This implies that a one unit
increase in the variable corresponds to a jump by one bucket. Since buckets are between 4 and 5%
wide, such a jump around the mean represents a 4.5 percentage points increase in the growth rate.
3.3. Measuring the focus on the demand-side
In order to capture the demand-side orientation of the focal firm in its digital efforts, we created a
dummy equal to 1 if the firm has included the following statement in its 3 top digital priorities:
“Better serving the needs of current customers (e.g., unbundling or rebundling existing products or
services, providing more opportunities for customization).” As shown in Table 1, this is the case for
62% of the respondents. We then interact this dummy with our measure of platform play. Overall,
among the platform adopters in our sample (12% of the total sample of firms), there are exactly as
3
For the sake of robustness, we have also built an alternative version of this variable considering platform
adoption as the inclusion of this strategy among the top 3 digital priorities of the focal firm, the results are
robust to this specification. Our results are also robust to the extension of platform play to the third option
“Tapping a previously inaccessible supply of given products or services in the market (e.g., ride sharing or room
sharing)”. This latter option was however selected by only a tiny fraction of respondents (less than 2%).
7
many firms combining their platform play with such a demand-side focus as not. Following H2a, we
expect the interaction term to be positively and significantly associated with firm performance.
3.4. Measuring the scale of digital efforts
To investigate our second set of hypotheses, we use Bughin and van Zeebroeck (2017a)’s measures
of digital reactions among incumbent firms. It consists in a 4-level scale from weak (no strategic
change or significant investment in digital technology) and medium reactions (limited change to
corporate strategy or average investment in digital) to semi-bold (significant change to corporate
strategy OR significant overinvestment in digital) and bold-at-scale (significant change to corporate
strategy AND significant overinvestment in digital). Bughin and van Zeebroeck (2017a) have
observed a linear relationship between their measure of digital reactions and the growth of firm
revenues and profit margins, with the bold-at-scale group being the only one able to compensate
the depressive effect of digitization in their environment. In our sample, there are 15% firms in this
bold-at-scale group of digital players. Among platform players, 24% qualify as bold-at-scale players.
Following H2b, we would expect the interaction of the strongest level of reaction (bold-at-scale) with
platform play to be significantly contributing to firm performance.
4. Empirical implementation and results
4.1. Which incumbent firms engage in platform play?
Before looking at our hypotheses, we first look at the incidence and character of platform players
among incumbent firms. A basic descriptive analysis reveals some interesting patterns. First,
platform play is found in only a fraction of incumbent firms: only 12% of our sample of firms report
platforms as their top priority (32% list platforms among their top 3 priorities, however). There is
substantial heterogeneity among industries, as reported in Table 2. The platform game is twice as
frequently played by firms in high tech & telecom (18%) as in financial services (9%).
Sector
Financial Services
High Tech & Telecom
Manufacturing
Public, Social, Healthcare and other services
Services: Prof., Media, Transport and Retail
Total
Non-platform players
310
91%
332
82%
404
89%
313
89%
528
90%
Platform players
31
9%
71
18%
49
11%
40
11%
57
10%
1887
88%
248
Table 2 – Incidence of platform play across industry clusters
12%
Total
341
403
453
353
585
2135
Looking at firm characteristics, Table 3 compares the share of platform players among different cuts
of incumbents. Most differences are not significant, except for B2B v. B2C firms: 13% of firms with a
primary B2B focus have engaged into platforms, v. 8% only for firms with a primary focus on B2C.
This result is appealing in itself, given the above discussion on popular platforms, which mostly
involve consumers and demand-side network effects, and is consistent with our first hypothesis.
Platform players among
NO
YES
Public firm
11%
13%
Revenue over 1B
11%
12%
Focus is on B2C
13%
8%
Firm is mono-product or mono-service
12%
9%
Firm has products
11%
12%
Table 3 – Share of platform players among different sets of firms
8
4.2. How do incumbents play a platform strategy?
Turning to our first hypothesis, related with the strategic orientations of traditional firms engaging
into a platform play, our empirical strategy consists in estimating the influence of various elements
of a firm’s strategic reaction to digitization on their likelihood of developing a platform. We estimate
these effects in a logit model with platform play as (dummy) dependent variable. Specifically, we
consider 2 dimensions: the perception by the focal firm of digital disruption in its environment, and
the strategic orientation of the focal firm’s reaction to digital turbulence. Their respective estimated
coefficients are reported in column 2 and 3 of Table 4 respectively. Column 1 in the table reports the
estimation of a model with control variables only and the last columns report the results for a full
model with all variables at once. A number of patterns emerge from these results.
First, looking at column 2, platform strategies are significantly triggered by the perceived threat of
digital disruption, in line with hypothesis 1a. However, this is particularly so when digital threatens
the supply chain. This suggests that platform adoption among incumbent firms is disproportionately
viewed as a way to react to digital disruption on the supply side.4
Second, looking at column 3, platform play is significantly associated with 2 particular strategic
orientations: it is negatively associated with the adaptation of products, services and channels to
customers’ needs and preferences, and positively associated with an increased speed of operations
in businesses and functions.
These results hold in a specification including all explanatory factors (in column 4). They are also
robust to perimeter changes in the dependent variable, as reported in columns 5 and 6. In column 5,
the dependent variable (platform adoption) is set to 1 as soon as the focal firm lists platform play
among its top 3 digital priorities.5 In Column 6, firms are considered as platform players if they
include either the core strategy (“Redefining the industry’s value chain…”) or the access to new
sources of supply such as ride or room sharing as their top digital priority. In both cases, the above
findings hold qualitatively, modulo some differences in significance levels. A few additional factors
turn significant too with these alternative measures of the dependent variable, which were all
identically signed but not significant with the default definition of platform adoption: “New
relationships created with external business partners”, “New models of sharing profits and value with
external businesses” and “Proactive adaptation of business model, even at risk of cannibalization”
turn significantly positive in one or two of these specifications. All of these are consistent with a
supply-side focus and inconsistent with a demand-side priority.
