Concerning anipan
Concerning anipan
This morning I was pointed to a blog which draws comparisons between anipan and inkr. inkr hadn’t actually been promoted yet, so I was surprised to see that somebody had already stumbled upon it and made a connection.
I know this is something that I need to publicly respond to.
I want to be clear that I have nothing to hide about my involvement with inkr, which itself is a really exciting proposal for artists that I’ll get to in a little while.
To begin with, I’d like to give a little bit of background about who I am.
I’m Edward Walsh, better known online as “eddz”, a British user experience (ux) designer. I’ve been involved with several other online services for creators, starting out with MangaBullet, which I built with Marco de Vita and Megan Mahoney. It actually started as a hobby project which, to our surprise, became quite well-known. So we ran it for about 4 years before re-imagining it.
After MangaBullet I worked with pixiv, a Japanese online manga art gallery. We talked on and off for several months before finally meeting in London. When we next spoke, I was asked to spend time advising pixiv on how to expand their international audience. As a ux designer with some background in building a gallery for “overseas” manga artists, I was well-prepared to help them out.
Megan and I demoing MangaBullet to the pixiv team.
While at pixiv, Megan and myself came up for the idea for anipan, which I’ll discuss beneath this section.
Since then, I’ve also worked as the interim chief technical officer for CUUSOO, which you might know for its work with Lego (Lego CUUSOO). It’s sort of like a Kickstarter for Lego builders. If your Lego creation gets enough love then Lego will turn it into a real set. The Minecraft and Back to the Future Lego sets were created in this way.
Now, I’m working on inkr, which is a project by some people involved in MangaBullet, pixiv, CUUSOO, and Tokyopop – as well as myself.
So, that’s my relationship with services for creators. Onto anipan.
–––––
The anipan story
As mentioned before, Megan and I came up with the idea for it while working at pixiv. The custom web framework for anipan was actually built out of pixiv’s office in Tokyo.
Working on anipan during the first days of the project.
Because MangaBullet was built as a hobby project at a time when none of us had a lot of experience creating web services, the site was quite bloated and expensive to keep online. Though MangaBullet had a premium account option and some ads, the funds we earned from this weren’t enough to keep the service online (let alone pay a salary or anything like that). As a result, most of the costs for MangaBullet came out of my own pocket.
So we wanted to build a better, more reliable, and cheaper-to-run version of MangaBullet, which is how the anipan project started out.
After coming up with the idea, we worked on the project for about 5 or 6 months while also doing our own things. At that time, Megan was studying at university, while I was working as a designer and engineer for an online school yearbook publisher (which was about as fun as it sounds).
We didn’t actually come up with the idea of releasing anipan as a new site until quite late into its development. Originally, we planned to release it as a rebuild of MangaBullet, hence the reason it looked and felt quite similar. We had a lot of problems coming up with the name for it – we wanted the name to be based on MangaBullet to make it feel like an evolution of the site. One thing we learned while in Japan is that nobody could pronounce the name “MangaBullet” properly. Plus, the use of the word “bullet” was a little intimidating. So Megan came up with a cuter name: manga -> anime -> “ani” + bakery -> “pan” (bread) = “anipan”.
A very early version of anipan, with live events working.
We wanted to complete development, but were troubled by time and costs. We knew it would take another 3 - 6 months of full-time work to finish, and we would also need to pay for its servers. Fortunately, we’d learned about Kickstarter as a way to raise funds for small projects, so we decided to pick that route. Since it turned out that Kickstarter wasn’t available outside of the US at that time, we picked the next best option – Indiegogo.
We decided that we’d need to raise a minimum of about $18,000 to pay for the rest of development (bear in mind that the average salary for a developer in the UK is about $80,000 per year) and server uptime, which we estimated to be $500 per month.
We also knew that to raise the funds, we’d need to provide some kinds of physical rewards, which we estimated would cost about $5,000 including shipping.
This is how we broke it down: • 6 months of full-time development ahead of launch: $15,000 • 6 months of server uptime after launch: $3,000 • Rewards for backers: $5,000 • Fees taken from Indiegogo and PayPal: $1,750 = $24,750
So we set the amount at $25,000. One barrier to getting there was that people hadn’t really heard of crowd funding or Indiegogo at that time, so it was difficult to communicate what we were doing exactly. But we reached the goal, right in the last few minutes, with the support of our friends, family, and fans. But it wasn’t an easy journey.
A screenshot of the anipan pitch video. The music for it was produced my brother.
While running the campaign, we found that we actually didn’t have any time for development. The campaign, for a month, was everything – we lived and breathed it. And after that, we continued to live and breathe it for more than another month. We had more than 400 amazing people who had backed our project, and we had to send at least one package to each and every one of these people. And we did. It took a lot longer than we expected, but we got there in the end (my own mother shipped the last package from the UK more than 6 months after the campaign ended). But that was only the beginning.
Though we’d spent a lot of time preparing for the ordering and shipping process, the rewards cost a lot more than expected. When people sent us emails and messages about not receiving rewards (there were a lot of them), we had to send replacements. Some people received two packages because of the time it took for the international parcels to arrive. In the end, the rewards cost us around $8,000 – which is approaching double of our original estimation.
This wasn’t even a fraction of what we ended up shipping!
When we finally got around to development, we had to work a lot faster than before, as we’d already burned through most of the funds that had been raised. What we had left needed to be split across development and servers.
We ended up finishing a stable-enough, semi-complete version of anipan to release. We expected that this would be fine for our audience of backers, as well as the people who requested invitation codes.
When we put anipan live, it felt sort of like a Thanksgiving sale. It was both a blessing and a curse: more users than we could have ever anticipated immediately poured in to visit the brand new site. What was fine running privately was not good enough to handle hundreds of times more traffic. Our server configuration fell over, and we had to set up new instances to handle the load. This was expensive. And by this time, we’d spent all of our funds. From the very first day of release, we were already self-funding anipan. And this was while continuing to self-fund MangaBullet.
We decided to run MangaBullet from the anipan servers for about 6 months before shutting the service down, which reduced costs by a substantial amount. But it took a lot of re-tooling, which took yet more time, and we were still burning a lot of money that we didn’t have.
The error page that displayed repeatedly while moving MangaBullet onto anipan’s servers.
Please remember that both anipan and MangaBullet were built and ran in the spare time of the people involved, and that unlike many other web services, they were not designed by businesspeople or with the idea that they would become businesses.
With that kind of unexpected momentum, we had to make the very quick decision to run anipan more like a business. It was either that, or we would most definitely fail.
At this time, Megan and I were back in Tokyo for her third year of university – the flights had been financed long before the campaign ended, and we’d saved some funds aside for living costs, which began being eaten into by anipan. My original plan was to get back into client work (which, shortly after, is what I ended up doing).
So thinking like a business, we turned to private investment. We looked for people and businesses who might be interested, and we found a few who were. One was of particular interest since we had a shared vision.
Back to pixiv for negotiations.
pixiv wanted to out-right buy anipan, create a subsidiary which focused on international operations, and make me CEO of it. We had several meetings, and the offer was so very tempting. We had an informal agreement – we shook hands with the CEO and COO and decided to work out the details at another meeting. But it turned out that what pixiv wanted and what we wanted were different things. We wanted to keep anipan running and working on it as usual. We didn’t want it to become swallowed up and shut down (which is ironic, given the consequences).
So in the meantime, while the pixiv deal was on hold, we kept working on development. Expecting to raise more funds from investment, we hired a developer, Jamey, to work on improving the code and server configuration, while I worked on the design and management.
We ended up running completely out of any funds whatsoever, so we had to take on client work in the meantime. This arrangement worked well to begin with, since it allowed us to work on improving anipan – fixing bugs and optimising the back-end, for instance – which we did for the next couple of months.
Jamey getting the anipan local server (a Mac Mini) running.
Unfortunately, this didn’t work out in the long-term, since anipan was eventually put on the back-burner in favour of the tight client deadlines. There were a lot of features which were started and abandoned or released half-complete, such as paint chat, live messaging, hand-drawn comments, and even an export function which allowed you to automatically upload artwork from your accounts at other services.
Although we were able to make the site a lot more stable, our focus had shifted to client work, and interest in anipan had understandably gone elsewhere. As quickly as we scaled-up, we scaled back down. Our developer left less than 3 months after joining.
Although the code had been overhauled, the site that released and the one we ended up with were more-or-less on par with each other from the perspective of a user. We started with a site that had great momentum and weak infrastructure, and ended with a site which had great infrastructure but weak interest.
The anipan app, which was not completed.
To continue financing the site (and my own costs of living) I continued with the client work. I was able to tinker with anipan in free moments, usually providing bug fixes or moving assets around. This continued from January until around June 2013. I considered putting up banner ads to finance anipan, but the math suggested it would still have to be self-financed for the most part.
As much as I wanted to continue spending time on anipan, the project had failed in every area imaginable. From the reward distribution, to the development schedule, to the site we shipped, to the servers collapsing, to the deals falling through, to the momentum disappearing, to our developer leaving, and to the project’s resulting humungous cost.
In December, anipan’s servers were suspended by our hosting provider after we were unable to pay the bill.
We did so much preparation, but ultimately failed publicly and humiliatingly.
I was able to learn a painful lesson that has shaped the way I approach personal projects, work, and life itself.
I consider myself fortunate that during this time I was able to have the help and support of the people who are close to me. They helped me in so many ways, from finding work to finding new ways to think and feel. I could have fallen flat on my face, but was lucky enough to always have a roof over my head, great friends and family, and an income.
For those who supported the anipan project – you guys are amazing in every way, and I’m deeply sorry that I haven’t been transparent about what happened to anipan behind the scenes.
I have a list of people who backed us for the digital rewards (which due to the state of anipan did not come to fruition), and I’ll be referring to this list in the near future.
–––––
As for inkr
I mentioned before that we haven’t started talking about inkr yet – we’re planning to start doing so next month, and have just opened up the Facebook page for it as of last week.
I’m disappointed that the first you’ve heard about inkr is a negative relationship with another project.
The difference with inkr is that it is a service which is being created by an incorporated company with independent financing and a stable source of income. I am involved, but make up just one fifth of the team – with probably only a tiny fraction of the total combined experience.
Late nights in inkr’s early days.
Because of my involvement in design and management, inkr obviously has a MangaBullet/anipan flavour to it which, depending on your taste buds, you might find either bitter or sweet. As an example, we’re building-in some of the live interactions, collaborative drawing tools, and gallery organisational tools that you might remember from the original anipan pitch video.
However, although there are some similarities, the service itself is very different – in the same way that other services I have worked with share similarities but are different from anipan.
A throwback to anipan, inkr features a grid layout – now with drag and drop functionality.
inkr is designed around the idea of artists collaborating together, and also allows independent artists to create multiple galleries using one account (such as to separate personal and professional works). inkr is calling these galleries “studios”, and you will be able to share a studio with multiple people.
It is also built on an api, meaning independent developers will be able to create apps and tools that allow artists and art appreciators to get more out of the service.
inkr has been in development for about 6 months now, and will be released once it’s ready. There will be announcements about the service very soon.
In the meantime, you can follow inkr at twitter.com/inkr.
–––––
I apologise for the wordiness of this blog. If you got this far, I’m genuinely impressed. I tried to write a tl;dr but couldn’t come up with anything meaningful without context.
I’m happy to answer questions that you might have. The easiest way is to contact me directly via Twitter: twitter.com/eddzw (if you’d like to send a DM, just drop me a tweet and I’ll follow back).
Thank you for reading.
