Hey Kent Kajitani.. thankx for doing this AMA
So, how long did it take Vasily to grow from 100,000 users to 3 million?
I ask this because I have seen startups take years to get to millions of downloads; so, is it possible to replicate that within a few months. If yes, what techniques would you use?
Actually we spent almost 2 years because we didn’t have any strong competitor so no need to hurry for acquisition.
We didn’t spend any money for ad until we got 1M users.
I think the period depends on ARPU.
cus if your product has high ARPU, you can boost Advertising.
So my point is improve retention rate first, then tackle on monetization and improve ARPU, then boost Ad so that you can acquire millions of users rapidly.
Unfortunately, our business model took time to validate monetization model so took time to reach 3M users.
What were the most effective channels for acquiring your users? If could get specific and say why as well.
It depends on stage.
After you build solid monetization model and get high ARPU, it would be definitely online marketing channel, for us it was facebook and twitter.
Before building monetization model, you should use your imagination like we did in the first stage.
(please take a look at another question)
Hi Kent, what are the major differences you have observed in Indian user acquisition and Japanese user acquisition?
Surprisingly, most things are similar.
But I found some difference.
Most Japanese startups focuses on SEO even if their products are smartphone applications.
I think it’s because we have only one language.haha
2) Indian startups use much much more transportation advertising than japanese one.
That’s because most japanese commute on train and their eyes are occupid by their smartphones while many indians drive their cars and bikes and they have to look around.
same situation with U.S.
Kent Kajitani I still wonder why Japanese creatives have too much of textual content than graphics? anything to do from culture.
I guess there are two major reasons.
1) Linguistic Differences
Logographic-based languages can contain a lot of meaning in just few characters. While these characters can look cluttered and confusing to the western eye, they actually allow Japanese speakers to become comfortable with processing a lot of information in short period of time.
People require a high degree of assurance, by means of lengthy descriptions and technical specifications, before making a purchasing decision.
It depends which stage you are in.
I highly recommend you to read Lean Analytics.
It give you the “KPI cheat sheet” !
Do you have a checklist that you verify for every release?
I don’t have cus KPI differs on each new features.
But I made it rule to ask female employee about the new feature BEFORE launching it.
And after launching it, I checked customer feedbacks and reviews more carefully than usual.
I know this is very hard but one strategy is building a partnership with internet provider company.
In Japan’s case, NTT and Yahoo Japan played a major role to increase the penetration rate.
And the startups which had a relationship with the company grew dramatically as the internet itself grew.
So the situation is the same, it’s extremely difficult but there is a way and it’ll bring huge benefit to your startup.
I made a simple formula.
User Onbording = AHA moment = Value Proposition + Tutorial
In order to make user onbord, you should make AHA moment happend.
AHA moment is a moment when a user gets what your product is really about and why it’s valuable to them.
To make aha moment happend, test what to tell users as Value Proposition and how to tell
There are many ways to tell it, like app store screen shot, walk through, movie etc.
Also you should make sure that new user understand how to use your product.
so you should test what to explain and how to explain.
Best way for this is User Testing.
Go to cafe, and talk to someone, let him use your product and see if he can explain what this service is and how to use it!
I know this is not a brand‐new hack but it’s personalizing the contents for each users.
Because fashion tastes varies so much based on age, height, budget, favorite brands, favorite style, favorite magazine and so on.
So we first made an experiment by personalizing by hand, I mean without algorithm with 10% of users.
The experiment showed reasonable increase in retention rate so we build an algorithm.
After optimizing the algorithm, we succeeded to increase retention rate by almost 50%, I mean 1.5x.