These results add to the earlier observation about the disproportionate amount of B2B firms among
platform players, which is itself confirmed by our logit estimates (in column 1), as the parameter
associated with a B2C focus is consistently negative across all specifications and significant in the
control-only model (it is significant at the 20% probability level in all others).
The picture that emerges from these statistics is more consistent with our first hypothesis than with
the view of platform strategy as a way to build demand-side economies of scale by focusing on a
new relationship with consumers. On the contrary, we observe that incumbent firms engaging in
platforms have a disproportionate focus on B2B and on their supply chain, with little attention to
4
Note that including each dimension of digital one by one in distinct regressions yields the same conclusion:
only the overall perception of digitization and the perception of digital turbulence in the supply chain turn out
positive and significant. All other dimensions (distribution channels, competitive landscape and core
operations) are not significant at any conventional level.
5
Instead of the first priority in the default specification.
9
end-customers and digital channels. This might be a sign that supply-chain platforms are an easier
game to play for incumbent firms.
4.3. Do platform strategies contribute to firm performance?
We finally explore our second set of hypotheses (the contribution of platform strategies to firm
performance) in an econometric model that builds on Bughin and van Zeebroeck (2017a).
Specifically, we aim to estimate , which reflects the elasticity of firm performance to the adoption
of a platform strategy as in equation 1.
𝑌𝑖𝑗 = 𝑐1 + 𝜔𝑃𝑖 + 𝑋𝑖
(1)
Where: Yij is the performance of firm i in its industry j. Pi is a binary variable equal to 1 if firm i has
adopted a platform strategy and 0 otherwise, and Xi is a set of controls at the firm level. Likewise
Bughin and van Zeebroeck (2017a), we normalize each variable by differencing out its industry
average.
To test H2a and H2b, we need to interact our variable Pi with a measure of the focus on customer
service improvement (i.e. on the demand-side), CFi and with a measure of the intensity of digital
efforts, Zij (the strategic digital reaction of firm i), as in equation 2. Our hypotheses H2a and H2b are
essentially a test of >0 and >0 respectively.
𝑌𝑖𝑗 = 𝑐1 + 𝜔𝑃𝑖 + 𝛼𝑃𝑖 𝐶𝐹𝑖 + 𝛽𝑃𝑖 𝑍𝑖𝑗 + 𝛿𝐶𝐹𝑖 + 𝛾𝑍𝑖𝑗 + 𝑋𝑖
(2)
We estimate equations 1 and 2 using OLS.6 The results are reported in Tables 5 and 6 with revenue
growth and profit margin growth as dependent variables respectively. We include the different
terms one by one, starting with a model including digital reactions and controls only, which we call
“Baseline” (column 1). In general, our results replicate those of Bughin and van Zeebroeck (2017a):
the stronger the digital reaction (from weak to bold-at-scale), the larger the benefit in terms of
performance. In contrast, in columns 2 and 3 (which report the estimates of equation 1), our
measure of platform adoption is not significantly associated with firm performance. At face value,
this suggests therefore that platform strategies do not pay off, neither are they detrimental to firm
performance.
This neutral effect of platform strategies may nonetheless hide substantial heterogeneity, as
predicted by our hypotheses, which we explore next. Looking at H2a first, the results in column 5
highlight a strongly significant and positive contribution of platform play on both performance
measures when combined with a digital focus on better serving the needs of existing clients,
typically through bundling or customization. This result holds in the full model reported in column 6
(for both performance measures). These results are consistent with our hypothesis.
We finally turn to H2b, which predicts that platform strategies should be beneficial only when part
of a bold at scale digital strategy. Let’s first recall that the “bold at scale” dummy capture firms that
reacted to digitization in an offensive way with significant overinvestments in digital technology
compared with their peers. We test this hypothesis by estimating equation 2, which includes an
interaction term between the platform dummy (Pi in equation 2) and the strongest category of
reaction (Zij in equation 2). The results are reported in column 3 of Tables 5 and 6. The positive and
significant coefficient associated with the interaction term clearly supports the view that platform
strategies only pay-off when part of an ambitious, corporate-level, strategy. This result, however,
only shows up with revenue growth as dependent variable, not with profit margin growth,
6
The results are robust to the use of a tobit model (see Robustness section below).
10
suggesting that scale in platform strategies is revenue-enhancing, but not necessarily profitenhancing. Profit contributions rely more on the customer orientation of the platform initiative. This
result is also consistent with the theory of platform envelopment of Eisenmann et al. (2011), which
suggest that incumbent firms can engage in platform strategies by bundling existing products,
services or functionalities.
We find further support to these findings in columns 7 and 8 of the same table 5, which report
subsample regressions of the full model. In column 7, the sample is limited to non-platform players,
whereas column 8 is run only on platform players. As it can be seen, the customer orientation is
significant and positive with platform players, but not with non-platform players, confirming that
customer orientation is a must when playing a platform strategy. This result qualitatively holds in the
EBIT growth model (columns 7 and 8 of table 6).
In contrast, whereas the bold-at-scale level of reaction is dominant in both subsamples, lower levels
of digital reaction (medium or semi-bold) are significantly contributing to revenue growth among
non-platform players, but not among platform players, supporting the idea that platform play, more
so than other types of digital strategies, require a bold and at scale commitment. This latter result is
less marked but also qualitatively holds in the EBIT growth model.
