Content uploaded by Paul Belleflamme
Author content
All content in this area was uploaded by Paul Belleflamme on Jul 14, 2016
Content may be subject to copyright.
Forum
CESifo DIC E Repor t 2/2016 (June) 6
Understanding the strategies
of crowdfUnding platforms1
paUl Belleflamme,2 nessrine omrani,3
and martin peitz4
Crowdfunding can be seen as an open call made through
the internet to provide nancial resources to support
new ventures. Several forms of crowdfunding coexist,
which mainly differ by the type of compensation that
they propose to funders. Compensation can be mone-
tary or may take another form. In the case of monetary
compensation, funders are investors and may be offered
equity stakes (‘crowdinvesting’), interest payments
(‘crowdlending’), or a fraction of prots (‘royalty-based
crowdfunding’). When compensation is not monetary
funders are consumers or donors and may be offered
a product in pre-sale, combined with some perks (‘re-
ward-based crowdfunding’), or some warm glow (‘do-
nation-based crowdfunding’).
Whatever its form, crowdfunding mostly takes place
on crowdfunding platforms (CFPs). Our objective in
this ar ticle is twofold: we wish to show why and how
CFPs facilitate the interaction between ent repreneurs
trying to raise funds (the ‘fundraisers’) and consumers/
investors willing to participate in the nancing of new
projects (the ‘funders’). The ‘why’ has to do with the
external effects that crowdfunding generates, not only
across the groups of funders and fundraisers, but also
within each of these groups; we argue that the complex-
ity of these effects is much more efciently dealt with
by a CFP than through bilateral relationships between
funders and fundraisers. As for the ‘how’, we present the
strategies that CFPs put in place to address the problems
raised by the various external effects.
1
This a rticle wa s prepa red for the CESifo DICE Repor t; it is la rgely
base d on Belleam me, Omra ni an d Peitz (2015).
2 Un iversité ca tholique de Louvain (COR E and Lou vain School of
Manage ment) and CESifo.
3
PSB Paris S chool of Busin ess.
4 Universit y of Mann heim, Mann heim Centre for Com petitio n and
Innovation (MaC CI), and CE RRE .
Crowdfunding and external effects
CFPs can be seen as ‘two-sided platforms’: they enable
the interaction between two ‘sides’ (here, fundraisers
and funders) whose demands need to be coordinated.5
The open and large-scale nature of crowdfunding ex-
plains why an intermediary (i.e., a CFP) can achieve this
coordination more efciently – i.e., at lower transaction
costs – than the members of the two sides by themselves.
The transaction costs stem from the presence of exter nal
effects across the two groups: the value that each group
attaches to the interaction depends on the participation
of the other group. Typically, the more funders partic-
ipate, the more crowdfunding becomes attractive for
fundraisers, and vice versa. We therefore expect these
so-called ‘cross-group external effects’ to be positive on
both sides; we also expect CFPs to manage these exter-
nal effects by choosing an appropriate price structure
for access and participation on the platfor m by the two
groups.
As we now show, things are slightly more complex.
Firstly, some cross-group external effects may be nega-
tive. Secondly, ‘within-group external effects’ may also
exist, according to which the value that a user attaches
to the interaction with the other g roup also depends on
the participation within this user’s own group. Finally,
CFPs also use a wide array of non-price strategies to
manage the various external effects.
Cross-group external effects on CFPs
Cross-group external effects arise when one group’s
valuation of the platform depends on the participation
of the other group. Let us rst examine the impacts of
funders’ participation on fundraisers. These effects are
positive without any ambiguity. A platform that attracts
a larger pool of potential funders benets fundraisers
in two ways. Firstly, and quite obviously, the presence
of a larger crowd of funders increases any fundraiser’s
chance of nancing their project. Secondly, entrepre-
neurs often use CFPs as marketing channels to evalu-
5 For an intro duction to two-sided platforms, see Bellea mme and
Peitz (2015, Ch apte r 22).
