In this post, you'll find the complete story of a hardware crowdfunding campaign: from the very beginning, up to successful funding. I'll try to share as many details as I can, interlaced with tips and things I've learned along the way.

It all started when...

Genesis

Before we dive into the details, let me introduce to you The Skull. I'm not entirely sure how the idea came to be, but it was sometime in late 2019, probably inspired by some Halloween stuff I've seen combined with my love for PCB art.

So what is The Skull? She's a bare skull-shaped PCB with built-in hardware/coding puzzles:

As you can see, She has two red LEDs that light up Her eye sockets. There's also a resistive-touch button on the back. When you touch it — Her eyes light-up.

I built a single prototype unit for myself and validated the basic hardware design: the sensitivity of the resistive touch button, the current consumption in deep sleep mode, and most importantly - my ability to create a challenging riddle using this hardware setup.

Fast forward to June 2020, I decided I wanted to make more of Her, and let others have the joy (or pain?) or holding Her in their hands and torturing their minds with Her riddle.

Crowdfunding?

I've never done crowdfunding before. I always thought that you need to have a beefy marketing team and tons of connections and resources to run a successful campaign. Or so I have heard.

Also, I didn't have much faith in crowdfunding due to my past experience backing hardware projects (e.g. 3D-printers) and receiving broken products that caused me endless frustration.

So the thought of crowdfunding never crossed my mind. I was planning to produce a small batch of Skulls, then put them up for sale on Tindie, just like I did with my previous Capture The Flag riddle.

I started doing calculations trying to figure out how many Skull pieces I should make — 20? 50? 100?

My smart life partner, Ariella, noticed I was deeply bothered with this and suggested: "Why won't you try crowdfunding? This way, you'll be able to figure out the demand in advance."

And so, I decided to take her advice and started researching...

Kickstarter? IndieGogo?

These were the first two names that came to mind. However, I quickly found out the bad news: they only work with specific countries, and Israel is not on their list.

Then I remembered I once backed a project on a smaller, niche one. Few minutes of digging, and I found the name: Crowd Supply. Checked out their website - they specialize in open source hardware and electronics projects. Perfect match!

I spent a few hours going through their launch guides. I learned that they can help with the imagery and video, but more importantly: the take care of the fulfilment logistics. I hate logistics. That bought me in.

Submit

The hardware was ready (or so I thought). The riddle not so - I only had general idea and some (very broken) prototype code. But my plan was to complete the missing pieces while the campaign was running, and so...

On August 19th, I submitted my project to Crowd Supply. You can find the complete submission here (excluding some minor details that would possibly spoil the riddle).

Looking back, I was very optimistic about the launch date: September 14th. As you'll soon find it, it took almost 6 weeks longer just to set everything up...

Accepted

Two days later, and sooner than I anticipated, I got an email saying:

Hello Uri,

Thanks for sharing The Skull with us. Your project looks great! We are happy to accept it.

Hooray! And just like that, the rollercoaster ride started.

We ping-ponged a few emails, and a day later I signed the Statement of Work. That's a 4-page contract, which was surprisingly written in a human-friendly language and unlike most contracts I've signed in life, felt fair and not draconian at all.

My favorite thing about that contract: it stated Crowd Supply deliverables first, and only then listed my commitments. This was also unusual for me, and made a very good impression on me.

Overall, it seemed like they were really going to care about the campaign and help as much as they could: video, image assets, feedback on texts, promotion...

But most importantly, for me —
They were going to take care of the fulfillment logistics!

In exchange, I had to pay 12% fee (of total funding raised during the campaign), as well as 2.9% + 0.30/transaction payment processing fees (pretty standard).

In addition to that, there's a per-item fulfilment fee (ended up being $1/item).

That's not as low as Tindie's flat 5% fee (+payment processing fees), but I think it's a good deal considering the all the extra value that you get.

And in case you are wondering, my obligations were:

There were also some terms related to the Microchip Get Launched program, but eventually I haven't got accepted into that, so there's no point in talking about that here.

GitHub

Another fun fact about Crowd Supply: they manage all their campaigns on private GitHub repositories.