4.4. Exploring robustness
We have tested the robustness of our results in different ways. We first run a Heckman selection
model, in which our first set of results (determinants of platform play) are used to form our selection
equation. The second and first stage results, reported in Tables 7 and 8 respectively, are in line with
our main estimates.
Table 8 reports a battery of extra robustness estimates. Throughout the table, odd columns are for
revenue growth as dependent variable, and even columns use EBIT growth instead. The first four
columns estimate equation 2 using the two alternative measures of platform play: columns 1 and 2
consider all 3 priorities (not just the first) and columns 3 and 4 extend the definition of platform play
to accessing new sources of supply such as ride or room sharing. Our core results (platforms pay off
only when combined with bold-at-scale digital strategies and/or with a strong customer focus) hold
qualitatively true, although significance is not always achieved, especially with the EBIT rate growth
as dependent variable.
Columns 5 to 8 report subsample estimates with two specific sectors: financial services (columns 5
and 6) and high-tech & telecom (columns 7 and 8). The result on customer focus is robust in both
sectors. The bold-at-scale interaction only holds in the financial sector, not in high tech and telecom.
Columns 9 and 10 provide additional support to hypothesis H2a by considering which C-level
executive is sponsoring the digital initiatives within the company: the CEO, the CFO or the CMO
(CIO/CDO serves as the reference). Each C-level sponsor is interacted with our core explanatory
variable (platform play). It turns out that only the sponsorship of the Chief Marketing Officer has a
positive and significant mediating effect on the platform-performance relationship. This positive
effect is however limited to revenue – not profit – growth.
Finally, columns 11 and 12 test the sensitivity of our results to an alternative regression model, using
a censored tobit (our dependent variables – due to their coding – are bounded in the [-8;+8] range).
The results are robust.
11
5. Discussion and conclusions
Platforms are receiving traction and raising concerns across an ever-wider range of industries. Once
the prerogative of innovative digital start-ups and unicorns, they are increasingly considered and
launched by incumbent firms, with some visible successes, although anecdotal. BMW and Daimler’s
respective car sharing initiatives (DriveNow and Car2Go respectively) seem to be paying off quite
well. Johnson Control, in contrast, had to close its platform Panoptix to external developers for lack
of traction. However, empirical evidence on the magnitude of incumbent platforms as well as on
their character and productivity effects is dramatically lacking. The present paper provides a first
step in addressing this gap.
First, it documents the phenomenon by showing that some 8 to 20% of incumbent firms are
engaging in some platform initiatives to a certain extent, with substantial cross-industry variation.
Second, it provides suggestive evidence that platform strategies tend to be played by incumbent
firms with an “old-world” mindset, disproportionately focused on supply-side economies of scale.
Finally, the paper further provides suggestive evidence that this is the wrong focus: platforms payoff more when they are designed to build demand-side economies of scale, aka network effects, by
focusing on the demand-side of the profit equation, and part of an ambitious and offensive digital
strategy. Overinvestment in digital technology, combined with an offensive strategic move (as
advocated by Bughin and van Zeebroeck (2017b)), is required to build and internalize network
effects. With other words, scale in digital investments matter in order to go beyond the “tipping
point” in the development of network effects. This is even more of a must when the threat of
disruption is stronger, and digital competition potentially larger, pleading for a differentiated
approach in platform play.
Our results suggest that successful platform play requires two key elements: being part of a broader
digital strategy that is bold and played at scale (consistently with the main result of Bughin and van
Zeebroeck (2017a)), and being combined with a strong focus on customer experience and service,
typically through bundling or customization. These are two difficult prerequisites for established
firms. The former (bold-at-scale digital strategies) requires a mix of ambition (building digital
solutions and models at scale) and courage (since the risk is cannibalizing existing profit pools), and
most certainly some risk-taking. The latter requires a change of mindset and most likely the
acquisition of new capabilities focused on demand-side rather than supply-side economies of scale.
This is not a digital capability per se, but a cultural mindset.
Leijon et al. (2017) offer some insights into the main obstacles faced by incumbent firms engaging in
platform strategies. Specifically, they discuss the difficulty of opening up to external value creation
and the resistance it may trigger. To handle the tensions between internal and external, their 4 case
companies ended up with specialized ecosystem setups. In short, their argument is that the shift
from products to platforms requires opening up the value creation process and giving away some
degree of control and planning over innovation and value creation. In this view, one plausible
explanation for our result is that incumbent firms find it easier to open up these processes to their
traditional supply chain than to their end consumers, hence their supply-chain bias in their platform
initiatives. More research is however needed to better understand how incumbent firms should
proceed to adapt their culture to a world of digital platforms.
References
Bresnahan, T. (1990), New modes of competition: implications for the future structure of the computer
industry, in Eisenach, J. and T. Lenard (eds), Competition, Innovation and the Microsoft Monopoly: Antitrust in
the Digital Marketplace, Kluwer, pp. 155-208.
12
Brynjolfsson, E., L. Hitt and H. Kim (2011), Strength in Numbers: How does data-driven decision-making affect
firm performance?, SSRN Working Paper 1819486, available at
https://papers.ssrn.com/sol3/Papers.cfm?abstract_id=1819486.
Bughin J. (2016), Ten Years of Enterprise 2.0: the Power Law of Europe 2.0 Revisited, in Encyclopedia of ECommerce Development, Implementation, and Management, Second edition.
Bughin, J., L. LaBerge and A. Mellbye (2017), The case for digital reinvention, McKinsey Quarterly, January
2017, 1-15.
Bughin J. and N. van Zeebroeck (2017a), The case for offensive strategies in response to digital disruption,
Université libre de Bruxelles, iCite working paper number WP021-2017,
https://ideas.repec.org/p/ict/wpaper/2013-247037.html.
