How we raised $484k on Kickstarter to make Glowing Plants

One of the goals of the Glowing Plant project was to inspire others to look at crowdfunding as a way to reach their own goals within DIY Bio / Synthetic biology. Consequently, we’ve been pleased to see a number of other projects launch. Sadly none of them have raised quite as much as our project, so we wanted to share some of what we did in the hope that these tips and tricks will be useful to others planning their own campaigns.

You can check out our project here: http://www.kickstarter.com/projects/antonyevans/glowing-plants-natural-lighting-with-no-electricit

Preparation

Once a crowdfunding campaign gets underway you will become overwhelmed and insanely busy. Also, crowdfunding campaigns have a significant momentum effect so it’s important they are as well presented and developed as possible at launch. This means planning for the campaign is critical.

We started formal planning eight months before the launch date, and Omri started telling people about the project eight months even before then. The most important thing at this stage telling everyone we knew and met about our plans to make a Glowing Plant, we went to meetups, gave talks and just told people we met. We literally talked to everyone we met (offline) about the project, it didn’t matter who they were. We didn’t discuss it online because we didn’t want to leak it beyond our control.

Sharing the project openly and transparently up front achieved three goals:

We wanted to know how people felt about this, if people were excited about the project. People wanted to talk about the project, and we learnt they wanted the seeds
It established a personal relationship with people, so when they saw the project they felt a personal connection to it, and
It allowed us to build partnerships broadening our social network. Especially in the first phase of the campaign your reach is as big as your network building that crucial early momentum. Many of these partners also became part of the project.

Our biggest fear at this stage was that someone would launch before us, but that turned out to not be the case.

We also did a ton of research online, looking at other campaigns (backing a few we liked) and seeing what did and didn’t work. Both Kickstarter and Indiegogo publish information helping project creators plan their projects and we read all of that, taking note of what they thought was important as well as what other bloggers online thought.

As part of this research we learnt that 80% of projects that get to 20% of their funding succeed, so this became a core planning goal. We realized that getting to 20% would be easier if we closed a few big backers early on so we went on the road and sold those rewards in person, sometimes giving them additional perks as part of the package. One of these was a $10k backer, Cambrian Genomics, who supported the project within the first hour giving us instant credibility that we could reach the target and igniting a wave of optimism online.

Another tip we got from the founder of Pebble who gave a talk at Singularity University. He convinced us to look at the landing page and treat it like a product in itself. You don’t launch a product without getting user input (I’ve made that mistake before!), so we showed the preview page to over 100 people in the month prior to launch. These people included our advisers (many of whom reviewed several times - thank you!), room-mates, friendly mailing lists and even a few people I stopped in the street in San Francisco.

The feedback we got from this group was incredibly valuable - we changed the reward pricing dramatically two days before launch to what people wanted to pay. Reward pricing is so important, you have to offer people value, and while there are cleverer ways of getting pricing right we found the simple question ‘would you support this project’ the most effective.

More feedback we got (from a friend of a friend who works at Kickstarter) was not to get too technical on the landing page. We got a ton of criticism about this after launch from scientists, but it was the right decision, the landing page should be simple and most importantly actionable (back the project today!). That’s why we stripped away all the text and just used images on the page, remember most backers don’t understand the details and trust you to know them yourselves. The landing page should be simple and actionable. Don’t use too much text (that’s why we had so many images). If you must give technical details, include a link to a page on your website. We did actually launch with such a link, but it was tied to a media-wiki page and was attacked by spam-bots on launch day after the traffic we got so that didn’t work and maybe it would have reduced conversion rates anyway as people navigated away from the page. Another reason for not going deep into technical details is that people will start discussing how you are going to do the science on social media, which drives traffic and maybe some new ideas you hadn’t considered.

