Hi folks and welcome to my Imposter Syndrome newsletter. How are you all holding up?
Some of you may already know I’ve been running another Kickstarter campaign for my sci-fi comic series ACAUSAL which is due to end very soon, and one of the many things I’ve been thinking about is: what makes some campaigns more successful than others? Sure, there are lots of contextual factors to consider such as the creators involved, the genre, timing etc. But I wondered if there were other factors that I needed to consider, to help my project be successful. For example, Kickstarter recommends you make a video in order to help get funded – but my video viewing statistics suggest no-one watches my videos, so how successful are they?
Rather than turning to my usual sources of advice, I decided to go to the scientific literature to see what they had to say. Surprisingly, there were a few papers on crowdfunding – but not specifically for comics – and they had some very interesting findings, which I’ll discuss below.
Before that, a few caveats. The papers use historical data to model future behaviour, which is not always guaranteed- and although some of these models are validated, at best they show association and not causation. The findings may - or may not - seem common sense to some of you, but I still think it’s instructive to see the evidence.
IMAGES ARE IMPORTANT
This paper was interested in the impact of image use in crowdfunding campaign stories. They examined a Kickstarter dataset comprising 16,439 projects, including successful (4,187) and failed (12,252) projects where there was complete data about the project. Using a Bayesian additive tree model, they found that several variables were important correlates for campaign success -these were: brightness of the images (defined here as perceived amount of luminance); number of images; and, contrast of the images.
For creators, this suggests we need to pay attention to the visual elements of page design by providing as many images as possible and ensuring they are sufficiently bright and of high-contrast.
SOCIAL SIGNALLING IS IMPORTANT
This study set out to understand the impact of social signalling on the success of a crowdfunding campaign through an experimental study. Social signalling in this context was specific to information about the magnitude of contributions to the campaign, and how frequent those contributions were. They were interested in exploring questions such as does having large initial contribution amounts and many early contributors to a fundraising campaign signal project quality and funders’ confidence in a project?
They recruited participants through Mechanical Turk who were screened for familiarity with crowdfunding and performed two different experiments using a fictitious crowdfunding campaign to test whether the amounts and timings of individual contributions were associated with success. Their model was based more on GoFundMe like platforms, where such data is regularly provided to external users. The study found that projects were 30%more likely to attract backers when there was a wider range of contributions (independent of project category and goal), and when there was a wider variance in the times when these donations were made.
What does this imply for Kickstarter type projects where this data is not publicly available?
Have reward tiers that cover a wide range. For example, if you have only three tiers backers are more likely to fund the project if the tiers are priced at $5, $25, $70 than if the tiers were $20, $30, $50.
Projects were more likely to be funded if other people saw contributions at a wider range of time throughout the campaign (rather than lots of contributions made on day one). This may seem intuitive, but let’s drill into it a little more. A key feature of the Kickstarter community is that backers can share that they have backed your campaign; seeing that someone you know and trust has backed a campaign may make you more likely to also back it. So this ability to share is crucial to a campaign’s success.
TONE AND NUMBER OF REWARDS ARE IMPORTANT
This paper examined data from all projects on Kickstarter between 2009 and 2014, excluding ongoing projects, cancelled, purged, and suspended projects. In total, the analysis considered data scraped from just over 150,000 projects.
The authors developed a linear logistic model that attempted to estimate the correlation between various variables such as tone of the story section, its language level, and number of rewards (for example). There is evidence from business literature that tone of a text should be moderate (not too positive or negative), so they included a quadratic tone term in the model to test this assumption. Their model results are shown in the figure below:
Here’s the important information to take from the above table.
Tone: holding all other variables constant and increasing the positivity of the tone by 1% is associated with 15.2 times greater odds of being funded. In terms of probabilities, this means that a 1% more positive tone increased the probability of being funded by 94%. This is a huge number and suggests tone is incredibly important for the story. However, they also included a higher-order correction term in the model and that is also very important as it confirms their initial hypothesis – that text should be positive, but not too positive (lest it be seen as overconfident and therefore reduce credibility).
Rewards: holding all other variables constant and increasing the number of reward tiers by 1% increase the odds of being funded by 2.6 (probability = 72%). This suggests that having a higher number of reward tiers is also important.
Factors least likely to be associated with success were a higher goal (lower goals are more successful) and also the duration of the campaign (shorter campaigns are more successful). The former is self-evident in that smaller goals are easier to reach, but the latter finding is not-obvious, as one may expect a longer campaign to have more time to fund. The model doesn’t show this to be the case.
Overall, what does this tell us?
Having numerous rewards at different price points appears to be a key enabler for success, as does having many (bright) images on the story page. Shorter campaigns, and campaigns with more social sharing also appear to be key enablers. The biggest surprise - having too positive a tone when writing about your campaign. Sadly, there was no information about how positive one needs to be, so I guess it’s a fine line between positivity and over-postivity!
Wanna hear me ramble?
I did a great interview with the Two Geeks Talking podcast where I discussed Acausal, String Theory, Star Trek and a host of other issues. For those who are interested, you can check out the 30 minute-long interview below.
Okay, that’s it for this instalment. Hope to see you next time!
John