AnyPerk, the first Japanese Y Combinator (YC) startup, pulled off an audacious pivot in the middle of their three-month funding cycle in San Francisco. They’re living proof that pivots needn’t always be delicate. Here is AnyPerk Founder Taro Fukuyama’s advice of when to uproot your product and start over.
Fukuyama’s story is a counterpoint to that of Mailbox app co-creator Gentry Underwood, who gave us contrasting advice in this piece: pivot your product, not your problem. Both are right—you’ll have to read on to figure out which approach is right for you.
What happened that caused your pivot?
Everyone in our class changed their product, but not the problem they were solving. So I think we're the biggest change—we changed totally. We got into Y Combinator for a dating site with mutual interaction—you’d sign up with Facebook and see friends of friends, and be able to ask for introductions. By the third day of Y Combinator we knew it wasn't working.
Does your new idea share anything with the old one?
Is there anything in common with these ideas? To be honest, no. I think it's really important that if you want to pivot, you need to pivot hard. If you make an adjustment that's really small, it might not make that much difference.
Some people say you should pivot your product, but not the problem you’re solving—because without a die-hard commitment to a problem, it’s just too hard to survive the startup process.
I think both are right. If the problem that you're solving is really big, then what you said is right. But we were solving a problem that wasn't really a problem, so we needed to change totally. If you're solving a really big problem that's actually a big problem, then you can just pivot the product and stick with the problem.
What was the "oh shit!" moment?
Well, if you start a company you feel like you're really changing the world. And when you talk to investors you're convincing yourself as well as them—you're drunk with the dream that you're trying to pursue. So even if it's not growing, you feel like it's going to grow eventually for some reason. But you also know that it's not going to happen forever. If it's not growing now, it may not grow even if you add a million features. But we kind of didn't want to be the first ones to say it wasn't working—we couldn't change it before we got into Y Combinator. But once you are accepted, the partners ask tons of questions: What are the numbers? And you have to admit, oh my god, this isn't really working. We had to be honest: We really should change it.
How did you feel at that point?
We were worried—demo day was coming three months later and we had to show hockey-stick growth. And we were excited to join Y Combinator because we were the first and only Japanese founders, so we were so excited as well, but nothing was working for us. I was scared we'd have nothing to show on demo day. In the first month, we ended up changing ideas seven times. That was a nightmare.
What did your YC advisors say?
How we made sure our idea was working: YC gave us advice: Sell before you build it. You don't need to make a product to make sure that this is the problem you should be solving. So instead of building a prototype, we made a mockup and asked people: Would you pay money for this? So other ideas people said, I won't pay for that. But for this idea it was different.
How did you think of a new idea?
I knew there was a Y Combinator discount list. If you're in YC you can get discounts from AWS or Gold's Gym. We showed the list to someone outside of Y Combinator and asked them, "Would you pay money to have access to all these discounts?" And people said they would be happy to pay for it. Then we remembered that in Japan this is a really popular business model and there are four public companies [doing this] there. So we pulled the model that was popular in Japan. A lot of companies have tried this business model in the United States as well, but unlike the Japanese companies, they really focussed on local—targeting restaurants to get a discount. But it's really hard to get discounts [from local merchants] because they have thin margins, and you can only provide the product to customers in that city, so it isn't scalable. There are no public companies like [AnyPerk] in the United States so far, so we decided focus on national perks.
How did you get your first customers so fast?
We had a chicken and egg problem: If you don't have any customers, you can't get any perks. But without perks, you can't get any customers. The other YC companies really helped us by becoming our customers even when we didn't have that many perks. Then we contacted all the vendors that participated in the YC discount list and said: We have all the YC companies signed up, and we’re making something bigger than this list, so you should give us other discounts. Why would they say no? They know it's expanding beyond the YC list, and that they can have access to other [startups] as well. That's why they gave us tons of discounts. We faked it. Usually companies copy from Silicon Valley to Asia—but we are the one copycat that took an idea from Japan and brought it to Silicon Valley.
[Image: Flickr user Lali Masriera]