Bughin J. and N. van Zeebroeck (2017b), The Right Response to Digital Disruption, MIT Sloan Management
Review, Reprint #58479, available at http://sloanreview.mit.edu/article/the-right-response-to-digitaldisruption/.
Eisenmann, T, G. Parker and M. Van Alstyne (2006), Strategies for two-sided markets, Harvard Business Review
84(10), pp. 92-101.
Eisenmann, T, G. Parker and M. Van Alstyne (2011), Platform Envelopment, Strategic Management Journal 32,
pp. 1270-1285.
Evans and Gawer (2016), The Rise of the Platform Enterprise A Global Survey, New York: Center for Global
Enterprise.
Evans, D. and R. Schmalensee (2001), Some economic aspects of antitrust analysis in dynamically competitive
industries, NBER Working Paper 8268.
Evans, D. and R. Schmalensee (2007), The Catalyst Code: The Strategies Behind the World’s Most Dynamic
Companies, Harvard Business School Press.
Evans, D. and R. Schmalensee (2016), Matchmakers: The New Economics of Multisided Platforms, Harvard
Business Review Press.
Gawer, A. (2009), Platform dynamics and strategies: from products to services, in Gawer, A. (ed), Platforms,
markets and innovation, Edward Elgar, pp. 45-57.
Gawer, A. and M. Cusumano (2002), Platform Leadership: How Intel, Microsoft and Cisco Drive Industry
Innovation, Harvard Business School Press.
Henderson, R. and K. Clark (1990), Architectural innovation: the reconfiguration of existing product
technologies and the failure of established firms, Administrative Science Quarterly 35, pp. 9-30.
Johnson, M., C. Christensen and H. Kagermann (2008), Reinventing your business model, Harvard business
review 86(12), pp. 57-68.
Leijon, E., J. Svenheden and F. Svahn (2017), Platform Thinking in Incumbent Firms: From Concept to
Capability, in Proceedings of the 50th Hawaii International Conference on System Sciences.
Parker, G., M. Van Alstyne and S. Choudary (2016), Platform revolution: How networked markets are
transforming the economy and how to make them work for you, WW Norton & Company.
Rochet, J. and J. Tirole (2003), Platform competition in two-sided markets, Journal of the European Economic
Association 1(4), pp. 990-1029.
Shapiro, C. and H. Varian (1999), Information Rules, Harvard Business School Press.
Van Alstyne, M., G. Parker and S. Choudary (2016), 6 Reasons Platforms Fail, Harvard Business Review 31.
13
Tables
Table 4 - Determinants of platform play (logit)
Controls
Overall firm perception of digital shock
Perception
of digital
turbulence
0.1211**
(0.0565)
-0.0091
(0.0605)
0.0870
(0.0694)
-0.0629
(0.0661)
0.1440**
(0.0610)
Firm
strategy
Full model
Platform play
is top 3 priority
0.1238**
0.1113***
(0.0576)
(0.0403)
Firm perception of digital shock in distribution channels
-0.0254
0.0008
(0.0616)
(0.0440)
Firm perception of digital shock in competitive landscape
0.0763
0.0527
(0.0704)
(0.0498)
Firm perception of digital shock in core operations
-0.0593
0.0017
(0.0667)
(0.0472)
Firm perception of digital shock in supply-chain
0.1223*
0.1153***
(0.0625)
(0.0447)
Higher business divestures
-0.2022
-0.2938
-0.1340
(0.3261)
(0.3402)
(0.2290)
Higher business acquisitions
-0.2249
-0.2796
-0.0696
(0.2333)
(0.2428)
(0.1687)
Higher frequency of corporate portfolio revisions
-0.0485
-0.0498
-0.1615
(0.1580)
(0.1640)
(0.1167)
Higher resources shifting across businesses
0.0315
-0.0014
0.0712
(0.1604)
(0.1673)
(0.1175)
Higher investment into portfolio
0.2302
0.1797
0.1237
(0.1591)
(0.1666)
(0.1196)
Higher changes to risk profile and time horizon
0.1862
0.1215
0.2662**
(0.1756)
(0.1839)
(0.1329)
More resources to understand customers' behaviors and needs
-0.0667
-0.0557
-0.1517
(0.1565)
(0.1620)
(0.1131)
Adaptation of offerings to customers’ needs and preferences
-0.3586**
-0.3560**
-0.2120*
(0.1565)
(0.1634)
(0.1138)
Digital metrics included in performance management system
-0.0639
-0.0840
-0.1686
(0.1717)
(0.1781)
(0.1246)
New relationships created with external business partners
0.0933
0.1304
0.3360***
(0.1559)
(0.1617)
(0.1123)
New models of sharing profits and value with external businesses
0.1911
0.1385
0.3972***
(0.1892)
(0.1944)
(0.1413)
Increased speed of businesses and functions operations
0.4360***
0.4358***
0.2211*
(0.1563)
(0.1637)
(0.1167)
Proactive adaptation of business model, even at risk of cannibalization
0.1887
0.1488
0.3009**
(0.1690)
(0.1766)
(0.1254)
Best people and resources reallocated to digital initiatives
0.3006*
0.1697
0.1112
(0.1647)
(0.1736)
(0.1243)
Firm is public
0.2032
0.0908
0.2269
0.1162
0.0162
(0.1842)
(0.1929)
(0.1860)
(0.1943)
(0.1393)
Firm is large (Rev>1b$)
-0.0542
0.0263
-0.0741
0.0184
0.1127
(0.1892)
(0.1981)
(0.1920)
(0.2009)
(0.1416)
Firm's main focus is B2C
-0.3240*
-0.3209
-0.2902
-0.2838
-0.2013
(0.1924)
(0.1988)
(0.1948)
(0.2013)
(0.1343)
Firm portfolio is mono-product or mono-service
-0.2105
-0.0901
-0.1806
-0.1004
-0.1459
(0.1962)
(0.2050)
(0.1997)
(0.2082)
(0.1414)
Firm portfolio includes products
0.1274
0.1409
0.0904
0.1225
-0.1380
(0.1538)
(0.1614)
(0.1563)
(0.1635)
(0.1149)
Constant
-2.1687***
-2.8081***
-2.2393***
-2.7468***
-1.5041***
(0.5442)
(0.5833)
(0.5535)
(0.5905)
(0.4113)
Log Likelihood
-741.07
-679.00
-725.67
-668.61
-1156.38
Pseudo-R²
0.03
0.04
0.05
0.06
0.07
Industry F.E.
