1747 words • 7~11 min read

The many, many ways I screwed up my first science crowdfunding campaign.

Four generations of field hardened OpenCTDs.

It’s been over five years since Kersey Sturdivant and I launched Oceanography for Everyone – The OpenCTD, my first attempt at crowdfunding science. Over the years, that initial effort has grown into Oceanography for Everyone, a community of researchers, educators, and citizen scientists, and has created new open-source tools for open-source, open-science hardware. The OpenCTD is the finest oceanographic instrument that you can build in your own home for less than $300.

The crowdfunding campaign was a total disaster.

Since then, I’ve written several articles on how scientists can launch and managed crowdfunding campaigns:

…but I’ve never written explicitly about what we did wrong during that campaign and how it impacted our success. Now that the final reward from that campaign has been delivered (yes, five years later, talk about the eternally delayed crowdfunding campaign), it’s the right moment to look back and think about how everything went so wrong.

went with lesser-known platforms. We launched the OpenCTD on RocketHub. At the time, RocketHub was hosting the #SciFund Challenge, a campaign to encourage scientists to launch science crowdfunding campaigns. Both the #SciFundChallenge and RocketHub were relatively small players in the nascent crowdfunding world. RocketHub doesn’t even appear to do crowdfunding anymore, they’ve pivoted to a “social network for entrepreneurs”. The old OpenCTD campaign page is long deprecated. #SciFund Challenge’s website hasn’t been updated in almost half a year.

Here’s the thing with crowdfunding, and especially crowdfunding in the early days: There are two dominant communities that you can rely on. There’s the community of people who want to support what you’re doing and there’s the community of people enamored with the idea of crowdfunding. Being a crowdfunding “investor” is a hobby in and of itself and many of the biggest donors are people who support dozens of different campaigns. So the larger and more popular the platform, the more crowdfunding enthusiasts you’ll attract. Heck, since backing the very first OpenROV, I’ve backed 23 other projects on Kickstarter, most recently Public Lab’s Balloon Mapping kits.

By going with RocketHub, I committed our campaign to a smaller potential audience. Considering Kickstarter was garnering huge press at the time, this was a near-fatal mistake.

I didn’t do an “all or nothing campaign”. In crowdfunding, there’s two basic funding models: “keeping what you raise”, and “all or nothing”. Keep what you raise means that even if you don’t hit your goal, you keep the existing contributions at the end of the campaign (usually with the crowdfunding platform taking a deeper cut). All or nothing means if you don’t hit your goal, you get nothing. Keep what you raise is great for disaster funding–raising donations for medical expenses, hurricane recovery, and other unexpected catastrophes. It is terrible for funding a project.

We had a budget. We knew what it would cost to get the OpenCTD off the ground (actually, I didn’t, but we thought we did. More on that later). We raised about half of what we needed. But we did a “keep what you raised” model. So now we had donors that we owed results, but hadn’t raised enough to really push the project forward. Had we done an all or nothing campaign, we would have put the project on ice, focused on other things, and then tried again in a year or two.

Here’s the thing for new crowdfunding campaigners, if you raise money and don’t deliver, you have zero credibility. We couldn’t launch another OpenCTD campaign until we delivered on this one. The last reward from our first campaign, funded five years ago, was delivered last month.

I made it too long. We thought we were being clever. “It’s summer,” we thought. “A lot of scientists are doing fieldwork. They won’t want to miss this.” So we made the campaign 3 months long. It dragged on and on. People donate to crowdfunding the day it launches and the days it ends. You need a big first day to generate some media buzz. You need the campaign to go long enough that, if it’s newsworthy, it gets some good press and spreads through social media. But you want people to land on the campaign and think “I need to fund this today.” If they see a deadline months in the future, they’ll assume they have plenty of time, close the tab, and forget.

Remember when we bought David Shiffman a new pair of sunglasses? That campaign ran for 2 weeks. 90% of contributions came in on the first or last day. That campaign, which is patently ridiculous, raised a bit less than half as much as the OpenCTD.

I didn’t budget my time or for other externalities. Our budget was way off. We thought $10,000 was reasonable to get the OpenCTD off the ground. In reality, we’ve invested almost 3 times that into getting Oceanography for Everyone to where it is now. Our budget didn’t include any of our time to work on the project, travel expenses, training costs, or sourcing hardware at scale. Part of the problem was that we thought we had the OpenCTD pretty much ready to go, only to discover that we had years of iteration and experimentation ahead of us. The OpenCTD M4 only came out this summer, and it’s the first version that we feel encompasses our full vision for what the device should be.

I had no idea what the tax implications were. You pay taxes on crowdfunding revenue, just like any other income. That was also a failure of budgeting. That year, I ended up eating the tax costs. Now I have an LLC that can accept crowdfunds and minimizes (but doesn’t eliminate) the tax burden.

I made the rewards complicated and confusing. We must have had a dozen different reward tiers. They were confusing, non-complementary, and unclear. No one wants to slog through that. Crowdfunding campaigns should never have more than 6 rewards (unless you making something that comes in different models). Stick to these guidelines:

  • $1 to let people show their support. You’ll be shocked how many people just hand you a buck for nothing bu access to updates.
  • $3-$5 for something tiny, a thank you on social media, a mention in some acknowledgements section.
  • $10 -$20 for a small, physical token. Just remember to budget the cost of your rewards into this. ‘
  • $50 – $100 for something more exclusive. A signed widget, a skype-in-the-classroom, some art, some form of access.
  • $XXX for the thing you are making. Whatever it really costs.
  • >$500 for something big. Remember to budget correctly and limit the number of people who can get it.

That’s it. Keep the rewards simple. Look, for example, at my Patreon campaign (which supports this site): Andrew David Thaler is creating tools for ocean science and conservation. There are 5 tiers, $1, $3, $5, $7, and $10 (Patreon is a bit different in that it’s a monthly donation, so multiply those by 12 for a year-long campaign). The first three are increasing level of access. The penultimate is a physical thing, and the final is access to something very specific which appeals to a narrow but enthusiastic band of my audience.

I targeted the wrong communities. We thought the OpenCTD would most appeal to formal researchers working on a limited research grant, so we hit venues like Nature and used professional mailing lists. What we missed was that the community that really cared about the OpenCTD and Oceanography for Everyone was citizen scientists and open-science hardware developers. Instead of enduring the Sisyphean task of convincing researchers already affiliated with institutions that had access to fancy commercial CTDs, we should have be liaising with the groups already working on low-cost open-source instrumentation and building our community from the ground up. We naively thought that they would find us rather than putting in the effort to reach out to those communities.

And that, more than anything else, prevented us from reaching our goal.

Crowdfunding can be a useful tool for funding scientific hardware and research, but too often we focus on the success stories, those that win big, rather than looking at the losers and thinking about what went wrong. I’ve seen many crowdfunding campaigns fail over the last five years and all of them have several of these traits in common.


Hey Team Ocean! Southern Fried Science is entirely supported by contributions from our readers. Head over to Patreon to help keep our servers running an fund new and novel ocean outreach projects. Even a dollar or two a month will go a long way towards keeping our website online and producing the high-quality marine science and conservation content you love.


Marine science and conservation. Deep-sea ecology. Population genetics. Underwater robots. Open-source instrumentation. The deep sea is Earth's last great wilderness.


Connect with SFS