Forum
CESifo DIC E Repor t 2/2016 (June)
77
ate (and possibly stimulate) demand for their product.6
Again, the larger the crowd of consumers/investors, the
more efcient this market testing can be. Yet, to the ex-
tent that fundraisers use a project on a CFP as part of a
price discrimination strategy, they may not always be
interested in an expansion of the pool of funders. A nec-
essary condition for this to happen is that not only the
size, but also the composition of the group of funders on
a CFP will have to change.7
As for the impacts of fundraisers’ participation on
funders, it seems at rst glance that they are also pos-
itive. We can indeed see two reasons why funders are
likely to prefer platforms with a larger number of fund-
raisers: rstly, platforms with more fundraisers provide
funders with a wider choice of projects to fund; sec-
ondly, when compensation is non-monetary, funders
are more likely to obtain rewards that t their tastes on
platforms that attract many f undraisers. A couple of
mitigating factors, however, spring to mind. Firstly, a
larger number of campaigns on a platform could reduce
the chance that any of them would be successful (i.e.,
would reach the required threshold), which would affect
funders negatively. Yet, as such risk results from coor-
dination failures among funders, CFPs can neutralise
it by guiding interested funders to campaigns that are
close to becoming a success. A second negative cross-
group external effect arises if asym metric information
problems become more serious as the number of fund-
raisers on a platform grows larger. As discussed below,
information asymmetries cause two types of issues for
funders on CFPs: funders may not only lack the nec-
essary information to assess the chances of success of
the proposed campaigns; they may also not be able to
control how fundraisers use the funds that they have
collected. W hether these problems are more acute (or
the cost for funders and platforms of alleviating these
problems is larger) on platforms with more fundraisers
is an empirical question.
6
The fact t hat large compan ies, which h ave easy ac cess to ca pital
market s, use CFPs is indi rect proof that crowdfund ing is not ju st about
fund ing. For instance, Sony has launched a CF P, First Flight, to te st the
popula rity (rat her t han to nance) its own n ew products (McCormick
2015) .
7
If fun ders di ffer in t heir in format ional require ments, certa in CFPs
may speci alise and cater to the need s of particula r types of fun ders. For
instance, by setti ng h igher prices o r by providing speci c s ervices, a
CFP may in duce self-s election among fu ndraisers a nd attract on ly high
qualit y project s, which, in turn, wi ll att ract on ly the more demandi ng
funders. Damiano and Li (2008) show in a styli sed set ting how a pla t-
form can u se prices to induce this for m of segmentation .
Within-group external effects on CFPs
Fundraisers and f unders also care about the partici-
pation of their own group members on the platform.
Within the group of fundraisers, external effects are
mostly negative: as fundraisers compete for funders’
contributions, the more campaigns a platform hosts,
the tougher the competition. However, a larger pool of
fellow fundraisers may promote the exchange of good
practices among them, or may attract a larger supply
of consultancy services adapted to crowdfunding. This
suggests that external effects may sometimes be posi-
tive within the group of fundraisers.8
Within the group of funders, external effects can be ex-
pected to be positive. This is certainly the case if a pro-
ject needs to reach a pre-specied threshold of nancing
to be car ried out. This is known as the ‘threshold-pledge’
or ‘All-or-Nothing’ (AON) model (an alternative is the
‘exible funding’ or ‘Keep-it-all’ (KIA) model, which
allows a fundraiser to collect any funds raised, even
when the target is not reached). In this case, the presence
of additional funders on a platform increases the proba-
bility that some project will be realised, which benets
all funders.