On August 26, The Skull project repository was born. I was invited to join it, and was immediately greeted by 7 outstanding issues:

Excited, I spent the rest of the day working out the messaging section (#1), Bill of materials quote (#5), and sent some questions about shipping the prototypes (#4), as I was determined get the campaign ready to launch mid September.

In a retrospect, having everything documented in a GitHub repo was very useful and made it so much easier to write this blog post.

Overall, by the end of the campaign, we had total of 47 issues with a whopping number of 995 comments!

Messaging

The first mission I got from Crowd Supply was to answer three questions:

1. What are the three most important things a potential backer should know about this project?

It's a Capture The Flag (CTF) Riddle, with multiple stages. The firmware (source code) for the riddle will be open, and solving the riddles requires reading and understanding the firmware code, as well as digging into the datasheet of the ATtiny45/85.

Apart from a riddle, it also double as decorative skull gadget that flashes its eyes when you touch a small grill at the back of its head. It's powered by a CR2032 battery that should last for at least one year in standby mode.

There is a background story I was thinking of, something along the lines of:

>    Many years ago, she was the queen of pirates.
>    Now, you are holding her skull in your hands.
>    Left with a tiny piece of the brain, barely alive,
>    it holds a dark secret.
>    
>    Awaken by a gentle touch, her eyes will glimmer,
>    and cease shortly after. They say eyes are
>    windows to the soul, and this lost soul is waiting
>    for you to unleash her secret and set her free.
>
>   Will you take up this quest?

The text is still draft, I'm open for feedback

My project manager, Chris, replied within hours and suggested:

I'm trying to decide if we can include nothing but some variation of the following on the pre-launch page. Put the tease back in teaser...

And so we did. He also did some word-tweaking to The Skull's background story, and you can find his improved version on the campaign's prelaunch page.

Little did I know how Chris's editing skill were going to play a major role in the campaign... But we'll get to that.

Onto the next question:

2. Who are your users and what problems does this project solve for them?

The users are people who are into reverse engineering / riddle solving / electronics and micro controller internals / code challenges. They need to be able to read and understand code, think out of the box, and enjoy solving complex problems.

I asked one of the users who bought my previous CTF riddle and that's what he said:

> Geeks, makers, Security researchers, hardware addicted, low level techie ppl.
> problem it solves: boring life i guess?
> Does it have to solve a problem? Why not have fun & educate myself?
> why not a listing of good things it brings into the table?
> alternative text: lack of a playground for hardware hacking which can be both fun & educational at a reasonable price.

The takeaway from here - when you sit down to work out the messaging, ask potential users for help. I would have never come up with "boring life" myself, yet it's so true.

The last question was:

3. Tell us your backstory. Why did you undertake this project?

Note: feel free to skip my answer, it's a bit long...

I love building physical things with hardware and electronics. I also love solving complex challenges (e.g. a few years back I spent about a month breaking into the encrypted firmware of my 3D printer). Last year, I spent a few months working on the hardware for a smart conference badge. As a side gig, I created a small CTF add-on for the badge, and after the conference concluded, I spent a week or so creating the actual challenge for the board, and put it up for sale on Tindie. When I published it, I got a lot of positive feedback, and some people liked it so much that they replicated the setup just so they could solve it immediately without having to wait for shipping the hardware to them.

This motivated me to think about the next riddle, combining my love for electronics with my passion for challenge solving, which resulted in The Skull. I kept going with the same principal as the previous riddle: minimal hardware, maximum challenge ;-)

The new challenge involves 3 milestones / riddles (and I'm thinking of another one, which we will probably be able to add as a milestone):

Hypnosis - making the skull respond to your commands
Capture the flag - extracting the secret string from the chip
Mind Control - make the eyes blink in a specific pattern (maybe SOS or so?)

All the steps must be achieved without any electrical connection to the skull (so without reprogramming the chip or connecting to any of its pins).

Presently, I'm still working on the exact details of these milestones. Ideally, the first riddle would be possible to solve with any ordinary smartphone, and probably the next two riddles as well, but I'll know for sure when I finalize the prototype code.