Y
Y
Y
Y
Y
Region F.E.
Y
Y
Y
Y
Y
N
2,122
1,959
2,122
1,959
1,964
Dependent variable is incidence of platform play. Standard errors in parentheses. Coefficient significant at the (*) 10%, (**) 5%, or (***) 1% levels.
The reference is a private firm with revenues smaller than $1 billion with a main focus on B2B, and multiple services in its portfolio.
Platform play
includes access to
new sources of
supply
0.1123**
(0.0535)
-0.0240
(0.0568)
0.0939
(0.0649)
-0.0483
(0.0621)
0.0910
(0.0579)
-0.3839
(0.3202)
-0.2518
(0.2247)
-0.0322
(0.1519)
0.0965
(0.1545)
0.1935
(0.1550)
0.1551
(0.1706)
-0.2201
(0.1522)
-0.3449**
(0.1523)
0.0371
(0.1635)
0.2081
(0.1500)
0.3560**
(0.1748)
0.3405**
(0.1533)
0.2404
(0.1628)
0.0914
(0.1627)
0.1362
(0.1825)
0.0349
(0.1884)
-0.0984
(0.1820)
0.1520
(0.1826)
0.0160
(0.1508)
-2.5558***
(0.5490)
-747.25
0.06
Y
Y
1,959
14
Table 5 - Impact of platform play on performance (revenue growth) (OLS)
1
2
3
4
5
6
7
Firm plays platform game
0.0563
-0.0232
-0.2612
0.0839
-0.2625
-0.5480*
(0.1746)
(0.1777)
(0.2067)
(0.1734)
(0.2734)
(0.2928)
Reaction is medium
0.4380***
0.4501***
0.4501***
0.4584***
(0.1374)
(0.1373)
(0.1365)
(0.1395)
Reaction is strong in only one dimension
0.6672***
0.6856***
0.6750***
0.6661***
(0.1400)
(0.1403)
(0.1396)
(0.1447)
Reaction is bold at scale
1.2384***
1.0686***
1.0725***
1.0717***
(0.1751)
(0.1864)
(0.1854)
(0.1865)
Platform play X Bold-at-scale
0.9948**
1.0288***
(0.3884)
(0.3899)
Firm has strong demand-side focus
0.2269**
0.1452
0.1462
0.1418
(0.1045)
(0.1096)
(0.1149)
(0.1149)
Platform play X Demand focus
0.6593*
0.5911*
(0.3463)
(0.3450)
Firm is public
-0.2195*
-0.2276*
-0.2189*
-0.2165*
-0.2158*
-0.2191*
-0.2694**
(0.1215)
(0.1226)
(0.1233)
(0.1217)
(0.1213)
(0.1236)
(0.1303)
Firm is large (Rev>1b$)
-0.3158***
-0.1789
-0.1817
-0.3249***
-0.3298***
-0.1922
-0.1634
(0.1183)
(0.1189)
(0.1195)
(0.1184)
(0.1181)
(0.1196)
(0.1254)
Firm's main focus is B2C
0.0871
0.1237
0.1317
0.0823
0.0847
0.1317
0.1493
(0.1156)
(0.1201)
(0.1197)
(0.1154)
(0.1154)
(0.1194)
(0.1226)
Firm portfolio is mono-product or mono-service
0.2156
0.4479***
0.4489***
0.2300*
0.2197
0.4507***
0.4079***
(0.1340)
(0.1440)
(0.1437)
(0.1343)
(0.1344)
(0.1440)
(0.1538)
Firm portfolio includes products
-0.0997
-0.0958
-0.0888
-0.0926
-0.0935
-0.0839
-0.1554
(0.1053)
(0.1094)
(0.1091)
(0.1052)
(0.1052)
(0.1090)
(0.1154)
Constant
-0.1267
-0.8440***
-0.8456***
-0.2797
-0.2220
-0.9389***
-0.8297***
(0.1668)
(0.1990)
(0.1983)
(0.1820)
(0.1825)
(0.2141)
(0.2263)
F-Stat
7.7
10.4
10.1
7.6
7.4
9.4
8.6
Adjusted R²
0.03
0.06
0.06
0.03
0.03
0.06
0.05
N
1,952
1,719
1,719
1,952
1,952
1,719
1,511
Industry F.E.
N
N
N
N
N
N
N
Region F.E.
Y
Y
Y
Y
Y
Y
Y
Dependent variable is incidence of platform play. Standard errors in parentheses. Coefficient significant at the (*) 10%, (**) 5%, or (***) 1% levels.
The reference is a private firm with revenues smaller than $1 billion with a main focus on B2B, and multiple services in its portfolio.