Other external effects may also come into play within
the group of funders. They result from the sequential
process that funding follows on CFPs, which induces
a form of dynamic behaviour among funders. We have
already mentioned that asymmetries of information pre-
vail on CFPs as, t ypically, funders have little informa-
tion about the reliability of fundraisers and the quality
of their projects. Because funding is sequential, funders
may try to infer information from the behaviour of fel-
low funders (even if the latter do not possess better in-
formation to start with). In particular, funders may rely
on the existing suppor t for a given project to gauge its
potential, thereby creating a type of peer-effect known
as ‘collective attention effect’. The sign of this effect
– i.e., whether it ultimately benets or harms funders
– depends on the rst funders. For instance, if the rst
funders had poor information or made a bad decision,
herding will lead subsequent funders to back the wrong
horse.9 There are reasons to believe that this scenario is
relevant in practice, as funders lack the capabilities and
the incentives to devote the appropriate resources to due
8 Strict ly speak ing, there are p ositive feed back effects bet ween fu nd-
raiser s and CFP-spec ic provider s of consulting service who constitut e
a third side of the platfo rm.
9
For evidence of herding on crowdlend ing platforms, s ee Chen and
Lin (2014), Lee and Le e (2012), Zhang and Li u (2012).
Forum
CESifo DIC E Repor t 2/2016 (June) 8
diligence. There is thus a ‘collective-action problem’ in
that funders nat urally tend to free ride on fellow funders
to collect information about the fundraisers’ chances of
success. Finally, another form of free-riding may cre-
ate a negative external effect: in the AON model, when
the nancing of a project comes close to the threshold,
it may become harder to induce funders to provide the
remaining nancing as they may rely on other f unders
to do it.10
Figure 1 summarises the various cross-group and with-
in-group external effects that crowdfunding generates.
Strategies of CFPs
The presence of strong and intert wined cross-group
and within-group external effects in crowdfunding
limits the ability of fundraisers and funders to conduct
transactions bilaterally in an efcient way. This creates
business opportunities for intermediation, which CFPs
try to seize by designing adequate strategies. These
strategies aim at creating value for the two groups by
driving agents to ‘internalise’ (i.e., to integrate into their
decision-making process) the effects that their actions
have across groups or within their group. Naturally, to
achieve a protable business model, CFPs must nd
ways to capture a sufcient share of the value that they
create for their users. We consider in tur n price and non-
price st rategies.
10
Kuppuswamy and Bayus (2013) nd that a dditional fu nder support
for a project on Kickstarter (a reward-based CFP) is negatively related
to its in itial suppor t, which is consist ent with this f ree-rider hy pothesis.
Price strategies
Currently, most CFPs charge
only one group or impose a “tax”
on a successful transaction. The
common practice is to charge a
transaction fee to f undraisers as
a percent basis for all successful
campaigns (unsuccessful cam-
paigns are generally not taxed).
As for funders, they usually do not
pay any explicit fee. Yet, insofar
as time elapses between the mo-
ment funders contribute money
and the moment that this money
is either passed on to fundraisers
(when the campaign is successful)
or returned to the funders (if it is
not), funders incur a foregone in-
terest when investing (early) in a project, which can be
seen as an implicit fee.
This reliance on transaction fees is common on two-sid-
ed platforms, especially in new markets where par tici-
pants have little understanding of the value they attach
to the interaction with the other group(s). The reason is
that imposing subscription fees may scare participants
away, which would jeopardize the launch of the plat-
form. Pushing this logic one step further, participation
may even be subsidised for some participants, espe-
cially in the early life of the platform. For instance, it
is not rare that CFPs do not wait for fundraisers to join
the platform, but actively look for the most interesting
among them and subsequently facilitate their campaigns
on the platform.
On top of the tax levied on transactions, CFPs usually
have two other sources of revenues. Firstly, as explained
above, they earn interest on the money pledged by
funders. Secondly, they may also offer additional pay-
ing services to the t wo groups; for instance, CFPs may
charge for handling payments, for suppor ting projects,
or for releasing information on previous projects.
Non-price strategies
CFPs also use non-price strategies to manage cross-
group external effects. First and foremost, CFPs have
to choose a mechanism for raising funds. As described
above, the choice is primarily between the ‘All-or-
Nothing’ (AON) and ‘Keep-it-all’ (KIA) mechanisms.
