YouTube marketing horror story
candyjapan.com
candyjapan.com
Less than 3,000 views across tiny channels with seemingly random creators, topics and locations? This is a waste of time and money.
Figure out who the buyer is first. Clearly there must be a certain disposable income level. What region buys the most? The USA? Europe? Asia? What are your customers interested in? Is it about Japan, or is it candy? Is it novelty gifts? Is it seasonal? Are there tie-ins to anime or music or some other Japanese export or culture? How big can the market even get and how do you know if you've hit saturation? These things need to be answered so you have a grid of interests and can focus on the most relevant audience.
Once you have that, only then can you actually come up with a plan to best reach them. Don't just send things out without a target audience. The advertising channel you pick (whether Youtube or not) is a function of your message, and almost every "viral" hit is either a game of chance (which you will lose) or much more orchestrated than you know. 99% of influencer marketing is useless and you do not have anywhere near the scale to break through it by force.
At the very least, try and talk to some (consumer product) marketers, there's plenty of advice you can get by just asking nicely, or even pay for a small strategy session. That will have far better ROI than this.
What about trying to get to the top YouTube SEO for search terms, "candy [...] Japan”, since his company is literally called that. If this feasible for a small budget?
I'm curious what it would take to pull off. Great insight, thanks for sharing.
Instagram is still wrongly focusing on the channel instead of audience. For example, if the buyers are younger people into fashion and food, and there are relevant profiles that showcase both, perhaps while traveling to Japan and showing how you can get the same candy at home, that might be effective. But you need the data to show that, not just assumptions.
Search is always a great way to advertise because you can target directly to people's intent as they type it in. Buying a few keywords around candy, japan, snacks, traveling, etc. is a good and affordable way to start. $5K spend on some ad strategies would provide decent data to start testing the market.
I only specialize in Amazon marketing but if my process is applicable to Adwords/Facebook, the startup method would be small tests with tiny budgets. If the article writer had made a video ad or just used one of these unboxing videos and tested it on Facebook, he would have gotten immediate actionable data.
He should have a Facebook pixel installed on his website and should have collected data on traffic as well as information on those who converted. Turn the data that you have to work for you. In Facebook ads, you can create custom audiences based on value. In marketing, there are ideas of superconsumers or customer segmentation. Who are your best customers? Why? You don't have to reinvent the wheel if you already have data. Just look at the data that you have.
You can put the data you have to work on Facebook. If you don't understand your customer market but have existing customers, lucky for you, Facebook can tell you and help you find more of them via lookalike audiences.
If you don't have any existing data and you are starting from scratch, then you can create audiences on Facebook ads and test them. Or better, ID your competitors and analyze or target their customers. I'm not a pro at Fb ads so this might be outdated but the idea I'm familiar with is testing $5/day ads. $5 day ad will give you a lot of data on impressions, CTRs, CPC or impression, conversion, engagement, etc. So you can easily test multiple audiences or ads and get data and numbers. Once you crack the magical formula where your ad is successful and all the numbers add up (customer acquisition cost), you light the ad on fire by boosting the budget and scaling it. The explosive growth part. That's what I do for my Amazon clients. I ID the best strategy to success, find the magic formula by making sure all the numbers add up to their goal, and then when I find it, I pour gasoline on it and dance around the flames as the money rolls in.
That's the basic idea for Amazon ads and maybe Facebook ads. But I know Facebook ads are amazing because there are so many tips, tricks, tools, strategies you can implement. For example starting with a video, getting engagement first, then creating a new audience based on those who watched the video the longest, etc. I have also heard that Google Adwords has way more bells and whistles than Amazon ads too.
Other people mentioned some of his mistakes, another mistake he made was not having a brief and agreement. Influencer marketing is a lot more complicated than here's my product, please make a post or video about it. Most people include a brief for what exactly needs to be in their post or video. A special format with guidelines on style. Key points that need to be made, time line, etc.
The article was interesting nonethelss: the writer was transparent and laying down his assumptions and how they carried out the work AND the process. Executing a thing like this might be trivial but it requires a lot of annoying work, I know this by experience: even for simply sending a t-shirt to a customer (which is a simpler item is a pain).
Nowadays luckily you can send t-shirts or other apparel with a single click directly from the company that produces the item (there are a couple of names out there who will do that). BUT if you've got custom items or boxes of items (like in this scenario) you have to think about packaging, costs and the "unboxing experience", that last one is critical.
Unboxing experience can make or break the whole video.
I feel bad for the author who did so much work (180 attempts!?!) to get 17 videos and 0 sales and I suggest next time he goes to these specialists if he's really committed in using YT as a marketing chanenl.
The term “influencer” is such an eye roll to me.
They aren't celebrities. They aren't just people with a following or a popular channel. They are specifically people with a following that are successful at converting sales.