8
0.2397
(0.6853)
0.7422
(0.6546)
2.1595***
(0.6609)
0.7168**
(0.3349)
0.1824
(0.4015)
-0.4529
(0.4234)
0.1131
(0.5028)
0.7281*
(0.4326)
0.4974
(0.3819)
-2.2628***
(0.7726)
2.9
0.09
208
N
Y
15
Table 6 - Impact of platform play on performance (EBIT rate growth) (OLS)
1
2
3
4
5
6
7
Firm plays platform game
-0.0832
-0.1893
-0.1247
-0.0461
-0.4748*
-0.4928
(0.1777)
(0.1851)
(0.2039)
(0.1764)
(0.2706)
(0.3003)
Reaction is medium
0.3519**
0.3485**
0.3503**
0.3199**
(0.1439)
(0.1441)
(0.1435)
(0.1476)
Reaction is strong in only one dimension
0.3653**
0.3602**
0.3473**
0.2816*
(0.1498)
(0.1502)
(0.1496)
(0.1552)
Reaction is bold at scale
0.8947***
0.9426***
0.9479***
0.9082***
(0.1854)
(0.1911)
(0.1889)
(0.1898)
Platform play X Bold-at-scale
-0.2759
-0.2342
(0.4737)
(0.4593)
Firm has strong demand-side focus
0.2995***
0.1960*
0.2213*
0.2247*
(0.1106)
(0.1171)
(0.1227)
(0.1227)
Platform play X Demand focus
0.8255**
0.7824**
(0.3521)
(0.3672)
Firm is public
0.0190
-0.0070
-0.0091
0.0209
0.0216
-0.0115
-0.0724
(0.1318)
(0.1322)
(0.1323)
(0.1318)
(0.1315)
(0.1325)
(0.1400)
Firm is large (Rev>1b$)
-0.2003
-0.0677
-0.0662
-0.2190*
-0.2250*
-0.0894
-0.0968
(0.1284)
(0.1282)
(0.1282)
(0.1285)
(0.1283)
(0.1284)
(0.1353)
Firm's main focus is B2C
0.0973
0.0644
0.0627
0.0941
0.0967
0.0655
0.1312
(0.1200)
(0.1248)
(0.1247)
(0.1197)
(0.1194)
(0.1240)
(0.1310)
Firm portfolio is mono-product or mono-service
-0.1127
0.0593
0.0593
-0.0930
-0.1075
0.0628
0.0798
(0.1439)
(0.1581)
(0.1582)
(0.1438)
(0.1442)
(0.1579)
(0.1623)
Firm portfolio includes products
-0.2172*
-0.1922
-0.1946*
-0.2084*
-0.2088*
-0.1876
-0.2342*
(0.1125)
(0.1177)
(0.1175)
(0.1123)
(0.1122)
(0.1171)
(0.1237)
Constant
-0.0479
-0.5456**
-0.5456**
-0.2488
-0.1739
-0.6857***
-0.6023**
(0.1761)
(0.2158)
(0.2160)
(0.1902)
(0.1940)
(0.2313)
(0.2422)
F-Stat
2.5
3.4
3.5
3.0
3.2
4.0
4.3
Adjusted R²
0.01
0.02
0.02
0.01
0.01
0.03
0.03
N
1,807
1,584
1,584
1,807
1,807
1,584
1,387
Industry F.E.
N
N
N
N
N
N
N
Region F.E.
Y
Y
Y
Y
Y
Y
Y
Dependent variable is incidence of platform play. Standard errors in parentheses. Coefficient significant at the (*) 10%, (**) 5%, or (***) 1% levels.
The reference is a private firm with revenues smaller than $1 billion with a main focus on B2B, and multiple services in its portfolio.
8
0.8283
(0.6601)
1.0795*
(0.6233)
1.2975*
(0.6604)
0.9300**
(0.3621)
0.3729
(0.4634)
-0.1562
(0.4296)
-0.2023
(0.4259)
-0.1352
(0.5662)
0.1789
(0.3937)
-2.0364***
(0.7729)
1.8
0.02
197
N
Y
16
Table 7 - Heckman Two-Stage Regression (Second Stage)
Revenue
Growth
Reaction is medium
0.2712
(0.6032)
Reaction is strong in only one dimension
0.7767
(0.6041)
Reaction is bold at scale
2.1468***
(0.6354)
Firm has strong demand-side focus
0.8122**
(0.3329)
Firm is public
0.0161
(0.4909)
Firm is large (Rev>1b$)
-0.3150
(0.4951)
Firm's main focus is B2C
-0.0626
(0.4307)
Firm portfolio is mono-product or mono-service
0.5798
(0.5202)
Firm portfolio includes products
0.5097
(0.3674)
Constant
-2.9234**
(1.3468)
Industry F.E.
N
Region F.E.
Y
Lambda
0.38
Lambda S.E.
0.67
N
1,931
N (uncensored)
198
EBIT
Growth
0.8386
(0.6384)
1.0560*
(0.6361)
1.3138*
(0.6735)
0.9633***
(0.3529)
0.2586
(0.5141)
-0.0218
(0.5154)
-0.3029
(0.4542)
-0.3074
(0.5376)
0.2172
(0.3912)
-2.5106*
(1.4239)
N
Y
0.36
0.69
1,922
198
17
Table 8 - Heckman Two-Stage Regression (Selection Equation)
Revenue
Growth
Overall firm perception of digital shock
0.1131***
(0.0334)
Firm perception of digital shock in distribution channels
-0.0403
(0.0351)
Firm perception of digital shock in competitive landscape
0.0379
(0.0404)
Firm perception of digital shock in core operations
-0.0256
(0.0374)
Firm perception of digital shock in supply-chain
0.0608*
(0.0353)
Higher business divestures
-0.0413
(0.1771)
Higher business acquisitions
-0.1840
(0.1364)
Higher frequency of corporate portfolio revisions
0.0104
(0.0909)
Higher resources shifting across businesses
0.0084
(0.0934)
Higher investment into portfolio
0.1573*
(0.0927)
Higher changes to risk profile and time horizon
0.0613
(0.1034)
More resources to understand customers' behaviors and needs
-0.0311
(0.0897)
Adaptation of offerings to customers’ needs and preferences
-0.1589*
(0.0908)
Digital metrics included in performance management system
-0.0092
(0.0988)
New relationships created with external business partners
0.0479
(0.0904)
New models of sharing profits and value with external businesses
0.0952
(0.1105)
Increased speed of businesses and functions operations
0.2138**
(0.0923)
Proactive adaptation of business model, even at risk of cannibalization
0.0900
(0.0985)
Best people and resources reallocated to digital initiatives
0.1043
(0.0977)
Firm is public
0.0759
(0.1114)
Firm is large (Rev>1b$)
0.0045
(0.1142)
Firm's main focus is B2C
-0.0748
(0.1082)
Firm portfolio is mono-product or mono-service
-0.0790
(0.1180)
Firm portfolio includes products
0.1001
(0.0922)
Constant
-1.8606***
(0.3419)
Industry F.E.