One big decision was Kickstarter vs Indiegogo. We went with Kickstarter because we thought it would drive more traffic to the site. Our project was initially rejected (with a very vague email saying we didn’t comply with 'guidelines’) which was a good thing as we weren’t ready at that stage. We used our personal networks to reach out to Kickstarter employees to get feedback on why we were rejected and then spent the next two months fixing those issues - this lead to a much better campaign and in the end 55% of our backers found the project directly on Kickstarter which is a much higher figure than Indiegogo. We hope they reverse their ban on GMO’s as rewards in the future, and they’ve indicated informally to us that this might be possible, but if you can meet their guidelines I think it’s worth taking advantage of their curation as it leads to higher quality projects.

One campaign we studied intensely was 'The Ten Year Hoodie’ - we were amazed at how much money they raised for such a simple product. We were really inspired by their video, it’s story arc and energy but what really struck us was how the campaign was about more than just the hoodie, it was about inspiring a whole new movement in manufacturing. Our campaign goal was to inspire people with synthetic biology, so we made that a core feature of the video. I think backers support a project and take a risk because it’s about more than just a product so we really spent time crafting and focusing on the 'why’ message even more than the what, on how our project fitted into broader trends.

A good video is really important. Kickstarter says you can make it yourself, but we were looking to hit the ball out the park so we decided to outsource it which turned out to be a great decision (If you want one talk to Rick Symonds at www.ricksymonds.com). Doing this properly take capital (think thousands of dollars) but it also signals to backers that you are serious about the project if you have put your own money at risk before launch. Choosing the music is critical, and I spent nearly two days reviewing clips on Audio Socket and The Music Bed. Again, good music costs money and also caused us some trouble as we bought a license for online broadcast that didn’t cover TV broadcast, but honestly that’s a success problem!! We reshot the video three times (thanks for your patience Rick!) based on feedback from potential backers.

Setting the target goal is hard, you want a low goal that looks attainable (and which you can get to) but not so low a goal that executing will be impossible. Momentum matters, and people like to back a winner, so the lower goal you set the better - most projects that reach their goal go significantly over it. We set our goal at the minimum amount of money such that if we raised it we would want to do the project - this approach means you can do an all or nothing campaign, as with less than our goal we were going to struggle to execute.

Running the campaign

We planned the campaign into three phases:

getting to 20% of our target funding goal, based on reaching our friends and family who all knew the campaign was coming in advance and had promised to back in the first few days
getting to the goal of $65k based on reaching out to the tech community/early adopters
going beyond, based on reaching the mainstream/national press and to the broader public (this phase was not well planned in advance, and what we did plan eg gardeners mostly failed) Within each group we wanted to create a 'surround sound’ effect so that people in that segment saw the project 3-4 times in a week and hopefully decided to back it.

We launched at 9.30am Pacific Time on a Tuesday, selected because it gave us nearly the whole week to generate press but wasn’t Monday morning when people were grumpy after the weekend. The 20% target all knew the campaign was coming, and what time it was expected to launch and we had a bunch of emails ready to go to them and some friendly mailing lists the second we went live. Somewhere between 10 and 20 people were anticipating the launch and ready to start sending messages to other people and on social media.

We then posted to HackerNews, which didn’t yield too many donations. However it did result in people posting our campaign on other social news sites like Reddit and Slashdot, resulting in some significant traffic over several days. Reddit was especially powerful, it brought more donations than the New York Times article!

Living in San Francisco Antony knew someone at Techcrunch and this helped us get one of the journalists to cover the story. We offered them an exclusive in exchange for them holding off until after we launched the campaign and that article led to many more - as we learned again, all the leading blogs are read by other blogs who will want to write follow up articles once you become a news story (this was also true with mainstream press). Thus the first story is the hardest to get.

After this it was mostly momentum and being prompt and responsive to incoming interest. Journalists are busy and if they reach out the least we could do was reply as soon as possible to schedule a time to talk. Antony worked 16 hours a day dealing with the onslaught, and we had two interns helping answer messages (we got hundreds per day). You can’t run a campaign like ours part time or in addition to other activities.

