The Champion Effect in Peer-to-Peer Giving: Successful Campaigns Highlight Fundraisers More Than Causes

Abstract

Online peer-to-peer giving is an emerging charity context that has rarely been investigated. Using a unique combination of survey and behavioral data from 1,647 online peer-to-peer fundraisers (whom we call “champions”), we tested empirically the influence of different best practices on fundraising success in this novel giving context. Across two samples, we found the fundraiser’s identification with the cause led them to engage in more best practice actions, which in turn led to greater fundraising success. However, not all actions were equally influential. Actions that made the champion salient—namely those relating to solicitation and those that signaled the fundraiser was highly invested in their campaign—were the strongest predictors of fundraising success, together explaining 28 times the variance accounted for by actions signaling charity efficacy. Thus, fundraisers will have more success by championing themselves than by promoting the charity in question: a finding with important applied and theoretical implications.

Keywords

fundraising charitable giving peer-to-peer social networks best practice

Introduction

New media and social networking are transforming the way people learn about and contribute to charitable causes (e.g., Guo & Saxton, 2014). However, little research has evaluated how our increasingly networked world influences charitable giving. Two notable exceptions show that online giving is at least partially driven by social network effects that are not typically observed in traditional giving contexts (Saxton & Wang, 2014; Scharf & Smith, 2016). However, it is not yet clear how the influence of social networks becomes manifest. Using behavioral and self-report data from two large samples of online peer-to-peer fundraisers, we analyze the best practice actions of fundraisers to test empirically four mechanisms known to drive charitable giving more broadly and evaluate their relative influence in predicting peer-to-peer fundraising outcomes. These are identification, solicitation, signaled investment (evoking reputation), and signaled efficacy (see Aaker & Akutsu, 2009; Bekkers & Wiepking, 2011).

We propose that success in the peer-to-peer fundraising context is influenced more by the champion than by the charity—a phenomenon we call the Champion Effect. Thus, while peer-to-peer fundraisers may be motivated by their connection with the cause, their donors are most likely to give because of their connection with the fundraiser (see also Scharf & Smith, 2016). We argue that, if champions are key determinants of fundraising success, then fundraisers will succeed to the degree that they make it clear that success is important to them. Below we outline the existing evidence base in relation to peer-to-peer giving, discuss several mechanisms that may influence giving within peer networks, and then test empirically the relative roles of fundraisers’ identification, solicitation, signaled investment, and signaled efficacy in determining fundraising outcomes. We later discuss how this evidence supports the notion that champions strongly influence peer-to-peer giving. Finally, we highlight the implications for nonprofit practitioners.

Charitable Giving Within Social Networks

Peer-to-peer fundraising harnesses the social connections of charity supporters to promote causes within social networks. Individual fundraisers become advocates for their favorite causes by asking friends, family, and colleagues for donations on behalf of a charity, often as sponsorship of fundraisers’ participation in endurance or symbolic events. In an increasingly networked society, where consumers are more likely to trust peer endorsement than traditional marketing communications (Miller, 2009), leveraging peer-to-peer networks is likely to become an increasingly important component of nonprofit fundraising success.

Mechanisms that promote charitable giving more broadly are documented (see review by Bekkers & Wiepking, 2011), including a long-standing body of research on traditional philanthropy that acknowledges the instrumental role that fundraisers themselves play in campaign success (e.g., Breeze, 2017; Tempel, Seiler, & Burlingame, 2016). To date, however, the specific factors at play in the peer-to-peer context have received less research attention. Two key studies, summarized below, show that giving in networked environments is influenced by the relationships between fundraisers and their networks.

Saxton and Wang (2014) evaluated the online fundraising success of 66 of the largest nonprofits in the United States. By comparing data scraped from charities’ Facebook Cause pages to information obtained from those organizations’ tax submissions, the authors demonstrated that the size of charities’ social media fan base was positively associated with the amount of money they raised on their Facebook Cause pages, but the organizations’ efficiency ratings were not. They concluded that online fundraising success is determined more by social network effects than traditional mechanisms like efficiency.

A second study specifically tested network effects in the peer-to-peer context. Scharf and Smith (2016) analyzed data from 35,571 online peer-to-peer fundraisers in the United Kingdom. They found that, after controlling for relevant fundraiser demographics and charity event factors, the number of Facebook friends a fundraiser had was associated with the amount of money they raised. Specifically, fundraisers with larger online social networks received a greater number of donations but smaller average gifts. The authors argued that results reflected a “local” public good that must be provided by the particular social group. These patterns of response indicate that donors are motivated by what the authors call “relational altruism”: they give because they care about the fundraiser and they perceive that the fundraiser cares about how much money they raise.

The two studies outlined above provide evidence that social network size influences online fundraising outcomes and suggest that peer-to-peer donors may be motivated by relational altruism (Scharf & Smith, 2016). These studies, however, do not speak to the way fundraisers may harness that motivation by signaling how much they care. We propose that the actions which fundraisers decide to take in their effort to raise money may determine their outcomes, and that different fundraising actions signal different priorities or degrees of care. In turn, the signals that these actions give evoke different degrees of responsiveness in donors.

Outside of the academic domain, the charity sector itself has highlighted particular “best practice” actions that individual fundraisers should take. Best practices are those actions that fundraising professionals identify as being most effective in raising money. For example, industry reports suggest that fundraisers who send more emails to their networks and who tell a story about why they are fundraising are more successful (Braiterman & Masterson, 2015; DonorDrive, 2017). Blogs targeted at fundraising practitioners suggest diverse tactics, including setting a low initial target, personalizing the fundraising page, asking for specific donation values, sending targeted emails, and sharing fundraising pages via social media (e.g., Classy, 2017; Francis, 2017). Within the scholarly research literature, however, there is little empirical evidence that these tactics work more effectively than alternatives. We aim to build an evidence base not only for what works in online peer-to-peer fundraising, but for why it works. To do so, we must first understand the mechanisms that promote giving in general.

