Publishing on Android

My Guiding Principle

2012-02-18T21:18:00.000-08:00

Inventing on Principle, a recent talk by Bret Victor, a prominent designer, talks about finding the guiding principle for your work. Since viewing the talk, I have been thinking about the guiding principle behind Kindergarten Math, and what I want to take from it going forward.

My guiding principle is Learning through Two-Way Interactivity

The first part is obvious. Learning - is about creating apps that people use to learn. If they don't learn, the app is not doing it's job. The second part is about Interactivity. Many apps on the market tell you the concept, or test you for it. For apps that tell you the concept, the interaction is one way from the app to the user. Here the app does all the work. e.g. an app that counts for you. For apps that test you for it, the interaction is the other way - the user does all the work e.g. an app that asks you what 2+2 is. In Kindergarten Math, the app guides you how to get to 2+2, but, it lets you work on it till you find the correct answer. That's two-way interactivity in the same exercise. The learning and two-way interactivity are connected. The Learning happens faster and deeper because of this two-way interactivity.

The two-way interactivity is hard to achieve at present. It requires more coding and effort than 1-way interactivity. That is perhaps the reason why not many apps offer it. But, that also means there is an opportunity to excel here.

Publishing Android Application to (B)lackberry Playbook

2012-02-13T20:47:00.000-08:00

RIM, the owner of Blackberry brand, has a tablet on the market - Playbook. It is originally not an Android tablet. But, with v2.0 of it's operating system, it is integrated with the android runtime. So, it will be running android apps. This is good news for android developers who want to reach a broader audience. However, the existing APK files need to be repackaged for Blackberry to a bar file, and submitted to blackberry specific market.

Why couldn't they just use the standard android packaging, and markets? I assume it's because they want a share of the app sales. The special purpose android tablets are not selling very well, even with high-quality-specs like the Playbook. It is because customers want apps that their friends have. And developers, like me, are too lazy to port their apps, or even repackage them.

Getting desperate for apps for the blackberry playbook, RIM announced they would give a free playbook per app submitted to the blackberry market. That is a pretty good offer for me, given that I need a tablet anyway to test my apps on a larger screen. So, I submitted Kindergarten Math to blackberry vendor portal today.

Creating the Bar file from an existing APK file was fairly simple. There was an online tool to convert the file. It did require Java installation for Applets, and the Android SDK pre-installed, but, I already had those for development purposes. Since we don't use any special permissions, the conversion was a breeze.

Testing the converted app was anything but easy. My husband spent a few hours last night setting up the VM, getting it all configured, installing the app on the VM. Finally, when he tried to run it, it would load the app, but not be able to take any user input. The emulator for blackberry playbook is pretty much useless. Not having an actual playbook, we just have to trust that it all works how it does on the android phone, and submit the app before the deadline. If RIM keeps it's promise and sends us a Playbook, we can test the app on it then.

The submission to the Blackberry App Vendor Portal was tedious, at a minimum. Vendor account took overnight to get approved. When submitting the app, you have to fill out 5 step wizard. You cannot save the steps. So, if you don't have some information right away, and want to come back to it later, you have to start all over again. The images are non-standard pixel sizes - 480 X 480 instead of 512 X 512. And, there is a confusing array of licensing, and tax information fields. I ended up with fixed pricing and static licensing to avoid last minute integration with blackberry playbook SDK. I am not sure what the tax reporting stuff meant. The documents were not clear on what you should choose if you are a corporation vs. a sole proprietor.

Still, kudos to RIM, for upgrading their OS to use Android and doing a promotion that publishers can appreciate. It seems windows phone will integrate more seamlessly with Android apps. Windows 8 has an Android app runner called BlueStacks. Now, if only iPhone and iPad would run Android apps... all the tablet and phone devices would be covered by the same technology stack for developers.

It is amazing that 15 years after it was created for phones, Java is finally the most prevalent platform for smart-phones.

Discovering the First Big Bug

2012-02-11T23:46:00.000-08:00

I build kindergarten math myself, and test it myself. The combination is not ideal. In a company, the roles of developer and tester are separated on purpose. As a developer, I have blind spots that a tester would catch. And recently, one of those blind spots, caused a big bug in Kindergarten Math Lite.

Kindergarten Math Lite was referring to sound files that were only available after the full Kindergarten Math was installed. But, most people would install Kindergarten Math Lite first. So, they would not get any of the voice instructions for the exercises. I always had both Kindergarten Math and Kindergarten Math Lite installed on the phone when I tested either. So, I did not see the problem. The problem was there for more than 1 months. I only just discovered it accidentally while preparing for an update and fixed it.

The downloads of Kindergarten Math Lite have been going up during this whole time, and so have the sales of Kindergarten Math. I had no indication that anything was wrong looking at the numbers or the crash reports. I wonder how it would've been different had the bug not been there.

Oh well. I live and learn.

A Sincere Thankyou to Technology Enthusiasts

2012-02-03T11:59:00.000-08:00

Kindergarten math Lite is different from most other educational math games on the android market. Where other games test what the kids have learned, or help them practice it, Kindergarten math teaches the concept.