As you can I tried to be upfront about the fact that I still had no concrete implementation of the riddle, only a rough idea.

But if you take a look at the final version of the challenge, you'll notice that the names of the milestones and their goals are pretty similar to what I originally had in mind.

BoM

I spent the second half of the day working settings up the Bill of Materials (BOM) on Mouser. For some unknown reason, I assumed this was mandatory, despite the issue that stated clearly this was optional:

Crowd Supply may be able to help get better pricing for some components used in your project. If you would like us to do this, you will need to let us know your bill of materials (BOM) by creating and sharing a Mouser project as follows:

For the early prototypes, I used parts from LCSC. Some of these parts were non-stocked on Mouser. Thus, I found myself wading through datasheets and looking for suitable alternatives on mouser.com for several hours, and ended up ordering these alternatives so I could test them.

Eventually, all that work was in vain. Even though Crowd Supply managed to get a competitive quote from Mouser, as they promised, I realized that shipping the parts from Mouser (US) to the manufacturer in China may cause delays and complicate things, so I just let the manufacturer deal with sourcing the parts.

In a retrospect, I really wish I paid more attention to the text and realized this was optional. This could have saved me hours of digging in datasheets and days of testing alternative parts.

En Route

Crowd Supply asks to send them a prototype unit for photographing, and also for sharing with potential reviewers. I realized shipping may take a long time, so I started bombarding them with questions right away:

  • What would be the best method to ship from Israel to the states?
  • How many units should I ship?
  • Should I solder the programming header?
  • What firmware should they include? (at that point I didn't have the actual riddle, only a very rough, untested prototype code)
  • What to write in the Item Description for customs? and which HTS code Should I use?

It took us a week to figure everything, but eventually I shipped two units, programmed with a quick demo firmware that flashes the eyes. One unit also had a programming header, so Crowd Supply could eventually upload the riddle code in case they find a reviewer (it never happened).

Ready to ship!

As for the courier, UPS was prohibitively expensive: $114 for a letter, $303 for a box. EMS was more reasonable, at $33 but couldn't guarantee shipping time.

DHL ended up being the best solution: $33 shipping with 3-5 day delivery time.

However, they wouldn't let me ship without filling in the Commodity code. Crowd Supply suggested a code that from another campaign:

Did I mention I hate logistics?

Like a game that is not fun at all. He nailed it!

Prelaunch

Crowd Supply received my Skulls on September 10th, but their photographer was busy above his head shooting other projects, so The Skull had to patiently wait for two more weeks. I started to realize that launching mid-September — not gonna happen.

Meanwhile, I worked on the Pre-launch Page GitHub issue (#3), filling out the required information:

Project title: The Skull CTF

Project teaser: Here are a few options. Let the best one win:
- Mind-bending hardware puzzle in 3-parts
- Spooky hardware challenge for the brave
- Can you hack into its ATtiny brain?
- Tough low-level quest for the hardware-addicted
- Serial.println("pwned!");

Project Creator: Wokwi
Wokwi creates tools for the Arduino and maker community, such as AVR8js, the Open source Arduino Simulator, Arduino Library Examples Playground and GoodArduinoCode. We also create fun hardware CTF puzzles from time to time!

Lists of product features and/or specifications:
- 3 challenging riddles built into The Skull's firmware
Powered by ATtiny45
- Runs on a single CR2032 battery that lasts for a year+ in standby mode

Product use-cases:
- Torture your mind with a creative reverse engineering challenge and sharpen your problem solving skills
- "Innocent" decorative Skull ?

Open source information
- Link to board schematics: I added the schematics to the repo, though I'm not sure if it is interesting/useful enough to be included in the pre-launch page?
- Link to firmware: Will only be available after campaign starts

Then, there was a moment I realized how lucky I am to have Chris as the project manager. He took my teaser suggestions, picked the one he liked, and turned it into a clever word play: Can you hack into ATtiny brain?

I find it amazing how taking out a single word from my original proposal made it so much better.