Mechanisms Driving Peer-to-Peer Giving

A range of mechanisms promote charitable giving more broadly (see Bekkers & Wiepking, 2011, for review). Four of these—solicitation, reputation, efficacy, and identification—may help us understand the determinants of successful peer-to-peer fundraising as well. Below we discuss each potential mechanism in turn as well as how it may relate to charitable giving within peer networks.

Solicitation, or the simple act of being asked to donate, is a major driver of charitable giving (Bekkers & Wiepking, 2011). Indeed, most donations are given in response to a request (Bryant, Jeon-Slaughter, Kang, & Tax, 2003). Solicitation may be a particularly strong mechanism in peer-to-peer giving because most donors to online fundraising pages are already in the social network of the fundraiser (Payne, Scharf, & Smith, 2014), and the closeness of the relationship between the donor and solicitor influences success. In traditional giving contexts, being solicited by an acquaintance rather than a stranger increases both the likelihood of donating and the value of the gift (Meer, 2011). Being asked by a family member or friend appears to be especially powerful in motivating response (Scharf & Smith, 2016). Solicitation is, therefore, expected to be a powerful determinant of peer-to-peer fundraising success because the person asking is likely to be known to and valued by the donor.

Reputation refers to the social consequences of making (or refraining from making) a donation (Bekkers & Wiepking, 2011). Social concerns about gaining status and avoiding shame are inherent in charitable decisions (Bénabou & Tirole, 2006). People are more generous when their charitable donations are visible to others, especially people that matter to them (Alpízar & Martinsson, 2013; Bereczkei, Birkas, & Kerekes, 2007; Satow, 1975; Soetevent, 2005). If reputation—specifically maintaining a positive relationship with the fundraiser—is a key consideration for donors choosing whether or not to respond to charitable solicitation, actions that make the fundraiser and their personal motivation salient should improve fundraising outcomes. The more the fundraiser cares about a particular charity, the more important it may be for the donor to support them through donating to “their” cause. We propose that fundraising targets, which have been shown to affect donor responses (Payne et al., 2014; S. Smith, Windmeijer, & Wright, 2015), are one way that fundraisers may signal their level of investment in the outcome. Targets are usually not linked to any specific funding need and the funds raised are passed to the charity regardless of whether or not the target is met. Therefore, the use of targets in online peer-to-peer contexts likely signals the fundraiser’s motivation and how much they care about the charity in question (see also Scharf & Smith, 2016). Other actions that may highlight the champion have been emphasized as good practices by charities, but to our knowledge they remain untested until now. These include personalization of the fundraising page, and sharing personal stories or connection with the cause (Classy, 2017; Francis, 2017). We propose that actions that highlight champions and signal their investment may be strong determinants of success in peer-to-peer fundraising contexts, even compared to actions highlighting charity efficacy.

Information that communicates efficacy, or the perception that donations will actually make a difference to the relevant cause, can encourage donors to make charitable contributions (Bekkers & Wiepking, 2011). The desire to personally make a difference is theorized to be a key consideration among philanthropists (Duncan, 2004), and there is evidence that people who think donations are more likely to help the needy report greater intentions to donate (J. R. Smith & McSweeney, 2007). Although field experiments show that information about effectiveness does not always affect donation likelihood, donors believe they are more likely to give to effective charities (Parsons, 2007; Scharf & Smith, 2016). However, efficacy has been demonstrated to be less influential in driving donations on social media, as noted earlier (Saxton & Wang, 2014), and may not be as important a consideration as simple personal preferences (Berman, Barasch, Levine, & Small, 2018). If champions are key determinants of fundraising outcomes in the peer-to-peer domain, the perceived efficacy of the charity in question may be less important than the perceived importance of the campaign to the fundraiser. Thus, a friend or family member may give to support a fundraiser’s charity effort without caring especially about the charity’s cause or perceiving the charity to be an effective agent.

Finally, identities have been demonstrated to influence charitable decisions, especially whom we choose to help (e.g., Chapman, Louis, & Masser, 2018; Wiepking, 2010). As mentioned above, we expect donors to be more motivated by their connection with the fundraiser than the cause. However, the degree to which the fundraiser is identified with their selected charity should also influence outcomes, insomuch as identification motivates them to exert effort to achieve fundraising success. People are more willing to help when they identify with the individual or group in need (e.g., Levine & Thompson, 2004; Zagefka, Noor, & Brown, 2013). Extrapolating out, we reason that fundraisers who identify more with their selected cause are likely to do more to help the charity.

The Current Research

Building on previous work that demonstrates social network effects in online giving (Saxton & Wang, 2014; Scharf & Smith, 2016), we propose that giving contexts may change the relative importance of fundraising techniques. The peer-to-peer domain makes the social network salient, and different techniques may, therefore, be effective in peer-to-peer versus traditional giving contexts. We aim to test empirically several mechanisms known to influence charitable giving in traditional contexts to assess their relative roles in motivating online peer-to-peer gifts. These are: the fundraisers’ identification with the charity in question (identification), efforts to ask for donations (solicitation), the degree to which the fundraiser signals investment in the outcome (evoking reputation), and the degree to which the fundraiser signals the efficacy of the charity (efficacy). To do so, we look to fundraising best practices highlighted by charities and assess their comparative influence on fundraising success.

While fundraisers themselves may be motivated by their identification with the cause they have nominated, we propose it will be the actions they choose to take in fundraising that will primarily determine their fundraising outcomes. If relational altruism motivates donors in peer-to-peer contexts (see Scharf & Smith, 2016), any action that signals the importance of fundraising success to the fundraiser will be particularly influential in reaching fundraising targets. Such actions include asking for donations, setting an ambitious target, and making their own identity and motivation a key component of their campaign.

