How I got more customers easily by changing where I compete
I used to think the problem was my fitness app.
The videos were decent.
The product was useful.
But every time I posted in the US market, it felt like I was fighting against thousands of other fitness apps, creators, coaches, and ad-heavy competitors.
At some point, I started asking a different question:
What if the product is not the main problem?
What if I’m just fighting in the wrong place?
I noticed that fitness and home workout content seemed to be getting more attention in Japan, so I decided to test the same app in a different market instead of changing the product again.
The idea was simple:
Same app.
Same type of content.
Different market.
I started posting Japan-focused content for the app, with Japanese captions and a local angle.
After one week, the app reached 20 downloads from Japan.
That was the first time I seriously realized that “where you compete” can matter as much as “what you build.”
The tricky part was distribution.
Posting from overseas did not reliably get the content in front of Japanese users. I wanted the test to actually reach Japan, not just be translated into Japanese.
So I used Reach Japan Lab, which helps post content through Japan-local devices and accounts. That made the test much cleaner because the content was actually being distributed from Japan, not just translated for Japan.
My biggest takeaway:
Sometimes the fastest growth experiment is not changing the product, the pricing, or the funnel.
Sometimes it is changing the market.
If you are building a game, language learning app, fitness app, or any product that works well with short-form video, Japan might be worth testing.
There are probably many products that look average in one market but become much more interesting in another.
I’m now trying to think less about “Is my product good enough?” and more about:
“Am I showing this product to the right market?”
Curious if anyone else here has seen a product perform much better after changing the target market.
Smart reframe, and I agree that founders over-iterate on product when the real variable is the arena. One caution I’d add from watching others enter Japan: 20 downloads proves attention arbitrage, not market fit. Less competition on the content side often coexists with harder monetization on the payment side. Japan in particular is forgiving at acquisition and brutal at retention, where shallow localization shows up fast in payment methods, support expectations, and UX conventions that captions alone don’t cover. So the experiment I’d run next is not more downloads but whether those 20 users are still active in week four. If they are, you’ve found a market, not just a quieter feed.
I completely agree with you. I also do not see 20 downloads as proof of product-market fit. It is only an early signal that the content was able to attract attention in that market.
That said, the reason I still think Japan is worth testing seriously is that Japan is not only a large acquisition market, but also one of the strongest monetization markets for apps. It is often considered one of the top three markets globally for app consumer spending, and users in Japan tend to have high spending potential and strong LTV when the product fits their expectations.
So I agree that the next question should not be “Can we get more downloads?” but rather “Do these users stay, engage, and eventually monetize?”
If those 20 users are still active in week 4, then I think the signal becomes much more meaningful. In that case, it would suggest that the Japanese market is not just responding to novelty, but that there may be a real opportunity to build a high-value user base there.
So I think your point is exactly right: downloads are only the first signal. Retention and monetization are what would make the Japan opportunity truly interesting.
The distribution detail is the part I would have underestimated. I’d have assumed Japanese captions were enough, so the point about the content needing to actually originate from Japan rather than just be translated is a good catch. Was there a clear reach difference before and after you started posting locally, or was the 20 downloads already the “posted from Japan” version? Trying to gauge how much of the lift was the market itself versus the distribution method.
The difference in reach was clearly obvious. The audience was almost 100% Japanese. Well, that makes sense, since we were posting from Japan.
There are methods such as VPNs and proxies, but they are troublesome to set up, and the quality varies a lot. On top of that, regulations are becoming stricter, so I think those kinds of methods will gradually disappear.
Really interesting experiment.
Distribution is something people underestimate a lot, the same product can feel completely different just by changing the market.
I'm curious how you validated that it was market fit vs just early algorithm luck.
Interesting. Do you think it was actually the market that changed, or just that the distribution channel happened to fit that market better? curious because I'm starting to think channel and ICP are much more connected than I first assumed.
That is a very interesting point.
I do not think the market itself suddenly changed. My guess is that the demand was already there, but the previous distribution channel was not reaching the right ICP effectively.
When we posted from Japan, the audience became almost 100% Japanese, so the difference in reach was very clear. That made me think that channel and ICP are much more closely connected than I originally assumed.
It is not only about whether there is demand in a market. It is also about whether the content is being distributed through a channel that naturally reaches the right people in that market.
This is a useful reminder that distribution is often more important than iteration. The interesting part isn’t just localization—it’s that “same product + different market constraints” can completely change perceived value. The challenge is turning that insight into a repeatable system rather than a one-off growth hack.
This comment was deleted 6 hours ago.