Kickstarter case study on the impact of social media
Exploring the effectiveness of social media platforms for promoting a four-day Kickstarter campaign.
Hi folks,
I recently ran a four-day Kickstarter campaign for my book “The Science of Crowdfunding,” which is a book focused on evidence-informed tips and strategies for optimizing crowdfunding campaigns.
Originally, I wasn’t going to run a campaign for this book at all. In fact some of you may recall a previous post where I discussed this issue in great depth. However, I realized that I was giving into fear. I was afraid of launching a campaign that wouldn’t be hugely successful. After much contemplation, I realized what I needed to do was launch the campaign and use it as a learning opportunity.
So, I did.
I launched the campaign on October 15 at 8:16am and set it to finish at 10am on October 19, meaning it was a fraction over four-days in length. I set a very low goal to ensure it would be funded, and launched with no promotion at all aside from a note on Substack, posted the night before launch.
As many of you know, this is not best practice for launching a successful campaign.
The purpose of this experiment wasn’t to see how much I could raise with no promotion, although that may be interesting. Instead, I wanted to learn more about my social media promotion strategy for a live campaign: how effective was it?
Case study: four-day Kickstarter promotion
Prior to launch, I set up unique tags for the following platforms and used them meticulously during the campaign: X/Twitter; Threads; Bluesky; Facebook, Substack, and my mailing list.
Now, X, Threads, and Bluesky are similar platforms because there is essentially a single feed that everyone posts to. The feed can be stratified and viewed differently by users, but essentially all posts go to the same space. I’ll refer to these as single-feed platforms for generality, although I acknowledge that both X and Threads use algorithmic feeds, which means your post could be seen by and engaged with people you don’t follow.
Facebook is more complicated because you can post messages to your followers but can also post to specific community/interest groups.
Substack notes appears to function like a single-feed platform, but there is also messaging functionality you can use to engage directly with those in your subscriber community, who have signed up to hear from you.
Mailing lists are already full of folks who are interested in hearing from you about your work, so this is (hopefully) a pre-engaged community.
It’s important to note that Kickstarter also has an organic promotion algorithm, that sends out emails to past backers, highlights/recommends projects on the platform, and also through newsletters. So even though I deliberately went against my own advice, and didn’t promote the pre-launch, Kickstarter automatically provides some support once the campaign goes live.
Single-feed platforms
I’ll start with the single-feed platforms (X, Threads, and Bluesky). I collected data on October 19, at 11am after the end of the campaign to use for analysis.
Comparing launch day posts
On launch day, I posted the same two messages on all three platforms. One message included only the appropriate link to the Kickstarter page, while the second message included an image featuring a mockup of the book, with the link showing up only as clickable URL text. I wanted to know how these two posts compared across all three platforms. Here’s the findings:
X (2112 followers)
Link only: 2 shares and 3 likes
Image: 4 shares and 5 likes
Threads (1129 followers)
Link only: 1 share and 1 like
Image: 1 share and 0 likes
Bluesky (402 followers)
Link only: 1 share and 1 like
Image: 2 shares and 1 like
Although X generated the most engagement (as measured by post shares and likes), the potential audience (followers) was much larger than the other platforms. To account for audience size, I normalized the data to focus on shares and likes per follower. The data then looks like the following:
X: link only (0.09% shares and 0.14% likes); image (0.19% shares and 0.24% likes).
Threads: link only (0.09% shares and 0.18% likes); image (0.09% shares and 0% likes).
Bluesky: link only (0.19% shares and 0.25% likes); image (0.5% shares and 0.25% likes).
The normalized data suggests engagement was much higher on Bluesky than the other two platforms, and that the post with image plus URL received more engagement than the URL alone. It also suggests X and Threads are similar in their engagement levels, which makes sense as they’re both primarily algorithmic feeds.
The above data allows us to compare two identical posts across three platforms, but let’s now explore campaign promotion on these platforms as a whole.
Posts across the duration of the campaign
Over the duration of the campaign I posted 9 times to X, 6 times to Threads, and 4 times to Bluesky. Posts were different for each platform (apart from the first two) in order to maximize engagement, and were a mixture of link or image posts. Overall, I found the following averages per post:
X (mean shares = 2.3; mean likes = 3.4)
Threads (mean shares = 1.2; mean likes = 1.0)
Bluesky (mean shares = 1.5; mean likes = 1.0)
Again, it looks like X has better engagement but when we normalize the data to take account of audience size we find the mean share/follower and mean likes/follower distribution becomes:
X (mean shares/follower = 0.11%; mean likes/follower = 0.16%)
Threads (mean shares/follower = 0.10%; mean likes/follower = 0.09%)
Bluesky (mean shares/follower = 0.37%; mean likes/follower = 0.25%)
which again suggests Bluesky engagement was better than both X and Threads.
Relationship between post type and views on X
One metric only available in X is the number of views per post. The mean number of views on X across the entire campaign was 163 (or 7.7% of followers), however we can also compare views for link posts to image posts across the campaign.
For links we find a mean of 2.3 shares per post, and 3.4 likes per post
For images we find a mean of 2.5 shares per post, and 3.5 likes per post
This suggests that likes were more common than shares as an engagement modality, and that image posts had more engagement than link-only posts, however a Chi-Squared test shows no statistically significant difference between post and engagement types.