Y
Region F.E.
Y
N
1,931
N (uncensored)
198
EBIT
Growth
0.1074***
(0.0339)
-0.0686*
(0.0365)
0.0492
(0.0411)
-0.0064
(0.0380)
0.0664*
(0.0358)
0.0134
(0.1775)
-0.1558
(0.1368)
0.0371
(0.0922)
-0.0331
(0.0955)
0.1752*
(0.0940)
0.0443
(0.1052)
-0.0520
(0.0916)
-0.1358
(0.0923)
-0.0166
(0.1004)
0.0317
(0.0924)
0.0965
(0.1125)
0.2602***
(0.0933)
0.0920
(0.1003)
0.0800
(0.0998)
0.0887
(0.1125)
0.0160
(0.1154)
-0.0461
(0.1096)
-0.0543
(0.1191)
0.0776
(0.0937)
-1.9071***
(0.3446)
Y
Y
1,922
198
18
Reaction is medium
Reaction is strong in only one dimension
Reaction is bold at scale
Firm has strong demand-side focus
Firm plays platform game
Platform play X Bold-at-scale
Platform play X Demand focus
1
0.4548***
(0.1363)
0.6822***
(0.1390)
1.0263***
(0.2053)
0.0199
(0.1301)
-0.4734**
(0.1911)
0.6061**
(0.3057)
0.5246**
(0.2310)
2
0.3456**
(0.1433)
0.3528**
(0.1490)
0.9053***
(0.2065)
0.2014
(0.1399)
-0.2481
(0.2040)
-0.0191
(0.3316)
0.3340
(0.2465)
3
0.4384***
(0.1366)
0.6597***
(0.1399)
1.1210***
(0.1842)
0.1421
(0.1163)
-0.3403
(0.2659)
0.5896
(0.3930)
0.5534*
(0.3235)
Table 9 – Robustness estimates
4
5
6
0.3404**
0.0652
0.0825
(0.1437)
(0.3122)
(0.3559)
0.3340**
0.2242
0.2119
(0.1504)
(0.3075)
(0.3666)
0.9751***
0.8306**
0.8242**
(0.1881)
(0.3631)
(0.3979)
0.2273*
0.0175
-0.0808
(0.1250)
(0.2756)
(0.3185)
-0.2877
-0.2032
-1.1571
(0.2681)
(0.5215)
(0.7226)
-0.4320
1.8298**
-0.5779
(0.4455)
(0.9213)
(1.5206)
0.6428*
1.3890
2.6589***
(0.3356)
(0.8784)
(0.7919)
7
0.9346*
(0.5098)
1.6318***
(0.4760)
2.1269***
(0.5136)
0.0675
(0.2915)
-0.9986*
(0.5592)
0.9623
(0.7011)
1.1623*
(0.6679)
8
0.8870*
(0.5081)
1.2437**
(0.4844)
1.9286***
(0.5107)
0.0594
(0.3046)
-1.0975*
(0.6408)
-0.1514
(0.7483)
1.5120**
(0.7297)
-0.2139*
(0.1221)
-0.1985*
(0.1193)
0.1276
(0.1200)
0.4377***
(0.1446)
-0.0903
(0.1091)
-0.8062***
(0.2170)
-0.0024
(0.1318)
-0.0910
(0.1288)
0.0699
(0.1249)
0.0753
(0.1578)
-0.1857
(0.1178)
-0.6799***
(0.2353)
-0.2247*
(0.1231)
-0.1916
(0.1194)
0.1271
(0.1203)
0.4482***
(0.1441)
-0.0919
(0.1091)
-0.9345***
(0.2166)
-0.0160
(0.1323)
-0.0912
(0.1283)
0.0641
(0.1243)
0.0671
(0.1581)
-0.1879
(0.1172)
-0.6928***
(0.2336)
-0.8904***
(0.3400)
-0.0844
(0.3296)
-0.3631
(0.4960)
0.4551
(0.4018)
-0.2377
(0.2784)
-1.5643**
(0.6383)
-0.3230
(0.3575)
-0.0170
(0.3420)
-0.3188
(0.5138)
-0.6316
(0.4149)
-0.2706
(0.2923)
-1.0645
(0.6919)
Platform Play X CEO Sponsorship
Platform Play X CFO Sponsorship
Platform Play X CMO Sponsorship
Firm is public
Firm is large (Rev>1b$)
Firm's main focus is B2C
Firm portfolio is mono-product or mono-service
Firm portfolio includes products
Constant
-0.2619
(0.2783)
-0.4642*
(0.2667)
-0.0267
(0.2431)
0.0875
(0.3019)
-0.3414
(0.2669)
-0.3995
(0.4661)
0.1463
(0.2891)
-0.5169*
(0.2690)
-0.3505
(0.2954)
-0.6712*
(0.4006)
-0.4206
(0.3174)
-0.0925
(0.5535)
9
0.4431***
(0.1363)
0.6597***
(0.1399)
1.0650***
(0.1855)
0.1464
(0.1150)
-0.8123**
(0.3152)
1.0017**
(0.4010)
0.5762*
(0.3415)
0.1444
(0.3773)
-0.2055
(0.4391)
0.7110**
(0.3579)
-0.2110*
(0.1235)
-0.2039*
(0.1196)
0.1268
(0.1183)
0.4463***
(0.1441)
-0.0944
(0.1092)
-0.9068***
(0.2136)
10
0.3398**
(0.1435)
0.3238**
(0.1503)
0.9361***
(0.1891)
0.2220*
(0.1228)
-0.7851**
(0.3205)
-0.3416
(0.4731)
0.7520**
(0.3641)
0.3755
(0.4045)
0.3347
(0.4534)
0.1793
(0.3477)
-0.0131
(0.1323)
-0.0951
(0.1281)
0.0527
(0.1242)
0.0617
(0.1583)
-0.1883
(0.1172)
-0.6534***
(0.2320)
Constant (sigma)
(Pseudo-)R2
Industry F.E.