Fundraisers Funders
Crowdfunding
platform
☺Higher chances to fund a project
☺Better market testing
Wider set of projects to fund
ܫLower chances of success for any campaign
☺Larger supply of
consultancy services
ܫMore
competition
Better fit of rewards
☺Exchange of good
practices
ܫLarger information asymmetry
☺Higher chances
projects get
funded
ܫFree-riding
in due diligence
External effects on crowdfunding platforms
Authors‘ ’ contribution. Source:
☺
☺
Figure 1
Forum
CESifo DIC E Repor t 2/2016 (June)
99
In the AON model, fundraisers rstly have to specify
a target, k nowing that they will not receive any of the
money that has been pledged if this target is not reached.
Although this mechanism may not seem terribly attrac-
tive for fundraisers (compared to the KIA model where
any money pledged can be kept), it has the advantage of
protecting funders as it drives fundraisers to set realis-
tic funding targets that match more closely the funding
that they need to achieve their project. As cross-group
external effects from funders to fundraisers are gener-
ally positive, choosing the AON mechanism to reassure
funders is an indirect way to make the platform more
attractive for fundraisers.
Another indirect advantage of the AON model for fund-
raisers is that it makes some funders ‘pivotal’ insofar as
it is their contribution that enables total funding to reach
the target. This is especially important in the context of
reward-based crowdfunding, whereby funders receive
the project’s product as compensation for their funding.
Fundraisers are then in a position to raise their prots by
charging different prices for their product to consumers/
investors who are pivotal and to those who are not.11
That said, fundraisers may prefer the exibility of the
KIA model (even although CFPs usually charge higher
fees on funds kept by fundraisers in cases where the tar-
get is not reached). If so, choosing AON would discour-
age fundraisers from joining the platform and, through
cross-group external effects, would discourage funders
as well. An alternative is to propose both models and let
fundraisers choose. One benet of this solution is that
it gives funders the option of drawing inferences from
fundraisers’ choices. Indeed it is documented that by
choosing AON, fundraisers credibly signal to funders
that they com mit not to undertake their project if they
do not reach the target; funders may then see the invest-
ment in such projects as less “risky”, which allows fund-
raisers to increase their chances of success.12 The choice
between AON and KIA should largely be driven by the
extent to which the benet from a project depends on the
funding level. If the use of funds below the threshold
level is highly inefcient, AON should be the preferred
funding format.
Apart from managing cross-group effects, CFPs also
design specic strategies to address asymmetric infor-
mation problems (acknowledging, as we have just seen,
11
See Belleamme, L amber t and Schw ienbache r (2014) for a stylized
model exam ining this possibilit y.
12 See Cum ming, L eboeuf a nd Schwienbacher (2014) for an empir ical
analysis of this iss ue.
that some strategies address the two issues at once).
We can distinguish between two generic types of prob-
lems: hidden information problems (funders often lack
the necessary information to estimate the chances of
success of the proposed campaigns) and hidden action
problems (funders have a hard time to control how fund-
raisers use the collected funds).
A rst instrument in mitigating hidden information
problems is direct screening: CFPs conduct due dil-
igence themselves and reject projects that are deemed
too risky. Alternatively (or complementarily), CFPs may
provide funders with a market-based screening mecha-
nism; for instance, some crowdlending platforms give
funders access to ‘soft’ information about fundraisers
(such as the maximum interest rate they are willing to
pay, a textual description of their reasons for the loan
application, or their pictu re). Studies show that funders
can predict default with greater precision on the basis of
such nonstandard information than with the use of more
traditional screening methods based on credit score.13
Finally, CFPs may also bring sophisticated investors on
board like institutional investors, venture capitalists, or
business angels, who have much larger capacities and
experience in due diligence. Their presence is thus like-
ly to reassure funders, as more information will be made
available about the chances of success of the proposed
campaigns.
As for hidden action problems (such as moral hazard),
a rst immediate measure is to invest in an adequate
monitoring system so as to avoid – or at least lim-
it – severe opportunism problems from being creat-
ed by fund raisers, such as outright fraud. In the same
vein, CFPs can prevent fundraisers from using arriv-
ing funds before the success of a campaign is assured
by taking control of making the nancial transaction.