However, is there an association between post type and views? A Mann-Whitney U-test shows no statistically significant difference (Z=-0.4307, p>.05), meaning the post type has no effect on views.
Impact of platforms on backer count
Most importantly, according to the Kickstarter dashboard, the four-day promotion across three platforms didn’t generate a single backer.1
Other channels
The other platforms provide limited data on individual posts, making it hard to compare them. I posted twice to my personal Facebook page, and four times to specific Kickstarter groups on the plaform, with extremely limited engagement in terms of post likes. However, 6% of campaign backers used the FB specific tag in my posts.
Similarly, 6% of backers came through my Substack link, where I have less than 400 followers and only 70 subscribers.
Finally, 19% of backers came through the solitary mailing list tag (where I have less than 300 followers).
The remaining 69% all came from the internal Kickstarter algorithm and recommendation process (according to Kickstarter).
Limitations
There are so many, I don’t know where to start. I focused on backers here rather than funding because social media is about people (in theory), but funding is ultimately the key measure for creators. I assumed that engagement is a realiable proxy for backer number, which is still a debatable assumption. The sample size for this case study is small, and may not be representative of other people’s experiences or outcomes, moreover it may not be indicative of future behaviour either. Context is also critical, and the nature and persuasion of each post also plays a role in whether people engage with it online, but that’s a more complicated analysis that I’ll defer to another time. I could go on. Suffice to say, this is a case study and should be treated as such - illustrative but not predictive.
Lessons for creators
In many ways, nothing about this case study should be surprising. The findings confirm what we should already know: that a warm (receptive) audience is more likely to back your campaign than a cold one.
The Kickstarter ecosystem is a closed-one, as the algorithm only promotes or engaged with people who have already backed at least one prior Kickstarter campaign; by design, the platform can’t email people who have never used the platform. But it’s important to remember that this audience is warm, they are primed to back Kickstarter campaigns, even if they may not previously have backed one of yours.
Those on your creator mailing list, or Substack, are also a warm audience because they’ve opted to hear from you. They may have come from past crowdfunding campaigns, or have purchased your work previously, or generally find you interesting for other reasons. This audience is one step removed from the crowdfunding platform, but they’re receptive to hearing from you as the creator, so it’s no surprise that some of them may back your campaign.
Social media, however, is a cold audience. You may have followers, but the nature of the platform doesn’t mean they’re necessarily interested in you or your work. They’re more like casual acquaintances; if they really wanted to hear from you, they would already be on your mailing list.
The case study has shown that warmer audiences are more likely to back your campaign. This is why you should share your pre-launch campaign widely before you launch, as this helps to warm the colder audience members.
Makes sense, right?
But if it does, then why do creators spend so much time promoting their campaigns on social media? The audience there is clearly cold, and even if there is engagement, there’s no guarantee it will translate into additional backers.
I think we do it because it’s convenient.
It’s easy to fire off a tweet or thread or skeet or whatever, asking folks to check out your campaign. It makes you feel like you’re promoting your work. You’re doing something. You can point to the post and monitor the engagement and reach as proof.
But while this can be a lot of work, it’s not valuable work.
If you haven’t converted your cold audience to a warm one by the time you launch, you may be wasting your time and energy on social media promotion.
This case study suggests meaningful engagement with a warm audience is what generates results; these are the people who already know about you, your work, or the platform. This means you need to grow your warm audience before you launch and then engage with them repeatedly during the campaign.
I’m not advocating that people stop using social media, but I think we need to recognize that it has limited utility as a channel for crowdfunding promotion. I know many creators on X who want to leave, but are afraid of not having a place to promote their work.
My counter-argument is that X is an ineffective place to do that anyway, and likely to become even more so as it enshittifies further. This also appears true of Threads, although given the newness of the platform and the small sample size in this case study, it’s hard to be definitive.
But if you’re only on X or Threads because you think it’s going to help you with promotion. Think again.
Only Bluesky, of the three channels here, appears to actually generate engagement, even with the small sample size. This is a newer platform, so perhaps this may change over time. Yet even in this case, the platform generated no backers for my campaign.
“The medium is the message,” said Marshall McLuhan, and he’s not wrong.
Although my (small) experiment showed posts with an image and URL outperformed those with just the URL, this is definitely an example of a platform effect.
What I mean, is that both X and Threads are algorithmically driven feeds, that can suppress or boost your post based on internal decision-making processes. They don’t show your posts to people who have asked to see them, because it’s more profitable for them. As a result, they can throttle URL posts or image posts or even text posts, depending on what you include in them. Whatever message your trying to share is therefore defined by the whims of platform leadership, and not by quality of marketing copy. For Bluesky, this may be different, but it’s still a relatively new platform with a small user base.
This campaign taught me a lot. I learned I need to work smarter, not harder, and I learned social media promotion is busy work, not meaningful work.
What did you think? Did this resonate with you? What are your thoughts about marketing and promotion for crowdfunding? Let me know in the comments.
It’s actually hard to know if this is true, because there is some uncertainty around how Kickstarter categorizes and tracks links.
This is incredibly fascinating. I have also been thinking about some of these same ideas with promotion and engagement. This also isn't the first post I have read about waning social engagement. I for sure am a fan of working smarter, not harder!
Very interesting stuff! I think creative people have been led to believe that social media is a necessary evil for so long that it’s hard to think of alternatives. I wish I could think of a solution.