Herding behaviour in the over-funding phase
Why goals and stretch goals are important for maximizing revenue
Setting goals for crowdfunding campaigns can be challenging.
On the one hand, you need to make sure you raise sufficient funds to deliver on what you’re promoting, but on the other hand - you don’t want to risk not getting funded. In the latter instance, you may want to set a lower (external) goal which you’re confident you can hit, but have a higher (internal) goal that you secretly need to reach to make the campaign successful in your own eyes.
Which option do you choose? Hopefully this article can help you make the right decision for you!
These two options represent a dichotomy facing creators on all-or-nothing campaign platforms such as Kickstarter, and could easily be mitigated by switching to other platforms/models where you keep whatever funds you raise.
But the other platforms have smaller audiences, and less discoverability than Kickstarter, so when it comes to crowdfunding platforms for newer creators, your best bet (for now) is likely to be on Kickstarter, so that will remain the focus for the rest of this piece.
If you peruse the Kickstarter platform, the new updates mean that projects now show you the share of funding they have raised (as a percentage of the goal). Projects that have yet to be funded have raised less than 100% of their funding, while projects that have met their goals, but are still live, have over 100% funding, i.e., they are over-funded.
Many of the over-funded campaigns have low campaign targets, so the over-funding is perhaps more indicative of the creator’s risk tolerance than the huge success of the campaign. This is likely true for those projects that are around 300% over-funded but only have a handful of backers, however, a few over-funded campaigns are genuine blockbusters, and see extraordinary herding behaviour which has driven their over-funding to extreme levels, with backers in the hundreds or thousands.
An interesting question to ask, is whether backers behave differently when confronting a campaign that has yet to be funded, versus one that is over-funded. The former carries a certain risk that the project may not ultimately be funded, however, altruistic backers may feel compelled to help it over the funding threshold. Conversely, although the over-funded project is guaranteed, there are still risks associated with quality and delivery, but there are also fewer reasons for altruistic backers to provide their financial support.
Does this mean that herding behaviour may be stronger for pre-goal campaigns versus over-funded campaigns?
Why does this matter?
Consider the following scenario. A creator needs $1000 to break even on their campaign, but fearing failure, sets a lower target of $500.
Let us assume that there are Observational Learning (OL) signals are easily available, so that backers can decipher the product quality. If the herding behaviour tails off once a project is over-funded, then the creator may easily hit their $500 target, but they may find themselves finishing with only $800. Conversely, if they chose their target to be $1000, strong herding behaviour may drive them to just over their target, and they finish the campaign with $1100.
In both cases, they were successfully funded, but only in the latter case did they actually make a profit.
Which model is right?
A recent study collected data on all Kickstarter projects between April and June 2018, including page data such as videos, number of images, word count etc., and also tracked dynamic variables such as daily funding received, daily backer count etc. This resulted in 166K observations of roughly 5.5K projects.
The authors used a multi-level logistic regression model to examine herding behaviour, by setting their dependent variable to be the daily funding amount of each project.
The independent variables were then the lag of daily funding amount, funding magnitude, time elapsed, and two interaction terms. The campaign information were treated as control variables, and a dummy variable was included to identify pre-post funding phases.
Their model found a positive herding effect present in the over-funded phase, however it was a smaller (statistically significant) magnitude than in the pre-funded phase.
Another important sub-finding was that herding was stronger in the second-half of the campaign (i.e., between the mid-point and the end) than in the first-half (i.e., between the start and the mid-point).
What does this mean for creators?
The lesson for creators is to be careful about setting funding targets. While there is a positive herding effect for projects that are over-funded, it is smaller in magnitude than for projects that have yet to hit their goal. Referring back to the earlier example, the creator who set their goal at $1000 actually benefits more than the one who set their target at $500 (although the former may have sweated more about their campaign success).
If true, this is an important finding, particularly for creators who deliberately set low goals and then hit their funding target in the first half of the campaign. These creators may be losing out on funding they would otherwise have received.
This suggests creators need to be strategic about their funding target. Ideally, they should set a target they believe can be reached in the second half of the campaign (but before the end). This way they benefit from strong herding behaviour in the latter stages to push them over the line, and then still benefit from residual herding in the over-funded stage.
An alternative strategy is to create new goals (i.e., stretch goals) in the over-funded phase. While these were not discussed in the paper, these artificial triggers may play a similar role for herding behaviour - where altruistic backers continue to back an over-funded project in order for it to hit the stretch goal. My own hypothesis is that the herding behaviour is still present in the stretch goal phase, but at a much reduced level.
A caveat to the above findings is that there was no information in the paper about what happens in the super-funded campaign case. Given the large number of backers for some of these projects, it seems likely there is a critical threshold for over-funding, beyond which the herding behaviour becomes strong again - thus driving these blockbuster campaigns to super-funding levels.
If true, this suggests a third strategy for creators, which is to over-fund as early as possible and then keep promoting the campaign as hard as possible until it reaches the critical threshold. At this point, herding should become strong again, and the project could shoot into the funding stratosphere!
For more information about herding behaviour and Observational Learning (OL) you can check out this previous piece.