Katie wrote about Kindergarten Math in her post at iste.org

...this Android program offers a variety of exercises that speak to the way young minds pick up mathematical concepts. For example, by moving balls in and out of a box users are able to actually see the process of addition and subtraction.

She expressed what I designed and built intuitively, in my attempt to teach my kids math. Quoting another commenter ...

(Kindergarten Math has a) Very simple interface with minimal distractions. My daughter loves it.

That also implies that to a lot of people, it can feel as though Kindergarten Math is not enough graphics or media-like. Kindergarten math is, in a sense, a new way of using technology for education. So, it may show the same growth pattern that new technology does. Below is the graph of the adoption of new technology or product - in this case.

The first adopters of Kindergarten Math are innovators and technology enthusiasts like Katie. Thanks to them, Kindergarten Math just crossed 500 paid downloads. It is now moving gradually into the zone of usage by early adopters and visionaries.

The chasm is the point beyond which many applications don't make it - perhaps because they fail to appeal to a broader audience. Kindergarten math has yet to cross the chasm. Will a lot more kids like and learn from the app? I hope so.

References:

Is Kindergarten Math Successful?

2012-01-29T19:55:00.000-08:00

Looking back at products that were successful, it is easy to say they were good, useful products. Look at this graph showing the growth of Android. It is clearly exponential growth. There is an inflection point at which it went from being linear i.e. slow and steady growth and took off.

Many successful products in technology show this type of exponential growth. But, when you are in the early stages of the product, looking at your linear graph, you may just be sitting at the early part of the exponential growth - that's linear. Or, linear - slow and steady growth is all that you may achieve. There is no way to tell in advance how your product or service will do in the market.

For Kindergarten Math, is the growth of active installs is starting to turn up?

Does the recent growth of daily installs looks somewhat exponential?
Not yet, I think. But, I'll be waiting, watching and looking for inflection point(s).

Usage and earnings are just one way of defining success. Another is the joy creating something I am proud of. In that sense, Kindergarten Math is already successful.

Will Kindergarten Math be available on the iPhone soon?

2012-01-29T15:25:00.000-08:00

A few people have asked me if Kindergarten Math is available for the iPhone. Not yet. Question is should I work on porting Kindergarten Math to iPhone now? I don't have a lot of resources - time or money. So, I have to chose whether I use those resources to build another android app for kids. Or, do I convert Kindergarten Math app to iPhone.

Android vs iPhone Development
Kindergarten math is a highly interactive application. It uses the full power of a 2 dimensional touch surface to create a realistic experience for the kids. To be quick to respond, it is written in native java for the android. I coded it using a 4 year old linux machine. It did not require me to buy new hardware. For testing the app, I borrowed my husband's 6 months old android phone. On days that I wanted to work on my app, he carried my old Nokia phone instead of his Android.

iPhone has a different native system. To create the same app for the iPhone, it must be rewritten using iPhone's native language - Objective C. Objective C is specific to Apple. It has it's own proprietary tooling, debuggers and ecosystem. I will need to invest time and money to learn it. Any iPhone apps need to be written using the Mac. And, they need to be tested on an iPhone. Just the hardware to get started writing an iPhone app is at least $2000. It is not a problem for someone who is already an iPhone fan. But, I have never owned an iAnything.

Android vs iPhone Market
Android is ~50% of the smart-phone market. iPhone is about half of the android market, give or take a few percent points. iPhone app sales, however, are higher than Android app sales. iPhone owners buy more apps than Android owners.

In tablets, iPad has 60% of the market share, and android has 40%. The trend, however, is in android's favor. Question is, will android take over iPad this year... just as it did iPhone last year?

My choice
The real question is...
Do I want to reach the largest audience - Android?
Or, do I want to make the most money - iPhone?

I would rather reach a larger audience, and have more kids learn math and logic. So, I'll build the next Kids App on Android.

I'll wait till end of 2012 to see if I should port any of the apps I have by then to iPhone or not.

Android App Sales Taxes in Washington State

2012-01-26T14:17:00.000-08:00

Infinut is based in Washington State. Turns out Washington state is one of the few states to tax sales of digital goods - and that includes Kindergarten Math.

10 days after posting Kindergarten Math for sale, I got a call of Washington department of revenue. I had made all of $10 or so at this point. It was a bit pre-mature. But, since then, I have been researching the Washington state tax rules. Here's what I have found so far.

Retail sales within the state of Washington by a Washington state company are taxable for B&O tax (~0.5%) and Retail tax (6.5%). I had 2 sales in Washington in Dec - $1.98 which I must report on the company's state tax quarterly return. Google market provides information about the state where the sale happened, so I can figure out which ones are in Washington. It also allows me to set tax percentage for the state, so, buyers are charged Retail tax if they live in Washington.

Amazon market does not provide detailed information about the sales. Also, they make it clear that what they are paying us is royalties. Royalties are taxed via B&O tax (~0.5%) in Washington state, so, I will report all amazon sales of Kindergarten Math as such.

