Candy Japan crosses $10000 MRR

It's been a while since I blogged about Candy Japan ( If you've been actively reading Hacker News for some time, you might have read my previous posts about it. If not, essentially Candy Japan is a "candy of the month" subscription service where I send subscribers random surprise candies twice each month in exchange for a monthly payment of $25. Initially I promised to send an envelope every two weeks containing 1-3 items, but eventually I have started to send larger boxes as well, since not all types of items can fit in a slim envelope.

It started from a simple HN post and a comment thread. It grew thanks to several blog mentions to about 300 subscribers. For a long time it seemed the subscriber numbers wouldn't grow past that. But now I'm happy to report that ceiling has been shattered, and recurring monthly revenue has now crossed $10k and customers are reporting that they are happy too. Here is a picture of the subscriber growth:

Initially I worried about issues with customs, but based on surveys customers are reporting that they are receiving their packages without customs issues. Well, we did have one package where the customs officials opened it and checked whether a pouch of powder was actually candy or something more suspicious, but that is the only reported case out of thousands of packages sent and did not result in any trouble for us or the recipient. Overall everything is working very well. Sometimes packages do get returned to me (usually due to wrong address or customer not being home to accept a delivery) but I always take the time to solve these issues with the customers and resend packages.

Out of that $10k / month revenue of course not all (not even most) is profit. Compared to a SaaS, the costs are very high, but I've carved enough profit margin to reach an income level similar to a Japanese salaryman now. The major costs are shipping (we use Japan Post) and the products themselves. I'm now spending enough on shipping that people at the post office know me by name. 

No special discounts yet, but it seems likely I can start enjoying a small discount in the future if the subscriber numbers continue to grow just a bit more. If we send more than 1000 items in a single day then there is a 10% discount. Last time we sent 700 in a single day, so just a bit more growth. Strangely it only matters that I send 1000 items at once. They don't all have to be candy shipments. Actually I've calculated that I might start saving money soon just by shipping some empty envelopes if I really wanted to cross that 1000 limit fast, since the discount is just based on a threshold ;)

Lately a small part of those shipments is going for free to blog writers, YouTube video artists and other creative people who have showed an interest in possibly featuring Candy Japan. To help them write their posts or create their videos, I send some samples out. Initially I hesitated to offer the free samples as it quite rarely results in any noticeable traffic at least in Google Analytics, but now I use these review samples to pad out the shipments. The thing is that candy manufacturers don't want to sell you incomplete boxes. You must buy whole boxes, but if a box has 200 pieces of candy in each box but you have say 550 subscribers then you must do something with those 50 extra pieces you would otherwise waste. So whenever I have a situation like that, I send those extra pieces to reviewers. I don't go out looking for reviewers, they seek me out. I get email almost daily from blogs that want to review the shipments, mostly from the community of "mommy bloggers" (that's a thing it seems) and others who are writing product reviews as sort of a hobby. Whether this is the best use for those extra pieces of candy I am not 100% sure, but it does result in some conversions.

Apart from shipping the other major cost is the candies themselves. I haven't been focusing on improving the margins there as much as I could. Instead I chose the path of convenience, where I have an arrangement with a local supermarket. I give them suggestions on which candies I think foreigners might be interested in and they then contact the suppliers to see which ones they are able to get. That saves me a lot of time and improves the chances of actually getting the items I wanted, but results in less of a discount than if I purchased the candies directly. 

The suppliers send the items directly to the supermarket I am cooperating with, so there is no longer any need for me to physically receive boxes of candy. Which is great, since we live in the third floor with no elevator and it was getting a bit ridiculous to do all the shipping ourselves in the beginning :-) The relationship with the supermarket built gradually. Since this has been ongoing since 2011, I went from being a strange foreigner bothering the boss with requests for a few dozen extra packs of candy to being the "Bemmu-kun" who casually walks to the back room while employees are slurping noodles in their breaktime. Having orders made for me is easy, but I am throwing away some margin there.