Unless you have something interesting to say on the matter, I don't think this forum needs more "le wrong generation"[0] posts.
You do not just become an "influencer", instead you gain influence by way of gaining a following for doing (and being good at) something. The term is definitely overloaded and mostly pointless now.
That’s just a word put on something which has always existed (and will always exist).
Any idea what are some of the criteria they use to find the perfect youtuber? I'd be surprised if this could be done reliably for many products, let alone for it to be worth it over common-sense random sampling when including the additional cost of their services.
Basically to me this smells like a lot of manual work by a team of humans, reading through countless of YT comments and reddit posts to understand exactly how a target audience reacts and "thinks" to various subjects.
But I could be wrong because they didn't say anything too specific but just "data analysis" which could be anything?!
The example with h3h3 they described as a successful campaign, was a vibrating toothbrush - electric and all that jazz, after being featured in the video "en passant", they drove so much traffic to the sponsor's webpage that their ecommerce timedout/crashed.
So most of the analysis is done by Youtube but with some snapshotting you can start to find trends, subsegments, seasonality ...
The one I used, you put up an ad of your product, and influencers applied, so you could quickly vet them by easily going through the data in one place.
What I’d do instead of unboxing videos is to have the channels call you out as a sponsor with a coupon in a number of their videos. And hopefully you can hit “meme level” like freshbooks, tunnel bear, etc.
The other thing is cosplayers seems like a weird target audience. I’d be focusing on gamers, they seem to like that blind surprise box stuff.
Previous: https://news.ycombinator.com/item?id=16461187
Relevant comment from that discussion:
"You targeted hardcore gamers, but you're not selling games. That's almost a complete non-sequitur. Your goal was to target people likely to purchase candy, or perhaps Japanese novelty items, via a mail subscription service. It doesn't appear that you made much effort if any to figure out what that audience looks like and how to reach it."
It doesn't seem like much has changed, he's just hellbent on putting effort into advertising campaigns without first figuring out who the actual target audience is that might reasonably complete a purchase, and what the product's best sales message is.
I mean, have you considered the possibility that people who make cosplay videos and the people who watch them mostly care about wearing costumes as their hobby, and that you might try targeting people who make videos about either various types of candy or various types of Japanese products because that's actually what your business sells?
Anyway, I was expecting murder when I went to the Candy Japan homepage, but there are changes they can make right now that would help. White text on light-colored header, doesn't use professional typefaces, as great as the animated candy SVG is it takes too long to see a real piece of candy, body text is too small and in the background (just unchecking the 80% rule improves the readability by a lot), testimonials distracting the body and flow, the list of past boxes is a list of blog posts that I have to individually click on to see the contents of (which is a shame since there is associated prose that I would love to be able to skim through quicker).
Just some of the things I would change. TokyoTreat seems to be part of a larger network selling Japanese loot boxes so it's seemingly stiff competition and maybe that's the key here. Perhaps CJ can pivot itself as something more personable and intimate? You can even make your own content (vlogs, perhaps) on how you go about hand picking a selection. You also already have a cute mascot.
I figured out it was much easier to use young affluent females influencers from ages 18-25 who had 10000-15000 followers (not too much, as that dilutes the network and those who had too many followers were too expensive anyway) as they're usually super-connected in their network, "popular" and their followers (usually their friends, etc.) trust their opinions.
Interestingly, I didn't need to pay anything even though I didn't know these people, most were happy to just do it.
With Youtube, it would seem that only the big-name You-tubers are worth any kind of sponsorship. I find that influencer marketing is really about trust and credibility as your customer sales really come from trusting the influencer promoting the product and in turn also meeting that need the customer wanted too.
> I didn't need to pay anything even though I didn't know these people, most were happy to just do it.
Maybe they liked the video and/or the app itself, and that's why they wanted to show it to their followers? So I'm a bit curious about the video, maybe one can learn something from watching it (?).
This was the app, I'm in Canada: https://itunes.apple.com/ca/app/lucid-see-who-likes-you/id12...
I contacted them through Instagram and explained my app and asked if they wanted to post the video for me. Some said yes. Some didn't. I think I also framed it as marketing experience for them too but I'm not too sure what I said exactly, I didn't really have any money at the time. Also some were friends of friends of friends and it wasn't as strange in asking them to post but I still didn't know these people.
I'm not as focussed on the app anymore and sometimes maintain it, I keep it more as a side project link for my resume.
Overall, you just need to be brave enough to message them, I was a bit shy and unsure at first and then afterwards I got used to it because it's easier to communicate online and it was the cheapest alternative I could think of.
I'm sure you'll find that almost every one of the YouTubers chosen for this experiment are churning out video after video of candy box samples they've received to an audience of nobody.
I think this part is weird, since the main part of his business is shipping packages internationally. Does the list of youtubers include too many from unusual countries?