Specifically, we hypothesize the following:

Hypothesis 1 (H1): Fundraisers’ identification with the cause, their solicitation efforts, signaled investment, and signaled efficacy will all positively predict amount raised.

Hypothesis 2 (H2): Fundraiser actions (solicitation, signaled investment, and signaled efficacy) will mediate the relationship between fundraiser identification and amount raised.

Hypothesis 3 (H3): Solicitation and signaled fundraiser investment, both of which make the champion salient, will be stronger predictors of amount raised than signaled charity efficacy.

The current research answers calls to create stronger links between researchers and practitioners (Bushouse & Sowa, 2012) by testing best practices identified by fundraising professionals. To our knowledge, this is the first research (a) to examine the relationships between self-reported fundraiser identification, best practices taken, and behavioral fundraising outcomes (i.e., dollars actually raised), and (b) to evaluate empirically the relative explanatory power of best practices on fundraising outcomes in the emerging online peer-to-peer domain. By evaluating the efficacy of various fundraising practices in this new giving context, the research can contribute to evidence-based practice while providing a more nuanced understanding of the ways that charity contexts may influence the psychological mechanisms underpinning the gift.

Method

Secondary data from Australian online peer-to-peer fundraisers were analyzed to investigate the influence of fundraisers’ identification with the cause, solicitation practices, signaled investment, and signaled charity efficacy in fundraising success. To increase confidence in the results obtained (see Asendorpf et al., 2013; Brandt et al., 2014; Schmidt, 2009), analyses were conducted on survey responses from peer-to-peer fundraisers at two separate timepoints.

Participants were surveyed in both 2013 and 2014 after taking part in a large Australian fun run. When taking part in this event, fundraising for a charity is optional and fundraisers nominate a beneficiary charity of their own selection. All runners who fundraised via everydayhero (N2013 = 5,609 and N2014 = 5,502), the official online fundraising platform for the event, were invited after the event to complete a survey about their fundraising experiences. More Strategic, a fundraising consultancy, designed the survey and collected the data via the Qualtrics survey platform on behalf of everydayhero. Data were subsequently anonymized and shared with the researchers free of charge for scholarly purposes. The researchers received no compensation for their work with this data set.

Participants

In 2013, 1,040 (19%) of fundraisers voluntarily responded to the survey, of whom 768 (74%) completed all measures of interest for this study and are included in the analysis. The majority of respondents were female (59%), with 31% male and 10% preferring not to disclose their gender. Their age ranged from under 18 to over 65 years, with 29% aged under 30, 27% aged 30 to 39, 20% aged 40 to 49, 14% aged over 50, and the remaining 10% preferring not to disclose their age. Collectively, they fundraised for 273 different charities.

In 2014, 1,180 (21%) voluntarily responded. Of those, a total of 878 (74%) completed all measures of interest for this study and are included in the analysis. Two thirds (66%) of respondents identified as female and 34% as male. Six participants (i.e., less than 1%) selected not to disclose their gender. Participants’ ages ranged from less than 18 to more than 65 years, with 32% aged under 30, 27% aged 30 to 39, 23% aged 40 to 49, and 18% aged over 50. Collectively, participants were fundraising for a total of 223 different charities.

Measures

Measures outlined below represent only a subset of those administered in the full survey, which is available on request. All measures were identical across the 2 years, except identification.

Fundraiser identification

In 2013, participants indicated the degree to which they identified with the particular charity they were fundraising for by agreeing or disagreeing with three statements (“Supporting them is an important part of who I am,” “It is a cause I have always been passionate about,” and “They are the most important cause I support,” all coded 0 = disagree, 1 = agree). A supporter identity scale was created by averaging scores, with higher scores indicating greater identification with their selected charity, α = .62. In 2014, respondents indicated the degree to which they agreed or disagreed with five statements (“I am really passionate about their work,” “They are the most important cause I support,” “Giving to them is an important part of who I am,” “I will try and live my life in a way that supports this cause,” and “I would proudly wear their T shirt when going out to be associated with them,” 1 = strongly disagree, 5 = strongly agree). A supporter identity scale was again created by averaging scores, α = .89).¹

Fundraiser actions

A list of 14 actions identified by More Strategic and everydayhero to be best practices in peer-to-peer fundraising were included in the survey. Fundraisers indicated which, if any, of the actions they had performed in the course of their fundraising efforts (each scored 0 = no, 1 = yes). One item (“Set myself a challenging target”) was excluded in favor of using the behavioral measure of target value that was preloaded into the survey, which is described below. A principal components analysis on the remaining 13 items from the 2013 sample revealed a 3-factor solution using either eigenvalues greater than 1 or the scree plot, which accounted for 44% of the variance. Three items cross-loaded greater than .30 on two factors (see Table 1). In the 2014 follow-up sample, two of the same three items cross-loaded on two factors and were therefore removed from analyses.² The social media post item (“Posted links and messages on my Facebook and other social media”), however did not cross-load but instead loaded at .57 on the signaled investment factor. This item was therefore retained for the 2014 sample. Identified cross-loading items were omitted and the analysis was rerun with oblimin rotation as factors were theorized to be correlated. Factor scores were computed by regression and saved for inclusion in the analyses below.³ Table 1 reports the final items and coefficients for the three factors, operationalizing the theorized constructs of solicitation, signaled investment, and signaled efficacy, which explained 50% of the variance in the 10 items retained in 2013 and 47% of the 11 items retained in 2014.

Table 1.

Item Wording and Factor Loadings for Fundraising Best Practice Questions (2013).