I did make a small breakthrough recently by making the very first purchase directly from a manufacturer. It happened thanks to a fortunate connection I made through Hacker News. Through HN I got introduced to a coworking space in Osaka called Knowledge Salon (thanks @yuzool) and met someone there who is experienced in dealing with Japanese manufacturers. He helped make the initial phone call and thanks to his introduction we made the first order, which was for 550 pieces of a larger candy variety box, which was apparently a large enough order that they felt it was worth their time. 

After having the initial phone call made by a fluent Japanese person (and seeing money really getting transferred to their account), they were also willing to deal with me directly. Obviously ordering from them was much cheaper than buying from the supermarket, which enabled me to send subscribers a larger shipment than usual that time. Sadly that manufacturer only makes a certain type of candy (ramune) and their selection is too limited to make many orders from there in the future. But it was encouraging to see that direct buying is possible. I also learned that phone and FAX were still preferred over email.

Going back to our arrangement with the supermarket that I usually use to place orders, after we decide what to send and the supermarket has received the items, I prepare a shipping list for them and the supermarket employees help do the shipping. I have a Python script running in Google App Engine that gathers all the subscribers that are supposed to get a shipment and a PDF file is generated from those (complete with customs forms) that the supermarket can then print and attach to the packages. We used to write each form by hand! Ah, so glad that is now automated.

Our relationship still doesn't really feel much like "drop shipping", as I am physically meeting with the people there several times each month (and buying my groceries while there!). There are many reasons to meet such as handing over my latest candy discovery for them to check with the suppliers, or stacks of cash to pay for a previous shipments or demonstrating how I would like them to pack some special shipment.

What am I doing now then that most laborous things have been automated? My task is now mainly curation; coming up with a mix of interesting tastes, striking a balance between adventurous and safe choices. To up the element of surprise I sometimes burn some extra money on larger than promised shipments and include DIY-candies and other specialities. Once I even commissioned custom chopsticks to be made with the name of each subscriber carved on them. Everyone loved them, but I ate up my profit that month with that extra gift. It's difficult to resist spoiling my subscribers and hard to remember to keep some profit too sometimes.

Besides curation other tasks that remain for me are responding to customer requests, dealing with bounced packages (because customer moved / entered their address incorrectly / post office made a mistake etc. it happens), content writing, marketing and site improvements. 

Content writing because  subscribers might not know what they are eating as all the candy labels are in Japanese, I send a twice-monthly newsletter which describes all the latest sweets. It takes me a day or more each month just to do the research and write the content for that newsletter. Next month I will get some help with this from another foreigner living in Japan who has some experience in blog writing and unlike me is a native English speaker (I'm Finnish), so perhaps he will be able to write some of the candy descriptions in the future. 

A big pain point I had for a long time was dealing with customer support. My personal inbox was getting clobbered by tons of Candy Japan -related support mails (my address has changed, I forgot to update my address and package was sent to where I used to live, my credit card expired etc.). I began to fall behind in support requests and as my inbox kept getting bigger it made me reluctant to check my email at all, resulting in even more email piling up. The whole thing felt very unorganized and I realized I need to take control of the situation, as it was starting to have an impact on my overall happiness. My solution was threefold:

1) Switch to a support ticket solution. While still the overall work is the same, now I have a more clear way to mark support requests as having been dealt with, better separation of my personal email from support email and some glimmer of possibility that in the future there is a clear path to delegate this task to someone.

2) Start writing a support manual. If there is some issue that keeps happening, I have started writing those down with clear steps on what to do in those cases. Support is actually easy to do when the response is already known beforehand. 

3) When getting an email, spending some time thinking WHY I received that email and how I could PREVENT similar emails from being sent to me in the future. I realized that many of the emails I get were because customers wanted to change or view their subscription details, so I added a simple page on the site where they can do this without needing to contact me. I will still make any requested changes myself as well, but taking steps like these will keep the workload getting smaller.

After these I am now in a place where my email support workload is gradually getting smaller and easier, instead of in a place where my inbox just gets longer and longer. Now I can deal with support in about 30 minutes each day.

Now that enough things have been automated or delegated to free up some extra time, I am spending that to try new things. The first new thing is a new club called Pen Japan, which is a stationery of the month club. For $14.95 / month I am sending subscribers a monthly shipment of cute / interesting stationery. Think cute stickers, pens, stamps etc. kind of things that a student might want to use at school (