"Item out for physical delivery" and then... nothing. Usually it should show "delivered" at the end.
Some countries were unusual such as Bahrain, but then the same thing would sometimes happen even for Canada or Norway.
a) During this experiment with youtubers I found that international package tracking often (24% of the time) does not actually work.
b) During my last few years running CandyJapan I found that international package tracking often (24% of the time) does not actually work.
[I interpreted a). Now I think you mean b).]
In the case above I would expect they they just didn’t scan it when delivering. We get cases with USPS, usually with appartment buildings that they will leave packages in common spaces and they will be stolen. Although that would usually have a delivery scan.
> August 20, 2018, 10:44 pm > Awaiting Delivery Scan > The delivery status of your item has not been updated as of August 20, 2018, 10:44 pm. We apologize that it may arrive later than expected.
Have you tried roasted a chicken? A whole chicken costs $6. Feeds 2 with leftovers, the bones can be used for broth, which can make lots of things, especially soup. The leftover meat is great for many things, like fried with potatoes. Sometimes I find that the chicken was used for 10 meals. Of course other things are added along the way, but veggies are dirt cheap.
That means you blew your 3-day food budget on half a chicken. That's just not nearly enough calories for 3 days. I'm not doubting you're able to afford chicken on that tight budget, it just seems that you're using it more like a condiment, rather than food.
I think many people take for granted that we can live so well here, and buy whatever looks good, and not planning ahead of time. But honestly it's fun to make meals yourself while saving money, and you can avoid some of the weird stuff they put in a lot of preprocessed foods. I admit we are trying to live frugally, we're trying to kill a house mortgage.
That's the nice thing about global markets: OP just needs a thousand customers a month = $30k/month gross. Maybe half that goes to shipping. If you can get each of those products for $1-$2 each in bulk, you're doing okay.
Some product might be free, highly-subsidized or he gets paid because the manufacturer wants to build an overseas market.
IMO if that’s what you’re spending on food, most sane advertisers would exclude you as well.
The problem being that the people who made it their profession, really don't like that idea, and thus it is not in their interest to see this fact.
The thing is that often it does work, and this business seems the type that probably has a least a few moves they could make with positive ROI.
What’s driving everyone nuts is he’s doing all this work and analysis but seems unwilling to stop making the most obvious basic rookie mistakes. Repeatedly.
What's more likely: That this company with a tiny budget had no idea what it was doing, or that a 12-figure industry with 2 of the most valuable companies on the planet is all fake?
Advertising is a core driver of every successful company and there are 100s of petabytes of data generated everyday proving it works. Even without any fancy analytics, measuring that every dollar spent returns more than a dollar in sales is about as simple as it gets.
This company's problem is that they need to learn how to do it right.
The same argument can be made for any non-journalist source.
> To get someone to agree to an unboxing, the most important thing turned out to be not asking for their shipping address up front. Instead first explain the service briefly, then ask if they are interested to do receive a sample for a video. If they say yes, only then ask for the shipping address. Asking for the address in the initial mail seemed to be off-putting and resulted in less people agreeing.
Even more details https://www.candyjapan.com/behind-the-scenes/ab-testing-cont...
Having $1m in rev a $1k marketing lesson isn't going to break the bank. https://news.ycombinator.com/item?id=15135372
After discounting cost of goods, operations, and taxes, $1k is not something to just write off so easily. There should've been more planning, although parent comment isn't quite the right idea.
In particular, this competitor video was just a lot more fun to watch: it's less about the product, or specifics of the subscription, and more just enjoying the two friends joking around with each other. And that helps sells the experience of the product: Buy this box, invite a friend over for the afternoon, and laugh at each other's awkward faces trying sour candies until the dog starts coughing and freaks you both out:
https://www.youtube.com/watch?v=3rwmezY0YZ4&t=7m44s
Or just have a fun afternoon trying to get the damn bottle open:
https://www.youtube.com/watch?v=3rwmezY0YZ4&t=9m32s
Of course, as much as that video had me laughing out loud & watching to the end, even that particular competitor video only has 250 views. And maybe you don't want videos like that, 'cos those girls really hated half of the candy the competitor sent to them.
The Candy Japan videos mostly had lower production values than the competitors ones. The Starbit one looked interesting, but their audio was so quiet & echoey that I couldn't bring myself to watch it. I did watch all of the "Sailor Moon & Kids" one (https://www.youtube.com/watch?v=aHF1p3ggq_w) and it was okay, but it kinda looked like the kids weren't enjoying the candy by the end.
So maybe the channels selected need to be filtered down further based on quality of production & whether their videos in general are fun enough to be shareable. Or maybe look at this as having bought $1000 of video recorded customer experiences & feedback.
Then I saw $29/month. I make a decent income, but that's just too expensive. I suspect your problem is less about advertising and more about value.