SkullCTF.com

The next two weeks were total silence. I waited for Crowd Supply to complete the Pre-launch page, and spent the time working on the firmware for the riddle, ordering parts from Mouser to test them, and... I bought the domain.

Skull.com was already taken. In a strange coincidence, it displays a photo that looks almost exactly like my Skull. But my Skull has better teeth.

So I decided to go for skullctf.com, and set up a basic "coming soon" landing page:

When the pre-launch page went live, I also added the green button saying "Subscribe to updates". Chris noticed it and wrote to me:

Lawyers, ha!

Financials

I also started working on the Financials issue (#6). You can see my answers to Crowd Supply's questions below:

1.  What pledge levels do you think you might offer?

The basic level would be one skull.

I'm thinking about offering a pair as well (that could be useful if we end up adding the 4th stage of the riddle, where you'd have to use 2 skulls to check your solution. Could also be nice if you wanted to give one to a friend.

Another thought that came to mind is offering a "classroom" edition with 10 or 20 units. Based on your past experience, is it a bad idea?

We can also get some custom logo (so given large enough quantity, the backer could choose their own logo and we'd add it in the available space on the back of the PCB, as we have plenty of such). I'm not sure it's worth the trouble in logistics though, WDYT?

Finally, I was thinking at some point to add an option to get a programming header + programmer PCB, based on the USBtinyISP board, though, again, I'm not sure it's worth the trouble given that most backers will probably be more interested in the riddle than re-programming the board, and that they can also use an Arduino Uno as programmer if they wish.

2. For each pledge level, about how much do you think it might cost to produce? For about how much do you expect it will sell?

I consulted the spreadsheet you prepared, first of all, thanks for that, it's super-useful!

I updated it based on PCBWay's pricing (which is somewhat expensive, we may be able to negotiate some discount with them).

Seems like 23$ per unit would do the trick.

If we go with pledge level that includes 2 or more units, then we can make them significantly cheaper per-unit (mostly due to the flat fulfillment fee). I think we can go with 31-37$ for two units, based on my quick calculation.

3. Approximately what do you expect your funding goal to be in US Dollars?

The sheet says $909 (based on $23/unit), and it seems like somewhere around this price could work. If we play along with the theme of the riddle, then $1337 could be a good choice.

And if you are wondering about the mysterious spreadsheet I mentioned in my replies, here it is:

Figuring out the pricing and funding goal

We had some back and forth discussing the optional add-ons, and especially the USBtinyISP programmer. At that point, I also learned the Mouser is actually going to do the fulfilment for The Skull (surprise!):

Chris suggested to ship an off-the-shelf programmer instead of making one myself, which was a very good call: we ended up selling just 3 units, and designing / manufacturing an extra PCB for this purpose would have been tons of work.

Eventually, we came down to a decision:

Glasses

I also came up with 3D printable glasses for the skull. As you can see, I kept myself busy while waiting for Crowd Supply to finish working on the prelaunch page:

Ariella painted them with nail polish to create this summer collection!

I shared these photos with Crowd Supply, and this got Chris intrigued:

This definitely ignited his imagination, as you can see:

In a retrospect, I'm really happy we went with just The Skull (and the off-the-shelf programmer), and didn't have to worry about providing other accessories. It allowed me to focus on delivering the best experience I could for the backers (more on that below, in the Updates section).

September 24th.

The day we prelaunched The Skull. The landing page was pretty basic, and included a sign-up form, the photo of the skull, and my little piece of poetry, with Chris's touch-ups:

The Skull's prelaunch page

I didn't spend any time promoting the campaign at this point, as I focused my efforts on launching the campaign. Nevertheless, the sign-up form amassed 68 subscribers over the following month.

Your mileage may vary - Roni, a friend who recently launched his own project on Crowd Supply, reports that he had 410 subscribers when he launched.

Shock!

As soon as the prelaunch page went live, my inbox filled with emails from GitHub. Apparently, Crowd Supply's bot bombarded me with a bunch of new issues:

You finish one, you get 10 others!

10 more issues before I was ready to launch... Actually, the funding goal was already set before ($1337, see "Financials" above), and we realized that Comparison table wasn't really relevant for this specific campaign:

Nothing compares to The Skull...

