Candy Japan 2014 Year in Review

Subscriber count triples


How it started

In 2011 I had recently moved to Japan after my wife landed a job here in Tokushima. Despite having studied the language and working as a programmer before, wasn't too keen on becoming a salaryman for a Japanese software company. Instead of becoming an employee, I decided to see if there might be something I could sell online instead. I considered a few things I might sell online (ramen? tea?), but settled on sweets quite fast as they seemed like a good balance between fun to subscribe to and easy to send. It was also a good fit for a recurring subscription. I liked the idea of having a subscriber base that would be somewhat stable. Candy Japan was born with the idea of introducing Japanese sweets to people around the world by sending samples via physical mail twice a month.

I managed to get a handful of people to sign up to validate the idea even before having a website. After being encouraged by that, I set up a site and did a Show HN post. Some blogs picked it up and quite rapidly I had 300 paying subscribers. As the newness of the idea faded it became a bit more difficult to find new members, but I managed to keep the number hovering around 300 for the next three years. That was just enough success to keep me going. 

2014

Instead of trying to aggressively expand or start a new project, I decided to go back to school to finally finish my M.Sc. degree (just got it for Christmas). While my university is in Finland, what I was mostly lacking was a thesis and a few remaining courses. My professor was kind enough to allow me to complete those remotely. At this point the site was running well enough with many tasks delegated, allowing for the extra time to do this.

To my surprise, while focusing more on writing a thesis and completing some courses rather than promoting Candy Japan, it actually started growing like mad. During the year subscribers tripled from hovering around 300 to over 900. There were days we sent over 1000 pieces of mail. 

One thing that did improve with the service was occasionally sending some larger boxes and commissioning a professional photo, which was easier to do than I had imagined. Walk to photo studio with box in hand and tell them to take an attractive picture of it, for about $50.

The pricing and landing page have remained almost the same as before. We kept fulfilling orders that came in and resolving any customer issues, but mostly I kept focusing on schoolwork.

What worked

Looking back the success seems attributable to many small trickles of traffic resulting in a bigger stream. It always seems challenging to know exactly WHY someone subscribed, as it is more likely a combination of things. Anyway, here is what I notice from looking at analytics:

- Someone wrote a BuzzFeed post that kept slowly sending a new trickle of traffic, which over the whole year resulted in 49 new subscriptions.

 - Procrastinating on Reddit! I noticed an askreddit post titled "What is the most ridiculous thing available on the internet for $25 shipped?". As my subscription happened to cost exactly that, I replied and that led to 28 new subscriptions. This felt like a rush, as I was just wasting time on Reddit and suddenly had 2000 visits that day from just a comment. 

- Blogging about the service here really helped.

 - Many subscription box review sites (that's a thing now) mentioned the site and sent subscribers. I did spend a ton of time sending review boxes to the bloggers, 230 of them in total. Most of those resulted in nothing. I suspect many of these conversions might be people that were already on the site, but went searching online to find reviews before signing up. Still it seems probably worthwhile overall.

 - Facebook sent 47 subs, Twitter 11. I'm not sure if these are from organic sharing or my own posts / paid ads. Google Analytics just shows all of these coming from root path of FB and Twitter just reports them as coming from "t.co".

 - All of the above made Google like the site more, resulting in twice as much organic search traffic. That was very significant, as search results in hundreds of subs. Or maybe there is just more search traffic for subscription boxes overall? 

 - My old AdWords campaign suddenly started working in July, again just a trickle but over the year 34 subscribers on top of everything else. Before that it was essentially dead, but without tweaking anything it started working. Again could be that I was just capturing more people who were already searching for the site.

 - Sites like msn.com, lifehacker and huffington post did some "top 10 subscription sites" type listings and mentioned the site with no prompting from me. While those are famous sites, these were very buried links, but still resulted in total 23 new subs.

What didn't work

In 2013 I had grand plans on featuring different Japanese prefectures, which still sounds like a fun idea, but didn't get around to doing. I also had plans to get the first envelope to subscribers sooner, which was a partial success in that I have a system of doing that now, but still need to reorganize things to be able to ship every day instead of twice a month as we are doing now. I did manage to buy some items directly from a manufacturer, but didn't keep doing it despite the better margin, as it was more helpful to have a middleman help find items for me.

Next

In 2015 I plan to simply keep iterating to improve the service. Outsourcing handling of returns (packages that come back because subscriber moved to new address etc.) is my January focus. I will also look again into offering another subscription tier and see if it might be possible to add a shop for individual purchases. Mostly keep doing what works. 

If you have ideas, questions or feedback, feel free to contact me@bemmu.com.