1. he only offers one pricing option.
2. go to amazon and search for "japanese candy". You can a lot more for your money without subscribing.
https://www.amazon.com/Japanese-assortment-Dagashi-Umaibo-ch...
3. after receiving a few boxes, you will find things you like. How would I re-order just those items? Amazon.
The problem: when a charge to my card was denied, they just canceled my subscription without notification. A simple (automated) email letting me know there was a problem would have had me fixing things within a day and they'd have kept a loyal customer. I did't sign back up because, IIRC, we'd gotten two shipments in a row that weren't very exciting and I just wasn't motivated to set it up again.
I feel like there's automation to be had, but they're having none of it.
You'd probably be better off asking them to give it screen-time in the same way that podcast advertising currently works, where the host devotes 20 or 30 seconds at the beginning or the end of the show to talk about Sponsor X.
What does a devoted subscriber base have to do with purchasing advertised products?
I also genuinely don't understand why you are spending so much on shipping. I have to assume that if you offered an option of monthly or every other month and spent half of the money saved on shipping on more candy that your customers would prefer it.
It's just a run of the mill monthly box that gives people cavities. Next.
Before I continue, its easy to stand here and say "You are doing it wrong". Please dont confuse this with me being such an asshole.
My goal here is to refine the idea, purpose, and outcome.
Idea
Having people dressed in cosplay review the candy. Possibly an attractive female.
Purpose
Get people to pay attention and buy some candy.
Outcome
Pay a actor/cosplayer to record videos for your own channel and promote that on social. Contact other youtubers and have them run your content on their channel. A cross channel feature, if you will.
Reason is that you control everything. There is little to be lost here, because your product wont need to be shipped, and the only thing exchanging hands is money. You also control the narrative and the direction of the leads to whatever collection method you decide to use.
Lets compare to the approach taken by OP.
Idea
Find a bunch of cosplayer channels. Send them free candy. Get a review.
Purpose
Get subscribers to click through and make some sales.
Outcome
Not all candy got to the channels, reviews were controlled by the channel owner (big no no unless it is a paid and scripted promotion), and no sales because you weren't in control of the narrative.
This is more common than youd think because people often think that youtube is a QVC like experience. People watch and buy. Happens often, but its not the main driver. Youtube's experience is in the "AS SEEN ON TV" product types. It is a way to get your offering out there on a platform, get some recognition and then market with the fact that you have been featured on the platform. This is why scripted and paid promotions work. You can say "AS SEEN ON $SOMEONE'S CHANNEL".
Where to go from here? Where would I go from here?
You already invested $1000+ of marketing and got a front page post on HN. Not a bad deal, but this has no long term value.
Id remove the line that says "Contact me to review my product". The outcome will be similar.
Then, I'd start looking for writers to produce me some good video scripts (less than a $100). Make sure the script is about the candy. Not the cosplayer. The cosplayer is a way to differentiate your content. It is not the focal point. Have your writers produce content about the actual candy. Think of Lucky Strike "Its toasted". Tell me how Pocky are made and how delicious they are. Come up with some ways to rate candy and rate them. So many ideas...
Id look for local cosplayers and record one of them following the scripts. Better yet, hire an actor and have them dress up. Same outcome.
Next, Id share the initial videos to my current client base. Make sure to let them know you will be reviewing their favorite candy on the videos. Share the story of the candy.
Publish it on your own channels. And pay someone to promote it on aggregators.
From there, reach out to influencers and ask how much do they charge by 30 seconds of promotion (important). Most will undercut their own rates by diving it in 30 second intervals. It also opens up Instagram videos. Then have them run your content as a feature. They love this kind of stiff because all they have to do is click publish. Easy money.
Main point is that marketing is not something you let others do for you. You can let them take control. You meed to live and breathe your own product. Not the boxes. Or the subscription. But the candy. Your goal is to have people feel excited about the candy offered in candyjapan. It will eventually force you to open an online retail store.
As always, this is not an attack. It is an outsiders perspective. From someone who has been, is, and will continue to be in similar trenches. Feel free to reach out privately. Best of luck.
PS. You are not in the subscription business. You are in the pleasure business. Candy is a luxury and a pleasure.
I have no idea why you're using this spiteful tone for someone openly sharing his experience as a YouTube marketing newbie. You also seem to be imputing some views on the author ("assuming that basic economies of scale aren't relevant") that he has not publicly espoused.
Anyway, looking forward for the first box!
Have a nice day :)
PS: Someone else here mentioned if part of the stuff could be something beside candy, dunno, could be interesting too.
They have hundreds of subscribers, and had many more in the past. https://www.candyjapan.com/behind-the-scenes/2017-year-in-re... Apparently there is competition in the space. https://www.candyjapan.com/behind-the-scenes/nobodys-going-t...