Item	Item wording	Oblimin rotated factor loadings
Item	Item wording	Solicitation	Investment	Efficacy
Emailed everyone	Emailed everyone I could (not just everyone I was willing to)	.53	−.15	.07
Email reminders	Followed up my initial email with a reminder	.64	.09	−.06
Asked in person	Reminded people to donate when I met them socially	.55	.06	.05
Updated page	Made changes to the standard Supporter Page provided by the charity	.02	.74	−.05
Shared reasons	Shared the reasons why I care about this cause on my Supporter Page	.10	.67	.10
Uploaded photo	Uploaded a personal photo to my Supporter Page	−.08	.76	.01
Donation impact	Told people what their donation could achieve for [Charity Name]	−.03	.00	.81
Fundraising impact	Told people what reaching my target could enable [Charity Name] to do	.00	−.05	.81
Shared charity info	Shared information from the charity with people I have asked	.08	.15	.60
Specific donation	Suggested a specific donation amount (e.g., US$67 will allow the Charity to do xyz)	−.01	.04	.39
Cross-loading items excluded from final solution
Social media post	Posted links and messages on my Facebook or other social media pages	.38	.55	−.22
Social media ask	Directly asked people for donations using social media such as Facebook	.55	.37	−.06
Thanked	Personally thanked every donor	.44	.32	−.02

Note. Factor loadings over .30 appear in bold.

Fundraising target

The target that fundraisers set was preloaded into the data in categorical form. For the purpose of analysis, value categories were coded in relation to the everydayhero default setting. Participants were coded 0 if they had selected or left the default target of US$700 in place, 1 if they had raised their target above the default, and −1 if they had lowered their target.

Funds raised

The actual funds raised by each fundraiser were preloaded into the survey from the everydayhero platform, allowing analysis of behavioral rather than self-reported data. Funds raised were reported in Australian dollars and cents.

Results

Means, standard deviations, and bivariate correlations for both data collection periods are summarized in Table 2. In both years, fundraisers actioned, on average, approximately half of the best practices (see Appendix for more detail). Furthermore, roughly one-third of respondents, respectively, set a higher target, lower target, or chose or left the default target setting. Respondents raised a total of A$753,764 for charity in 2013 (M2013 = A$981.46, SD2013 = A$2,380.97) and A$885,804 in 2014 (M2014 = A$1,008.54, SD2014 = A$1,603.00). The amounts raised, however, ranged from A$5 to A$50,230 with a strong positive skew (i.e., most people raised small amounts). Values were log transformed in order to meet the assumptions of normality for regression modeling. As shown in Table 2, all predictors were positively associated with the amount of funds raised (rs > .12, ps < .01). Most predictors were significantly associated with one another, although all collinearities between factor scores were low (rs < .32).

Table 2.

Descriptive Statistics and Zero-Order Correlations Between Fundraising Predictors and Amount Raised in 2013 (Below the Diagonal) and 2014 (Above the Diagonal).

		2013 M (SD)	2014 M (SD)	1	2	3	4	5	6
1	Fundraiser identification	0.67 (0.35)	3.92 (0.78)	—	.09**	.15***	.07*	.20***	.12**
2	Solicitation	0.02 (1.00)	0.01 (1.00)	.15***	—	.20***	.32***	.26***	32***
3	Signaled investment	0.03 (1.00)	0.01 (0.99)	.10**	.13***	—	.15***	.22***	.26***
4	High target	0.13 (0.84)	0.08 (0.79)	.14***	.22***	−.01	—	.19***	.49***
5	Signaled efficacy	0.01 (1.01)	0.01 (1.00)	.23***	.25***	.21***	.18***	—	.23***
6	Raised (log)	981.46 (2,380.97)	1,008.54 (1,603.00)	.14***	.30***	.17***	.49***	.24***	—

Note. N2013 = 768; N2014 = 878 (Listwise). Solicitation, signaled investment, and signaled efficacy use factor scores from Principal Components Analysis. Fundraiser identification was measured on a 0 to 1 scale in 2013 and a 1 to 5 scale in 2014. Target coded −1 = less than A$700; 0 = default value A$700, 1 = more than A$700. Mean and standard deviation reported for Raised are untransformed.

p < .05. **p < .001. ***p < .001.

The analyses below report results based on factor scores for solicitation, signaled investment, and signaled efficacy rather than the individual actions. The use of aggregated scores is preferred here for parsimony and to remove potential instability in the model caused by multicollinearity. Results with individual actions may be of particular interest to fundraising practitioners, however. Details of the prevalence and relative influence of individual actions are therefore reported in an Appendix as well.

Hierarchical multiple regressions (summarized in Table 3) were conducted to regress log-transformed donations raised on fundraiser identification, solicitation, signaled fundraiser investment, and signaled charity efficacy. Identification with the nominated charity was entered in Step 1 to assess fundraiser motivation. Fundraisers’ solicitation factor score was entered at Step 2, while their signaled investment score and fundraising target were entered at Step 3 to assess the influence of making it apparent that their fundraising was important to them. Finally, signaled efficacy was entered at Step 4 to assess its unique impact over and above the champion-relevant signals.⁴

Table 3.

Hierarchical Regressions With Fundraisers’ Identification and Actions as Predictors of Funds Raised in 2013 and 2014.

	2013 amount raised: Log (β)				2014 amount raised: Log (β)
	Step 1	Step 2	Step 3	Step 4	Step 1	Step 2	Step 3	Step 4
Fundraiser identification	.14***	.10**	.04	.02	.12***	.09**	.05	.04
Solicitation		.28***	.17***	.16***		.31***	.18***	.17***
Signaled investment			.15***	.13***			.17***	.16***
High target			.44***	.43***			.41***	.41***
Signaled efficacy				.09**				.06*
R ² ch.	.02***	.08***	.20***	.01**	.01***	.09***	.20***	<.01*
Model R ²		.10***	.30***	.30***		.11***	.31***	.31***

Note. N2013 = 768; N2014 = 878.

p < .05. **p < .01. ***p < .001.