Another strategy is to install a reputation system.
Naturally, such a system can only work if fundraisers
repeatedly use a given CFP. In such cases, the CFP
can use the t rack record of a given fundraiser to pro-
vide funders with useful information about that fund-
raiser’s reliability. To enrich this reputation system, the
CFP can also tap into the wealth of information avail-
able on social networks: It has indeed been document-
ed that the number of friends that f undraisers have on
Facebook can be used as a predictor of the success of
their projects.14 Finally, CFPs may also nd ways to
13
See, e.g., Berkovich (2011) or Herzenstein, Sonenshein and Dholakia
(2011).
14
See, for example, Mollick (2014).
Forum
CESifo DIC E Repor t 2/2016 (June) 10
insure funders against a number of risks; for instance,
some crowdlending platforms choose to partner with
banks to insure against market risks.
Conclusion
Our goal in this ar ticle was to show that crowdfund-
ing platforms are at the heart of current developments
in the different forms of crowdfunding. Without the
intermediation services that these platforms provide,
fundraisers and funders would not be able to interact in
an efcient way. To make our point, we have described
the ‘why’ (i.e., the complex web of external effects that
crowdfunding generates for funders and f undraisers), as
well as the ‘how’ (the price and non-price strategies that
platforms deploy to address these external effects).
Even if crowdfunding is still nascent and is thus bound
to evolve, we believe that the framework of the analysis
proposed here will remain relevant and help us under-
stand future developments.
References
Belleam me, P., T. Lambe rt and A . Schw ienbacher (2014),
„Crowdfu nding: Tappi ng the Rig ht Crowd”, Journ al of Busin ess
Ven tu r in g 5, 585 –60 9.
Belleam me, P., N. Omra ni and M. Peitz (2015), “The Economics
of Crowdfu nding Platforms”, Information Economics and Policy 33,
11-2 8.
Belleam me, P. and M. Peitz (2015), Industrial Organization:
Market s and Strategie s, Second Edition , Cambr idge Univer sity Pre ss,
Cambridge.
Berkovich , E. (2011), “Search and Herd ing Effects in Peer-to -Peer
Lendi ng: Evidence from Prosper.com”, Annals of Finance 7 (3),
389– 40 5.
Chen, D. and Z. Lin (2014), “Rational or Ir ration al Her ding in O nline
Microloan Markets: Evidence from China”, mimeo.
Cumm ing, D. J., G. Leboeuf and A. Schwienbache r (2014),
“Crowdfu nding Models: Keep-it-All vs. All-or-Nothi ng”, mimeo.
Damia no, E. and H. Li (2008), “C ompeti ng Match making”, Jour nal of
the European Economic Association 6, 789 –818.
Herzen stein , M., S. Sonen shein and U. M. Dholakia (2011), “Tell Me
a Goo d Story and I May Lend You Money: The Role of Narr atives in
Peer-to-Pe er Lending Decisions”, Journ al of Marke ting Research 48,
138 – 49.
Kuppuswamy, V. and B. L.Bay us (2013), “Crowdfu nding Creative
Ideas: The Dynamics of Project Backers in Kickstarter”, UNC Kenan-
Flagler Research Pap er No. 2013 -15.
Lee, E. and B. Lee (2012), ”Herding Behav iour in Onl ine P2P Lend ing:
An Empirical Investigat ion”, Electron ic Commerce Resea rch and
Applications 11, 495–503.
Mollick, E. (2014), “T he Dyn amics of Crowd fund ing: Det ermi nants of
Success a nd Failu re”, Journal of Busin ess Ventur ing 29, 1–16.
McCorm ick, R. (2015), “Sony Launches Crowd fund ing Platfor m for its
Own New P roducts”, Th e Ver ge, 01 July, http://bit.ly/1It2VbT (accessed
17 May 2016).
Zhang, J. and P. Liu (2012), “Rational Herding i n Microloan Markets”,
Management Science 58, 89 2–912.