So just eight more issues to go! Time to start working...

Video

Crowd Supply was supposed to handle that part, but they still needed my input. So I did some research (actually, asked Chris to do it for me):

Chris's answer we very insightful:

Definitely Mr. Hoverboard. :)

From among my relatively recent campaigns...let's see. Inkplate 6 at least demonstrates a few of the techniques we often use. I'm trying to think of a good game- or puzzle-oriented video, but Snekboard is the only one that comes to mind. it's quite frantic, though, which might be hard to pull off for The Skull. Same with video-game products like Gameduino Dazzler. Even something like Circuit Stickers, probably has more motion than can count on. And we really try to avoid voiceovers and talking heads (like in the Circuit Classics video). Maybe something like Very Useful Circuits?

A few of us really like the x0x-Heart video, and that style could be used for pretty much any campaign. Unfortunately, we're not really set up for stop-motion, so you'd have to put a lot of those shots together yourself.

I suppose you could do something like the component-callout segment (at 0:32) of the USB Armory Mk II video, but as a parody. Point to the skull's tooth—using very serious looking pointer—and identify it as: "Tooth." Same for the eyes, jaw, etc. Then flip it over and point to the "Brain." :)

And maybe we could shoot a handful of different-looking skulls in various, real-world locations? (A bit like Glow Glyphs video, I suppose.)

However, after looking at all those campaign videos, my conclusion was:

Seems like a "cool" video doesn't make a lot of difference...

Chris sort-of agreed with me:

So I decided to stop worrying about the video and let Crowd Supply take care of it, and instead focus on the campaign page's text.

Text

When the campaign went live, I got messages from several friends. They all said they really loved the copy, and asked if I wrote it. That was the moment I decided to write this blog post. Yes, the one you are reading right now.

The short answer was: "It's text I wrote, upgraded by Crowd Supply's staff". And now, finally, for the longer answer...

Crowd Supply supplied the campaign page outline:

It took me about a week until I forced myself to sit down and write the first draft of the copy. It's too long to include here, but I encourage you to take a look and compare it with the current campaign page.

Chris added his own voice to the text. He removed my jolly emojis and replaced them with "spooky influence", as he said in his own words:

I liked it. If you were too lazy to compare my original draft with the final campaign page, here are just few examples of "spooky" influence that Chris added:

lunatic void.

But my personal favorite is the term Chris coined for possible future expansion boards (a.k.a shields / daughter boards):

Slaughterboards ;-)

Pledge Levels

We only discussed a single pledge level (see Financials above). The issue had plural wording, "pledge levels", so  I suggested 2-pack, 3-pack, and a classroom pack:

Chris ran the numbers and gave his own input:

And eventually we came to a mutual agreement on a pricing scheme:

Eventually, we had to re-do the pricing, due to changes to Crowd Supply shipping scheme. But before we get there...

A baby

I tried to finish working on all the issues before the end of September. You could imagine I was very excited about the campaign and wanted to launch it already. But it wasn't the only reason:

The baby arrived a little earlier than anticipated, and so, I took a little break from working on the campaign. This gave Crowd Supply enough time to finish the video.

Meanwhile, Chris and I entertained ourselves while finalizing the pledge levels:

Repricing

Things went into high gear again as the end of October approached, as we had a very concrete reason to launch:

To make things even more challenging, Chris updated about the new shipping calculation methods that they introduced, so we had to re-work the pricing:

When I woke Friday morning and read those comments, I suddenly realized: Halloween was the next day, so we had to finish everything that day!

Chris and I kept ping-ponging throughout the day (actually, evening for me — time zone difference, ya know...), until we eventually figured out the new pricing (and rejected some of my "wild" ideas):

So the first 42 skulls will sell for $16, and then $20. And now you know how this was decided.

What else?

Announcement

Before we could launch, I also had to come up with "launch announcement" update. I kept it simple:

Ahoy Mateys!

We're live! A dead-skull is pleading to get on your desk and challenge you with a mind-bending reverse engineering puzzle. Let's see if we can collect enough coins to bring it (back) to life...

and Chris "upgraded" it:

Ahoy Mateys!

This long-forgotten skull yearns for a place on your desk...a place from which, upon waking, it will endeavor to shatter your mind with a horrifying reverse engineering puzzle. Show it you are not afraid by helping us gather enough doubloons to raise it from the dead!
Doubloons? I learned so many new words during this campaign...

Ok, so now, are we ready yet to...

Launch! 🚀

It was already midnight for me, and while we nailed down most of the open items, there was still no sign of the campaign video. Not all hope lost, as it was still afternoon in Portland. I let Chris know I was signing-off:

When I woke up the next day, I ran to my computer to check if the campaign was up. All I saw was a single message from Chris:

And no campaign. I figured out they probably still had work left on the video, and realized it means we'll postpone the launch to the following week. Oct 31 was Saturday, and I didn't expect them to work on the weekend.

But Chris and Gabe were determined:

The video was truly brilliant, in my opinion. Totally surpassed my expectations. If you haven't watched it yet, it's waiting for you at the top of the campaign page. 🍿

We did it! The campaign was live!

Early Birds

What if nobody cares? What if all this work was in vain? Or maybe everyone is in hanging out in Halloween parties in it wasn't a good launch date?

I anxiously refreshed the campaign page every minute, waiting to see if anything happens.

Fortunately, I didn't have to wait too long. The first two sales happened within just 5 minutes!

And they kept coming...

Two hours into the campaign, I sent Chris this screenshot:

Two more hours, and we doubled that:

Things were looking really bright! I had 14 orders of a single skull, and one double Skull. Soon I'm gonna be rich!!1

But then things started to slow down...

Sales

I knew you'll ask. So I summarized the daily sales into a graph, just for you.

See that slope on the left? On the launch date, I had 17 orders totaling $288. But this moment of glory was quickly gone. The pace of new orders slowed down:

I was still pretty optimistic about reaching the funding goal of $666. A week later we did!

Roni, the friend I mention above, says he's been seeing a similar pattern with his campaign: many sales in the first day, and then a major drop.

I also opened a Twitter account for The Skull, where I tweeted daily about the campaign's progress (inpired by year_progress):

Updates

A side effect of the campaign going live: my email was bombarded again with GitHub issues opened by Crowd Supply's bot. This time it was one issue for each weekly update I had to write. For instance, the first one was:

Hmm... what should I write? If I was a backer, what would I want to know?

Crowdfunding projects usually send out updates telling you about new features they are working on, challenges they are facing, etc. In my case however, things were a bit different.

I was working on the firmware of the riddle, perfecting the fine details, and trying to achieve a fine balance between challenging riddle and positive user experience (believe me, it's hard!).

If I shared any part of the work-in-progress code, or anything about my own challenges I faced while working on the firmware (and my creative solutions / workarounds for them), I'd risk spoiling the riddle and giving away parts of the solution.

I ended sharing up only a single piece, one that wasn't directly related to the riddle but was still deeply technical: Removing a Curse from the ATtiny85 Fuses.

Fictional

I looked for inspiration in the past updates of other campaigns, when "Conversation with my dog" caught my eyes.

It was special. I couldn't share all the j̶u̶i̶c̶y̶ boring technical details with my backers, but I could entertain them with some made up stories. And so I did.

The first campaign update told about the Shaman we hired to bless our compilers, and led us to find two mysterious scrolls in the depths of the ocean (compare with my raw version, before Chris's touch ups).

I created a script that would let me reshape part any piece of C++ code according to an ASCII text file template. I then took a piece of The Skull's WiP firmware code and used the script to mold the code into two ASCII skull templates:

Input: ASCII skull templates (+ some code I can't show here)
The first version of "Skull Code" scrolls, that didn't make it.

I posted the result to twitter in form of a survey, and also asked the backers to vote which one will make it into the firmware.

Overall, I spent about an hour coming up with the idea, and then nearly 3 days executing it: writing the code-reshaping script, fine-tuning the output, creating the graphics, and making up the story.

Some of backers seemed to really resonate with the unusual update:

For the rest of the campaign, I kept the same line of fictional, exaggerated updates. Every week I spent 1-2 days coming up with crazy ideas, researching, creating graphics and even inventing smaller riddles for the backers:

Chris kept spicing up the updates with words I never heard of before, lawyers threatening to sue me out of existence, and whatnot.

You can find all the in-campaign updates below (and even more post-campaign updates):

  1. Blessing the Bytes
  2. Eldrich Assembly
  3. Completing the Ritual
  4. Black Friday Special! — we raised the price just for black Friday
  5. Harnessing the Power of the Beaver Moon
  6. SPELLunking & SKULLduggery
  7. Answers Begin to Manifest...
  8. Spiralling Deeper
  9. Into the Darkness

If you go through the updates, you'll also notice that they always begin with campaign funding progress.

What started as a simple ASCII progress bar evolved into another mini-riddle, where I'd always encode the funding percentage, each time in a different way:

[▓▓▓▓▓▓▓▓▓▓▓▓░░░]81%
[▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓] NaN%
0b10010111%
CLXXXVII%
٢٠٢%
Two Hundred Eighteen PercentThe 48th Prime Number%
(1+2+3+4*5!-6-7+8+9)%
"{}%".format(ord('Ƕ'))
[זנב תולע]%
(2⁹+2⁵+2⁴+2³+2¹+2⁰)%
….. ——. ..—- %
五百九十六 %

I also spent a fair amount of time researching about "monkey hitting keys at random on a typewriter keyboard to recreate the work of Shakespeare", learning about nostalgic BBS text-based RPGs, and even about illegal entry to the Catacombs of Paris.

If you look closely in the campaign updates, you'll find traces of all these topics.

In some cases, I actually had to stop and not take these updates too seriously, as I was getting too specific with the description. For example, compare the draft for update 5 with the much terser final version.

As you can imagine, working on these updates was very taxing.

Sure, it was fun experience (for the most part), a good way to improve my writing skills (especially when you work with a word-savvy person like Chris), and provided an excuse for my brain to take breaks from working on the actual firmware.

But I don't think these updates really moved the needle. We had 6-12 sales per week, so it's hard to tell if there's any correlation with the updates. In any case, the updates were not an effective way to promote the project.

If you want your campaign to be successful, people need to know about it. In my case, I shared the campaign on my Facebook page and Twitter profile.

Many of my friends (and maker-friends) are from Israel, and 17 of the 78 backers were Israeli. That's more than 20%.

This indicates that sharing the campaign on Facebook was a good call.

I also looked at the "As Featured In" section of past Crowd Supply campaigns. This section includes links to blog posts and news articles that featured these campaigns. I then found the people who wrote these articles and reached out to two of them.

One of these emails actually landed an Article on Geeky Gadgets:

I also mailed Hackaday about The Skull, but after asking around in Crowd Supply's discord, I realized they rarely write about crowd funding projects. And just like everyone else, they ignored my email.

However, Hackaday and I did an AVR architecture and reverse engineering course together, and I sneaked The Skull into the intro video for the new course:

Pardon my messy hair!

I also sent Hackaday a link to an article I wrote about the ATtiny fuses, which indirectly mentioned The Skull. And this time they wrote about it in their blog.

Maximum

We hit the funding goal one week after launching the campaign. I kept thinking about ways to advertise and promote the campaign. But then I realized I never asked myself the most important question:

What do I consider as success? How many units do I need to sell to call this campaign successful?

After thinking about it for a day or two, I realized that the optimal number lies between 100 and 200 units. You can probably imagine why I wanted to sell at least 100 units. But what about the maximum? why stop at 200?

I'm coming from software background, where scaling is usually merely an engineering problem. Add another cloud VM or optimize your cache and serve thousands (or million) more users.

Scaling is hardware a whole different story. You want 200 ATtiny45 chips? You can usually find some supplier with stock. You want 20,000 units? that's a whole different story (especially nowadays with the chip shortage).

So more units usually mean longer lead times. And often also more logistics. I hate logistics!

More units also means bigger risk. If something goes wrong, you might have to redo the entire batch, doubling your costs.

200 units seemed like the sweet spot: the risk isn't too big, lead time should be pretty short, and yet it feels like a satisfactory number of people the enjoy the merits of the skull.

I only managed to sell 97 skulls during the campaign, just below the minimum number I was hoping to sell. But there was a surprise waiting for me...

Doubling

When we launched the campaign, Crowd Supply bot bombarded me with GitHub issues. I missed one of them, only to rediscover it a month later:

Digging in, it turned out that Crowd Supply were going to double the number of orders:

Perfect! and so... the campaign concluded with the following numbers:

Crowd Supply decided to purchase 102 additional units, bumping the total number to 199. Goal achieved! 🎉

In retrospect, I wish I paid more attention when initially reading the description of the GitHub issues. If you remember, this also happened with the BoM, where I assumed it was mandatory to buy the parts from Mouser.

Certifications

Crowd Supply requires you to figure out the CE certification for your product (so they can ship it to Europe). They have a guide on their website, which is a good starting point.

I was hoping they could help me with figuring this part out (as I was clueless), and their guide pointed out that the project manager have the relevant experience:

Apparently, this wasn't the case and they weren't able to provide guidance. Following my feedback, they revised the text in their guide, and no longer offer advice on the subject.

In favor of those of you who are going to launch a hardware campaign, this is what I gathered:

  1. Arduino Uno is CE/FCC certified and includes both marks on the back of the board (at least in Rev. 3). They also make the CE/FCC certs available for download.
  2. SparkFun does not certify most of their products, and do not include CE nor FCC marks on their boards. Yet they are able to distribute them through Mouser. Most of their products are RoHS compliant, but they do not provide documentation. Also, their take on FCC certification (source):
    We will seek clarification on this point and update this document in the future, when such clarification is available, but at this time, SparkFun will not be doing FCC testing on kits. An exception is their Artemis BLE module, which went through a full CE/FCC/ISED certification process.
  3. AdaFruit also sells boards without CE/FCC marks, such as the Circuit Playground Bluefruit (which includes a 2.4GHz BLE radio).  They discuss this on their forum. They also do not provide RoHS documentation, as they explained on their forum, they seem to cause more harm than good.

To summarize: if your product is similar to something Adafruit / Sparkfun sells, you may be able to do without CE marking / certification. Of course, take this with a grain of salt, I'm definitely not an expert (and in fact, know very little) about certifications.

Success

I consider the campaign successful.

I learned a lot from doing this campaign. In a retrospect, I'd probably spend less time working on the weekly campaign updates.

I had many moments where I wished I had a "normal" project, and not a riddle, so I could actually share my progress with the backers. Working on the Skulls I really felt alone many times.

Luckily, there are some people who really helped along the way:

  • Ariella, my amazing life partner. She was the one who came up with the crowd funding idea at the first place.
  • Benny who helped with beta-testing the riddle.
  • Irina Ilina, my talented graphic designer. She did the artwork for the final version of the skull, the glasses 3D model, and was responsible for much of the graphics in the campaign updates.
  • Roni who helped with the manufacturing logistics in China.
  • Amit who gave initial feedback and came up with the pirate theme.
  • and Chris from Crowd Supply. He was a great partner and I really enjoyed working with him.

Overall, running this campaign was fun for the most part. Sure, there were some glitches, mostly post-campaign/fulfilment related, but that's a story for a different time.

At last, the campaign was successful!

The Bottom Line

Running a Crowdfunding hardware campaign isn't a great way to make money. It is a good platform to test your idea, and get funding for building and manufacturing the first version of your product, but don't count on it to pay the bills.

I spent about 4 months working on the campaign. At the end of the day, my revenue was $2,905.35, and the production costs were $1,333.5, leaving me with about $1,570. And that's without considering the prototyping costs, shipping expenses, and the countless hours I spent on this project.

So the bottom line is: A good reason to run a hardware campaign is because you really want something to exist. If you are just about making money, software is probably an easier way.

And if you decide to run a crowdfunding campaign, give Crowd Supply a go. They are really good at what they are doing.