2013 Fundraisers

In the 2013 survey, fundraiser identification explained 2% of the variance in funds raised, F ch.(1,766) = 16.31, p < .001. Fundraisers who identified more strongly with the nominated charity raised more money, ß = .14, p < .001. Solicitation actions significantly explained an additional 8% of variance, F ch.(1,765) = 65.11, p < .001. As expected, fundraisers who asked for donations through more channels raised significantly more money, ß = .28, p < .001. Signaled investment explained an additional 20% of the variance in amount raised, F ch.(2,763) = 107.48, p < .001. Fundraisers who signaled investment through personalization actions, ß = .15, p < .001, and setting a higher fundraising target, ß = .44, p < .001, raised significantly more money. Finally, actions that communicated the efficacy of the charity explained an additional 1% of variance, F ch.(1,762) = 7.47, p = .006, with fundraisers who shared more information about charity efficacy also raising significantly more, ß = .09, p = .006.

Taken together, the full model explained 30% of the observed variance in dollars raised, F(5,762) = 66.98, p < .001. In the final model, all predictors remained significant except for fundraiser identification, which became nonsignificant once fundraising actions were accounted for. Bootstrapping analyses conducted in R using the lavaan package (Rosseel, 2012) with 1,000 resamples confirmed that the actions fundraisers took (relating to solicitation, signaled investment, and signaled efficacy) fully mediated the relationship between their identification with the charity and their fundraising success, combined indirect effect (IE) = .39, SE = .07, 95% confidence interval (CI) = [.26, .52].⁵ Results of the mediation model are presented in Figure 1.

Figure 1.

Mediation model showing how fundraiser’s identification with the cause predicts fund’s raised indirectly via their best practice actions of solicitation, signaled investment, high target, and signaled efficacy (2013 sample).

2014 Fundraisers

In the 2014 survey, the same pattern of results was returned (see Table 3). Fundraiser identification, entered at Step 1, explained just 1% of the variance in amount raised, F ch.(1,876) = 12.65, p < .001. Fundraisers who identified more strongly with their nominated charity raised significantly more money, ß = .12, p < .001. The inclusion of solicitation factor scores at Step 2 significantly improved the model, explaining an additional 9% of the variance, F ch.(1,875) = 93.29, p < .001. Fundraisers asking through more channels raised significantly more money, ß = .31, p < .001. Entered at Step 3, variables related to fundraiser investment significantly explained an extra 20% of the variance, F ch.(2,873) = 126.53, p < .001. Fundraisers who signaled their investment by personalizing their campaign, ß = .17, p < .001, and setting a higher target, ß = .41, p < .001, raised significantly more money. Finally, signaled efficacy factor scores explained less than 1% of additional variance, F ch.(1, 872) = 4.39, p = .036. Fundraisers who shared information about the efficacy of the charity in question raised significantly more money, ß = .06, p = .036.

Taken together, the full model explained 31% of the observed variance in amount raised, F(5,872) = 79.40, p < .001. All variables remained significant in the final model except for fundraiser identification, which was no longer significant once fundraiser actions were included. Bootstrapping analyses conducted in R using the lavaan package (Rosseel, 2012) confirmed that the actions fundraisers took (relating to solicitation, signaled investment, and signaled efficacy) fully mediated the relationship between their identification with the charity and their fundraising success, combined IE = .13, SE = .03, 95% CI = [.07, .19].⁶

Discussion

Across two field samples with online peer-to-peer fundraisers, we find support for all hypotheses related to the notion that fundraising outcomes within peer networks are influenced strongly by the fundraising champions themselves. Specifically, across both samples, we find that fundraisers who identified more with their selected charity, and who took more actions to solicit donations, signal their personal investment, and signal the efficacy of the charity, raised more money (supporting H1). The fundraiser’s greater identification with the cause was associated with taking more actions, which in turn was associated with greater fundraising success (supporting H2). All actions are not equal, however. Asking for donations through more channels and, especially, actions that signaled the fundraisers’ personal investment in the outcome were stronger predictors of fundraising success than actions that signaled the efficacy of the charity in question (supporting H3), respectively explaining at least 8 and 20 times the variance in amount raised.

While these findings will be intuitive to practitioners and to many scholars, they have not been demonstrated empirically before. To our knowledge, these are the first studies to (a) combine self-reported actions with behavioral outcomes in the peer-to-peer domain, and (b) evaluate the relative effectiveness of fundraising best practices in this emerging charity context. Findings have both theoretical and practical importance as they highlight how effective fundraising actions change in the new, increasingly important online peer-to-peer context. Furthermore, using best practices identified by fundraising practitioners helps to overcome the research-practice divide (see Bushouse & Sowa, 2012) by ensuring that actions have relevance to nonprofits themselves.

The focal hypothesis that the champion would be an important determinant of fundraising success in the online peer-to-peer context was strongly supported. In both samples, actions that made the fundraiser and their investment salient—such as uploading a photo, personalizing their fundraising page, articulating their reasons for fundraising, and setting a high target—were those most strongly associated with fundraising outcomes. Although we do not test donor motivations, these findings are consistent with Scharf and Smith’s (2016) assertion that people give in peer-to-peer contexts because they care about the fundraiser and they know the fundraiser cares about raising money. In our view, the current results show how fundraisers can effectively harness the power of such relational altruism by signaling their commitment to potential donors. They also echo evidence from research on traditional giving contexts that individual fundraisers are important for campaign success (Breeze, 2017; Tempel et al., 2016). Results show a strong Champion Effect in the peer-to-peer domain, with the fundraiser themselves being a key component of fundraising success.

Fundraisers who were more identified with their selected charity raised more money, apparently because they put more effort into their fundraising and performed a greater number of best practice actions. Because our data were collected after fundraising was complete, we cannot be certain that identification led to success via actions. It is also possible that fundraisers who raised more money later felt more identified with the cause. However, our contention that identification with the cause led fundraisers to exert more effort to raise money aligns with previous work showing that people who are more identified with an individual or group are willing to do more to help them (e.g., Levine & Thompson, 2004; Zagefka et al., 2013). In addition, we identify concrete mechanisms through which identities can affect outcomes in the charitable domain: the best practice actions fundraisers take.