One trick we learnt was syndication. An obvious example of this is tweet or facebook shares of articles, but some news agencies also push articles out to other sites in syndication deals, for one agency this led to five times the traffic from the original article so it’s always worth asking journalists about this. This was the reason behind the incredible performance of the SingularityHub article which was our highest source of backers.

Another thing we learnt was the value of controversy, so don’t be afraid of it. We had NGO’s who didn’t like our project pitch mainstream news agencies for us, resulting in press we could never have got ourselves. You are not building consensus with your campaign, you are firing up your core supporters, your core believers, so much that they want to give you money now even though you don’t have a product. This is really hard, because if you get this right a lot of people are going to hate on you. But you want discussion and debate, that’s what drives passion and that’s what drives social media. People share your project when they care, and controversy makes them care. Anything in Syn bio will be controversial, don’t be afraid of that but just remember stick within clear ethical lines (define these yourself, don’t let others try to define them for you) and of course the law.

In the last days of the campaign we used Hootsuite to schedule hourly tweet/fbook messages counting down until the end. It’s hard to track if these worked, but they created urgency at the end to go along with the Kickstarter reminder email.

What didn’t work

Lots of things went well with our campaign, but there were a few which didn’t:

Stretch goals: we hoped for success but we didn’t plan well enough for it with our stretch goals. I wish we had made them more frequent milestones and things which benefited all backers. Every project is different, but we should have outlined which additional experiments we could do with additional funds and linked those experiments to the stretch goals explaining how they would make the plant brighter. Still I’m very excited about the Glowing Rose!
Duration: Kickstarter recommends 30 days, we didn’t listen and did 45 days. Most campaigns are a U shape, with a huge peak at the start, then a long lull, then another rush at the end after Kickstarter sends the reminder email. If you extend the campaign all you do is extend the lull which means more stress and longer until you can start working.
Facebook ads: We tried running facebook ads to drive traffic in the lull, they didn’t work though I wish we could have targeted people who liked the page and their friends.
PR Agency: We had a good conversion rate (2-8% depending on source), so figured more traffic would lead to more conversions and hired an agency that contacted us. They got a few articles but nothing compared with what we did ourselves, journalists want to talk directly to the founders you don’t need a middle man.

Post campaign

We vaguely thought we’d plan the post campaign during the actual campaign, big mistake as we didn’t have time. We should have developed the strategy for that before launching as the campaign is just crazy. The most important thing is to plan where to direct traffic after the campaign ends, as Kickstarter locks the page the second you finish. If you will continue taking pre-orders you want to decide that upfront, so that it can all be setup. Trycelery.com and Shopstarter.com seem to be the leading platforms to help with that.

You also want to think about SEO post campaign. We were recommended to give a redirect link out to journalists (eg www.glowingplant.com/kickstarter) for the page rather than the actual page. That way you can change it later and keep the SEO juice for yourself.

You also need to plan for defaults on Kickstarter (less a problem on Indiegogo which collects money upfront). Mostly these defaults come from expired credit cards, and you can expect to spend the first week after the campaign chasing these people up to get new card details so don’t go on holiday too fast! After about one week the campaign funding gets locked and a week or so later you get the funds. Our net proceeds were $432k so about 89% of the total (the remainder is defaults, amazon fees and Kickstarter fees) - Don’t forget to budget this 11% cost in your planning!

Conclusion

The campaign was one of the most rewarding events in my career. Intense, but an incredible feeling seeing the numbers go up and knowing people were putting their faith in us. We take the trust and responsibilities of the backers very seriously, and the plants are already glowing. We can’t wait to ship them to everyone next fall!

We benefited from so many people’s advice before launch, we want to pay that forward to others. If you are planning a similar campaign please do get in touch - just make sure to give us at least a couple of weeks before launch date to get back to you.

Appendix:

For the geeks here are our traffic stats: