Hi, I'm Bemmu, and I run a twice-monthly Japanese candy subscription service called Candy Japan.
I studied computer science in Finland and visited Japan for two years as an exchange student to study Japanese. I liked the country enough that after graduating I came back and have been here ever since.
Twice a month we package a huge lot of boxes of candy and send them to people around the world. It has been running for 5 years now, and in 2016 had revenues of around $15,000/month.
Even before Candy Japan, I had already been running an online store that sold Japanese comic books. I ran it as a side thing during the exchange study in Tokyo. I would send out the shipments during our lunch break.
From what I understood, the work life in Japan is quite hectic, so after coming back to live here I was looking for some project I could take on instead of applying for a local company. The choices of employment would have been very limited anyway, as we live in the countryside.
Before moving here I went on a holiday with a friend to Thailand, and we started bouncing around some business ideas. The first "subscription boxes" had just started popping up, and I thought maybe there could be something fun we could send while living abroad.
What that might be exactly wasn't clear. Initially I thought to just send random souvenirs from Asia. There was also no acute need to pursue the idea, as I had a MySpace app which was doing great. You can imagine how that turned out. Only later did I pick Japanese candy as the actual product to send.
I wanted to start really simple, so I didn't even have a website when we launched. I just sent a PayPal subscription link to potential customers over email and then started sending candy to their address periodically.
It seemed that the customers were enjoying this, so I set up a simple website for the service. Since then I've built it out to automate things. It's a Python app running on Google App Engine (probably a mistake in retrospect). With my background I could just build it myself, so there was no investment involved besides my own time.
Some blogs covered the launch and sent a few dozen more subscribers my way. The blog mentions propelled the site to the top for Google searches for "japanese candy", sending even more customers. After about 3 months there were 300 subscribers, which seemed like a huge success since the site was now covering my cost of living in Japan.
I think a lot of luck was involved, but on the other hand I've tried a lot of things which have failed as well, and feel like the key is to just keep putting stuff out there, make sure people hear about it, and see if anyone bites.
After the first 300 subscribers I hit a ceiling that was very difficult to break through. I stayed at that level for three years, until finally I managed to break through to a new level of subscribers we are at now. There was no single reason for it, rather those subscribers came from a variety of different sources that started to align favorably.
Traffic sources are ever-changing, but over the years I have seen subscribers join through BuzzFeed, Reddit, Facebook, Twitter, AdWords, directories, and "top subscription boxes" lists on sites like Lifehacker and MSN.
Recently marketing is getting more competitive with literally thousands of new subscription boxes having launched since I started. That is a good thing in that people are now familiar with the concept, but it is getting difficult to stand out.
Nowadays I've had to resort more to paid channels, which is a quite different approach, as you have to very carefully calculate whether the traffic you are buying is actually profitable.
Customers get billed monthly either through PayPal or directly by credit card. We then send them two boxes of candy each month. With 600 members, sales are currently about $15,000/month.
Since I get paid in USD but buy everything in JPY, margins are quite unpredictable due to changing currency exchange rate changes. Some years I've barely made minimum wage, others I have been much better off than if I had worked for a company in Japan as a programmer.
As for payment integrations, PayPal was easy, but back before Stripe was available in Japan it was challenging to find a credit card processor that would fit my situation of being a Finnish expat in Japan. It took months to find the processor, get accepted as their customer and finally get recurring payments working.
The future seems unpredictable. I hope to grow subscriber numbers to over 1000 members, as that is where many of the shipping and product discounts would kick in, which would be a big boost to margins. Where exactly these new members would come from is unclear, but I will keep my eyes open as the internet landscape keeps changing each year.
It's the old joke of "making it up in volume", but in this case the math shows that the volume would really help. Getting any scalable paid traffic source to even break-even profitability would likely be enough to break through to that level, but also comes with risks of wasting a lot of money with unprofitable promotion experiments.
If I could go back and do things differently, at least I would try to be more prepared for credit card fraud. I had a very naive payment integration at first, which didn't take fraud into account at all. It was fine for years, until one day it wasn't.
I got attacked by fraudsters that used the site to test the validity of hundreds of cards in rapid succession before I even knew what had hit me. It caused thousands of dollars in losses due chargebacks, lost products and unnecessary shipping fees, as the purchases had actually been done with stolen credit card numbers.
From that I've learned to keep an eye on incoming subscriptions and cancel them before attempting a charge if anything seems suspicious.
Blogging openly about progress with Candy Japan has been helpful, as it has often resulted in hundreds of helpful suggestions on sites such as Hacker News and Reddit. It can be quite overwhelming, however, when you have only limited time and energy, but unlimited amounts of suggestions for things that you should be doing.
Having studied Japanese has obviously been helpful, and I still often try to think up new ideas which would combine my programming and Japanese language background.
Timing for the launch was great, as the unique concept garnered a lot of initial interest.
I feel I'm not in the place to be giving out advice, but were I forced at gunpoint, here's what I might say.
Try a lot of things, see what sticks. Don't build out more than necessary, keep it simple. Use existing solutions where possible. Try to promote your ideas early and see what happens. If it looks promising, iterate.
The two books that inspired me the most are Hackers & Painters by Paul Graham and Founders at Work by Jessica Livingston.
Beware of the echo chamber, and try to have your own ideas.
To read more about the business, I have a behind-the-scenes blog on the site. If you have ideas, questions or feedback, I will be monitoring the comments section below.
If you'd like to try some Japanese candy for yourself, you can subscribe to Candy Japan here.
How do you guys handle shipping? I did some research and shipping food to USA seems to be not trivial.
This is the most interesting business model in IndieHackers!
You mentioned that in retrospect, using Google app engine was probably a mistake. Could you expand on that a little bit more and maybe describe some of the reasons for that? I'm working on a project that I put on app engine because of the generous free tier, but I'm seeing that it's quite easy to get "locked in" to Google's infrastructure.