As hypothesized, solicitation—or asking for donations through more channels—was shown to be an important contributor to fundraising success (corroborating Bekkers & Wiepking, 2011; Breeze, 2017; Bryant et al., 2003). In the fundraising literature, it has been shown that people are more likely to give when they are asked to donate by someone known to them, and that donors tend to respond more favorably to fundraisers who are close to them (Meer, 2011; Scharf & Smith, 2016). In the peer-to-peer domain, fundraisers are almost certainly known to the donors (Payne et al., 2014). We argue throughout that signaling investment may make reputation salient to the donor. One important way that fundraisers signal their investment is to ask more often, and through more channels. In this way, solicitation may amplify the perceived relational consequences—whether positive or negative—of the donor’s response. Indeed, within the wider literature, concerns about reputation have been observed to affect charitable responses (Alpízar & Martinsson, 2013; Bekkers & Wiepking, 2011; Bénabou & Tirole, 2006; Bereczkei et al., 2007; Satow, 1975; Soetevent, 2005). In the peer-to-peer context, where relationships between fundraiser and donor are personal, solicitation factors may be especially powerful.

Beyond mere solicitation, fundraisers can also draw attention to themselves (and indirectly highlight potential reputational consequences) by signaling personal investment in their campaign. It has been argued before that peer-to-peer donors give, at least in part, because they care about the fundraiser, they understand the fundraiser cares about the cause, and they want the fundraiser to succeed (Scharf & Smith, 2016). We propose that signaling investment in the campaign evokes reputation because it makes clear that the fundraiser will be paying attention to donor responses. Results here show that actions that signal investment are indeed strong determinants of fundraising success.

Setting a high target was the strongest unique predictor of fundraising success. We interpret this finding as showing that people who are more identified with their cause also set a higher fundraising target, indicating that targets are determined at least in part by fundraiser motivation (see also Scharf & Smith, 2016). Fundraisers could, however, adjust their targets during the campaign, and the current data cannot differentiate between those who set a high initial target and those who set a lower initial target but raised it as their campaign progressed. It must also be acknowledged that targets may also be determined by such pragmatic concerns as perceived wealth and size of the network in question. That is to say, in addition to their personal investment, fundraisers surely consider how many people they know well enough to ask for a donation, and the relative resources those people hold, when determining an appropriate target. Future research would therefore benefit from controlling for network size and resources to assess the impact of the champion effect over and above these practical considerations.

When solicitation and signaled investment were considered in unison, they accounted for almost 30% of the observed variation in fundraising success. This effect size is substantial. To contextualize, it is comparable both to the combined predictive power of household income, debt, and demographic make-up on the size of household charitable donations (28%; Hughes & Luksetich, 2008), and to the combined impact of fundraising expenditure, price of giving, organizational age, and social network size on the value of donations received on Facebook Cause pages (30%; Saxton & Wang, 2014). Regardless of their motives, champions who both ask for donations through more channels (solicitation) and ensure their social networks know how much they care about the outcome (signaling investment) are those who raise the most money. Although we distinguish solicitation and signaled investment, it is hard to disentangle the effects of these two mechanisms in the peer-to-peer domain because the solicitor is known to the donor (Payne et al., 2014) and the very act of asking signals investment. As mentioned previously, both types of action make champions salient and, we argue, both evoke reputation. It would be interesting to explore their joint operation or interactions in future research.

The perceived efficacy of a charity generally influences fundraising success (Bekkers & Wiepking, 2011; J. R. Smith & McSweeney, 2007), and as hypothesized, signaling charity efficacy was associated with greater donations in the two samples here. Yet the role of signaled efficacy was less important in promoting fundraising success than other factors (echoing Berman et al., 2018; Saxton & Wang, 2014). After accounting for champion-related actions—solicitation and signaled investment—promoting the efficacy of the charity or donations explained no more than 1% of extra variance in the amount raised.⁷ These results support the idea that champions are more important drivers of fundraising outcomes in the peer-to-peer domain than charities are. Nonetheless, it is also possible that donors may simply be relying on sources other than the fundraiser for information about charity effectiveness. Future research could ask if, when, and how donors seek effectiveness information in the peer-to-peer domain.

Successful peer-to-peer campaigns highlight fundraisers more than causes: a phenomenon we call the Champion Effect. On a theoretical level, these findings highlight the apparent motivational duality of peer-to-peer fundraising: it may be that in peer-to-peer contexts the fundraiser and donor perceive different targets as the beneficiaries of the altruistic response. Donors appear to be focused on giving to the fundraiser (Scharf & Smith, 2016). However, fundraisers who take more best practice actions are those who are more strongly identified with the charity, suggesting fundraisers are focused on the charity as the beneficiary of their actions. The fundraiser’s pivotal role as a champion of the cause can be enhanced according to the actions that they take. How to most effectively equip the champion with best practices that signal their investment emerges as an exciting direction for researchers and practitioners to explore.

Strengths, Limitations, and Future Directions

Strengths of the current research include the use of behavioral outcome data, large community samples of actual fundraisers, practitioner-led survey development, and the close replication of methodology and results across two samples. However, several limitations warrant mention.

First, response rates were relatively low, with only 19% to 21% of invited fundraisers choosing to participate in the research. We, therefore, cannot rule out the possibility that results may only reflect the experiences of a subset of highly motivated fundraisers who are particularly passionate and dedicated. Second, because fundraisers were free to select the beneficiary charity for their fundraising, a vast array of different charities were included in the data (273 in 2013 and 223 in 2014). Research has shown that some charity types are more effective in fundraising via social media (Saxton & Wang, 2014) and that different types of people are motivated to give to different types of charities (Chapman et al., 2018; Wiepking, 2010). Future research should, therefore, consider how the type of charity may moderate the relative importance of the mechanisms evaluated here. Third, the data were collected exclusively in Australia. Given that some cultures exhibit different patterns of giving (Charities Aid Foundation, 2017), it will be important to also test the Champion Effect in different cultural contexts. Finally, although the model tested here explains substantial variation in the amount of funds raised, it does not explain all of it. As previously mentioned, practical concerns such as the size and wealth of fundraisers’ social networks surely matter. Furthermore, factors such as awareness of need, altruism, norms, values, and prestige have all been demonstrated to influence charitable giving in traditional contexts (see reviews by Bekkers & Wiepking, 2011; Sargeant & Woodliffe, 2007). Future research on peer-to-peer giving would benefit from studying the impact of such factors in networked contexts.

Evidence presented here suggest that donors are influenced by the perceived investment of the fundraiser more than by the efficacy of the charity. An interesting avenue for future research will be to understand the boundary conditions of this effect. In particular, research should investigate whether there are certain issues or causes that are so polarizing that donors would be unwilling to support them, regardless of the enthusiasm of the fundraiser.

Applied Implications

Overall, evidence presented in this article supports the assertion that peer-to-peer giving is influenced by a champion effect, where campaign success is determined more by actions highlighting fundraisers than actions highlighting causes. Two factors are of particular importance: asking through more channels (solicitation) and fundraisers’ signaling the importance of the outcome to them personally (signaled investment). Fundraisers themselves are likely to be motivated by the cause in question (Payne et al., 2014) and may, therefore, select tactics aligned to their own motives (e.g., promoting the effectiveness of the charity in achieving its mission) while potentially neglecting tactics that could motivate others (e.g., promoting themselves and their connection with the cause). Our data speak to this phenomenon. Therefore, charities should intervene and educate individual fundraisers to help them to prioritize their efforts. Fundraisers should be encouraged to ask for donations through as many channels as possible and to ensure that their campaigns are personalized with photos, high targets, and articulation of their investment and motives.

Conclusion

In two large community samples of online peer-to-peer fundraisers, using behavioral outcome data, we found that fundraiser identification, asking for donations through more channels, signaling personal investment in fundraising success, and highlighting the efficacy of the charity were all significantly associated with raising more money. However, actions that highlighted the fundraiser themselves (solicitation and signaled investment) accounted for substantially more variance in fundraising outcomes than the individual or charity factors. Results highlight that fundraising best practices will depend on the giving context. We demonstrate an important “Champion Effect” in online peer-to-peer fundraising and suggest that efforts to equip the fundraiser with tools to convey their own personal connection with the cause will lead to greater success than efforts to highlight the effectiveness of the charity or its overarching mission. Fundraising practitioners should, therefore, evaluate the communications, toolkits, and other support they provide to individual fundraisers with the relative importance of these motivations in mind.

Footnotes

Appendix

In the article above, factor scores were used to evaluate the fundraising actions in the principal analyses, allowing greater methodological robustness, parsimony, and theoretical focus. However, the influence of specific best practices is surely of relevance to fundraising practitioners and is, therefore, reported here. The relative frequency of individual best practices (i.e., the percentage of fundraisers who reported they had done each practice) for both the 2013 and 2014 samples are reported in Appendix Table 1 as well as the zero-order correlations between each action. Hierarchical regression analyses for both 2013 and 2014 retaining original best practice items are presented in Appendix Table 2.

Acknowledgements

The authors wish to thank everydayhero and More Strategic for generously sharing their anonymized data for scholarly purposes. We also thank the editor and four anonymous reviewers for their helpful feedback on an earlier version of this manuscript. Both studies received ethical clearance from the University of Queensland (15-PSYCH-PHD-73-JH).

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The first author is supported by an Australian Government Research Training Program Scholarship but the authors received no financial support for the research, authorship, and/or publication of this article.

Notes

ORCID iD

Cassandra M. Chapman

Author Biographies

Cassandra M. Chapman is a PhD candidate in the School of Psychology at the University of Queensland researching the social psychology of charitable giving.

Barbara M. Masser is the Australian Red Cross Blood Services Chair in Donor Research at the University of Queensland researching blood product donor recruitment and retention and gender within the criminal justice system.

Winnifred R. Louis is a professor at the University of Queensland whose research interests are in the areas of identity, norms, and decision making.

References

Aaker

J. L.

Akutsu

(2009). Why do people give? The role of identity in giving. Journal of Consumer Psychology, 19, 267-270. doi:10.1016/j.jcps.2009.05.010

Alpízar

Martinsson

(2013). Does it matter if you are observed by others? Evidence from donations in the field. The Scandinavian Journal of Economics, 115, 74-83. doi:10.1111/j.1467-9442.2012.01744.x

Asendorpf

J. B.

Conner

de Fruyt

De Houwer

Denissen

J. J. A.

Fiedler

. . . Wicherts

J. M.

(2013). Recommendations for increasing replicability in psychology. European Journal of Personality, 27, 108-119.

Bekkers

Wiepking

(2011). A literature review of empirical studies of philanthropy: Eight mechanisms that drive charitable giving. Nonprofit and Voluntary Sector Quarterly, 40, 924-973. doi:10.1177/0899764010380927

Bénabou

Tirole

(2006). Incentives and prosocial behavior. American Economic Review, 96, 1652-1678. doi:10.1257/aer.96.5.1652

Bereczkei

Birkas

Kerekes

(2007). Public charity offer as a proximate factor of evolved reputation-building strategy: An experimental analysis of a real-life situation. Evolution and Human Behavior, 28, 277-284.

Berman

J. Z.

Barasch

Levine

E. E.

Small

D. A.

(2018). Impediments to effective altruism: The role of subjective preferences in charitable giving. Psychological Science, 29, 834-844. doi:10.1177/0956797617747648

Braiterman

Masterson

(2015). The 2014 Blackbaud peer-to-peer fundraising study. Retrieved from http://charity.everydayhero.com.au/wp-content/uploads/2015/07/2014-peer-to-peer-fundraising-study-1.pdf

Brandt

M. J.

Ijzerman

Dijksterhuis

Farach

F. J.

Geller

Giner-Sorolla

. . . Veer

(2014). The replication recipe: What makes for a convincing replication? Journal of Experimental Social Psychology, 50, 217-224. doi:10.1016/j.jesp.2013.10.005

10.

Breeze

(2017). The new fundraisers: Who organises charitable giving in contemporary society? Bristol, UK: Policy Press.

11.

Bryant

Jeon-Slaughter

Kang

Tax

(2003). Participation in philanthropic activities: Donating money and time. Journal of Consumer Policy, 26, 43-47.

12.

Bushouse

B. K.

Sowa

J. E.

(2012). Producing knowledge for practice: Assessing NVSQ 2000-2010. Nonprofit and Voluntary Sector Quarterly, 41, 477-513. doi:10.1177/0899764011422116

13.

Chapman

C. M.

Louis

W. R.

Masser

B. M.

(2018). Identifying (our) donors: Toward a social psychological understanding of charity selection in Australia. Psychology & Marketing. doi:https://doi.org/10.1002/mar.21150

14.

Charities Aid Foundation. (2017). World giving index 2017: A global view of trends. Retrieved from https://www.cafonline.org/docs/default-source/about-us-publications/cafworldgivingindex2017_2167a_web_210917.pdf?sfvrsn=ed1dac40_10

15.

Classy. (2017). Tips for successful fundraising. Retrieved from https://www.classy.org/fundraising-tips

16.

DonorDrive. (2017). The state of peer-to-peer fundraising 2017. Retrieved from http://www.donordrive.com/library/

17.

Duncan

(2004). A theory of impact philanthropy. Journal of Public Economics, 88, 2159-2180.

18.

Francis

(2017). Top 10 tips to help your peer-to-peer fundraising participants raise more. Retrieved from https://probonoaustralia.com.au/news/2017/01/top-10-tips-help-peer-peer-fundraising-participants-raise/

19.

Guo

Saxton

G. D.

(2014). Tweeting social change: How social media are changing nonprofit advocacy. Nonprofit and Voluntary Sector Quarterly, 43, 57-79. doi:10.1177/0899764012471585

20.

Hughes

Luksetich

(2008). Income volatility and wealth: The effect on charitable giving. Nonprofit and Voluntary Sector Quarterly, 37, 264-280. doi:10.1177/0899764007310416

21.

Levine

Thompson

(2004). Identity, place, and bystander intervention: Social categories and helping after natural disasters. The Journal of Social Psychology, 144, 229-245. doi:10.3200/SOCP.144.3.229-245

22.

Meer

(2011). Brother, can you spare a dime? Peer pressure in charitable solicitation. Journal of Public Economics, 95, 926-941. doi:10.1016/j.jpubeco.2010.11.026

23.

Miller

(2009). Community fundraising 2.0—the future of fundraising in a networked society? International Journal of Nonprofit and Voluntary Sector Marketing, 14, 365-370. doi:10.1002/nvsm.373

24.

Parsons

L. M.

(2007). The impact of financial information and voluntary disclosures on contributions to not-for-profit organizations. Behavioral Research in Accounting, 19, 179-196.

25.

Payne

Scharf

Smith

(2014). Online fundraising—the perfect ask? (IDEAS Working paper series). Retrieved from https://warwick.ac.uk/fac/soc/economics/staff/kascharf/online-fundraising-payne-scharf-smith.pdf

26.

Rosseel

(2012). Lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48(2), 1-36.

27.

Sargeant

Woodliffe

(2007). Gift giving: An interdisciplinary review. International Journal of Nonprofit and Voluntary Sector Marketing, 12, 275-307. doi:10.1002/nvsm.308

28.

Satow

K. L.

(1975). Social approval and helping. Journal of Experimental Social Psychology, 11, 501-509.

29.

Saxton

G. D.

Wang

(2014). The social network effect: The determinants of giving through social media. Nonprofit and Voluntary Sector Quarterly, 43, 850-868. doi:10.1177/0899764013485159

30.

Scharf

Smith

(2016). Relational altruism and giving in social groups. Journal of Public Economics, 141, 1-10. doi:10.1016/j.jpubeco.2016.06.001

31.

Schmidt

(2009). Shall we really do it again? The powerful concept of replication is neglected in the social sciences. Review of General Psychology, 13, 90-100. doi:10.1037/a0015108

32.

Smith

J. R.

McSweeney

(2007). Charitable giving: The effectiveness of a revised theory of planned behaviour model in predicting donating intentions and behaviour. Journal of Community & Applied Social Psychology, 17, 363-386. doi:10.1002/casp.906

33.

Smith

Windmeijer

Wright

(2015). Peer effects in charitable giving: Evidence from the (running) field. The Economic Journal, 125, 1053-1071. doi:10.1111/ecoj.12114

34.

Soetevent

A. R.

(2005). Anonymity in giving in a natural context—A field experiment in 30 churches. Journal of Public Economics, 89, 2301-2323.

35.

Tempel

E. R.

Seiler

T. L.

Burlingame

D. F.

(2016). Achieving Excellence in Fundraising (4th ed.). Hoboken, NJ: John Wiley.

36.

Wiepking

(2010). Democrats support international relief and the upper class donates to art? How opportunity, incentives and confidence affect donations to different types of charitable organizations. Social Science Research, 39, 1073-1087. doi:10.1016/j.ssresearch.2010.06.005

37.

Zagefka

Noor

Brown

(2013). Familiarity breeds compassion: Knowledge of disaster areas and willingness to donate money to disaster victims. Applied Psychology, 62, 640-654. doi:10.1111/j.1464-0597.2012.00501.x