Region F.E.
N
0.07
0.03
0.07
0.04
0.12
0.10
0.16
0.10
0.08
N
N
N
N
N
N
N
N
N
Y
Y
Y
Y
Y
Y
Y
Y
Y
1,719
1,584
1,719
1,584
286
261
331
304
1,719
Dependent variable is incidence of platform play. Standard errors in parentheses. Coefficient significant at the (*) 10%, (**) 5%, or (***) 1% levels.
The reference is a private firm with revenues smaller than $1 billion with a main focus on B2B, and multiple services in its portfolio.
All OLS estimates, except columns 11 and 12 (Tobit)
0.04
N
Y
1,584
11
0.4517***
(0.1361)
0.6763***
(0.1394)
1.0752***
(0.1857)
0.1474
(0.1146)
-0.5514*
(0.2918)
1.0465***
(0.3923)
0.5998*
(0.3447)
12
0.3484**
(0.1429)
0.3443**
(0.1491)
0.9460***
(0.1881)
0.2216*
(0.1221)
-0.4926*
(0.2987)
-0.2344
(0.4568)
0.7821**
(0.3653)
-0.2174*
(0.1231)
-0.1909
(0.1192)
0.1331
(0.1195)
0.4587***
(0.1441)
-0.0866
(0.1088)
-0.9432***
(0.2133)
2.1024***
(0.0457)
0.02
N
Y
1,719
-0.0084
(0.1323)
-0.0911
(0.1280)
0.0635
(0.1237)
0.0628
(0.1571)
-0.1874
(0.1165)
-0.6841***
(0.2302)
2.1310***
(0.0451)
0.01
N
Y
1,584
19
WORKING PAPERS 2013
001 - Exploring europe’s r&d deficit relative to the us: differences in the rates of return to r&d of young leading r&d firms - Michele Cincera and Reinhilde
Veugelers
002 - Governance typology of universities’ technology transfer processes - A. Schoen, B. van Pottelsberghe de la Potterie, J. Henkel.
003 - Academic Patenting in Belgium: Methodology and Evidence – M. Mejer.
004 - The impact of knowledge diversity on inventive performance at European universities – M. Mejer
005 - Cross-Functional Knowledge Integration, Patenting and Firm’s Performance – M. Ceccagnoli, N. van Zeebroeck and R. Venturini.
006 - Corporate Science, Innovation and Firm Value, M. Simeth and M. Cincera
WORKING PAPERS 2014
007 - Determinants of Research Production at top US Universities – Q. David
008 - R&D financing constraints of young and old innovation leaders in the EU and the US – M. Cincera, J. Ravet and R. Veugelers
009 - Globalization of Innovation Production; A Patent-Based Industry Analysis – J. Danguy
010 - Who collaborates with whom: the role of technological distance in international innovation – J. Danguy
WORKING PAPERS 2015
011 - Languages, Fees and the International Scope of Patenting – D. Harhoff , K. Hoisl, B. van Pottelsberghe de la Potterie , C. Vandeput
012 – How much does speed matter in the fixed to mobile broadband substitution in Europe? – M. Cincera, L. Dewulf, A. Estache
013 – VC financing and market growth – Interdependencies between technology-push and market-pull investments in the US solar industry – F. Schock, J.
Mutl, F. Täube, P. von Flotow
014 – Optimal Openness Level and Economic Performance of Firms: Evidence from Belgian CIS Data – M. Cincera, P. De Clercq, T. Gillet
015 – Circular Causality of R&D and Export in EU countries – D. Çetin, M. Cincera.
016 – Innovation and Access to Finance – A Review of the Literature – M. Cincera, A. Santos.
WORKING PAPERS 2016
017 – Effectiveness of Government intervention in the SME sector: Evidence from the Brussels-Capital Region – G. E. Fombasso, M. Cincera.
018 – A review of corporate R&D intensity decomposition – P. Moncada-Pastemò-Castello.
019 – The laws of action and reaction: on determinants of patent disputes in European chemical and drug industries – R. Kapoor, N. van Zeebroeck.
020– How do the normativity of headquarters and the knowledge autonomy of subsidiaries co-evolve? – M. Hansmans, G. Liu.
WORKING PAPERS 2017
021– The case for offensive strategies in response to digital disruption – J. Bughin, N. van Zeebroeck.
022– Access to finance as a pressing problem: Evidence from innovative European firms – A. Santos, M. Cincera.
023–Platform play among incumbent firms: the wrong focus? – N. van Zeebroeck, J. Bughin
20