I am still not sure if the city also charges retail tax on digital goods. It will be time to call the Department of Revenue at tax filing time to verify that I got it right the first time.

Amazon Android Market - Pop and Flop?

2012-01-24T11:02:00.000-08:00

Since I wrote this post, I have discovered a bug in my code that prevented sound instructions from playing for 99% of the users. The update with the fix just went live on Amazon Market - 2/16. Since then, the conversion has gone up considerably. I will post the numbers soon. The below may not be correct since the upgrade.

Kindergarten math lite on Amazon's android market started off with a pop. It reached 400 downloads a day during the holiday weekend. But, since then, it's all been downhill. The ranking for Kindergarten math within amazon's market has not changed much during that time. So, I have to assume, the same story is true for the rest of the kids apps on amazon market.

for the last few weeks, it's about 100 downloads a day for the free app, and about 1 order a day for the paid app. That gives a conversion ratio of about 1%. The conversion ratio on the Google android market is closer to 3%.

I wonder if the conversion ratio is affected by the buggy Test Drive the App on your computer capability that amazon provides. It does not play sound correctly. Or, have Kindle Fire sales dropped?

Full vs. Lite App

2012-01-17T23:01:00.000-08:00

Nothing in life is free. Developers also need to earn.

There are two revenue models for content on the web, and for apps. Predominant one is advertising supported revenue. Second revenue model popular for apps is to have a free lite version that demonstrates the app, but is restricted in the number of times or types of things or the amount of time it allows the user to use the app. For a Kids educational app, advertising is not suitable. It is distracting as well as inappropriate. So, we decided in favor of having a lite app, and a low cost full version.

Kindergarten Math Lite is the free version with 25 exercises. Kindergarten Math is the full version with 250 exercises. The conversion ratio is the percentage of people that buy the full version vs. ones that use the lite version. Typically conversion rates in the android market are 1-5%. Kindergarten math lite has about 10,000 downloads. Kindergarten Math has roughly 300 downloads. So, the conversion ratio is about 3% for the Google android market.

One very interesting difference is in the conversion rate for phones vs. tablets -
For android 2.x (phone operating system), Kindergarten Math has ~125 active installations, and Kindergarten Math Lite has ~5000 active installations. So, conversion ratio is about 2.5%.
For android 3.x (tablet operating system), Kindergarten Math has ~70 active installations, and Kindergarten Math Lite has ~500 active installations. So, the conversion ratio is about 14%.

That is quite a difference. Does this mean I should start writing games optimized for tablets? I am not sure. It is not sufficient data to say. But, this year is supposedly the year of the android tablet. I am curious how the ratios and totals will look at the end of the year.

Sprinting past ten thousand mark

2012-01-16T15:46:00.000-08:00

Kindergarten Math Lite crossed ten thousand downloads on the Google android market today. There is additional four thousand downloads on the Amazon android market of Kindergarten Math Lite. It took about 2 months since release to get here.

Getting here was more about building the app that kids like, not about marketing it. People who liked it have been kind enough to post good reviews. So far, this is a market where apps sell themselves. Once there are larger number of publishers and apps per topic, I expect it would require more marketing to get thousands of downloads.

There is about 220 downloads a day on average for Kindergarten Math Lite on the Google android market plus another 100 a day on the Amazon android market. At this rate, if it sustains, it will take about 6 months to get to the next point of fifty thousand downloads on the Google android market. That should be sometime in July. It will be interesting to see what happens between now and then.

The ratio of actives to downloads is about 62%. Percentage of actives was higher in the beginning, and has been slowly going down. People who are done with it are un-installing it slowly. The chart showing total active installations is below.

Along with crossing the 10,000 download mark, the ranking of Kindergarten Math Lite jumped about 10 places higher on the google android market for the search for Kids Math today too!

First month of Paid Kids App on Android Market

2012-01-13T20:31:00.000-08:00

I released Kindergarten Math on the android market on Dec 14th, a month after the release of the free (lite) app with the same name but lesser content. I changed the free app to direct to Kindergarten Math for customers who complete the free app... about a week later.

The first week was slow. Without the thousands of users of Kindergarten Math Lite, getting any indication of the presence of the new full version, there wasn't much sales. Sales picked up when the free app was used to direct to the paid app.

For the first month from Dec 14 to Jan 14, sales were roughly 250 copies - that comes to about $170 after market fees. It is my first $100+ that I have made from directly selling software to consumers. I know it's not much to brag about, but, I couldn't be prouder of it.

Sales are mostly US customers, about 90% at this point. France and Australia are next in sales. Followed by Canada, UK, Thailand, Indonesia. I am happily surprised by the broad appeal and the international market that Android phones and tablets provide for me... just sitting in my house, writing away software for kids I love.

Sales are almost twice as high on weekends and holidays than on weekdays. More time with kids, results in more sales, I think. The peaks in the graph below are weekends, and the valleys are weekdays.

Android's many markets

2012-01-12T14:03:00.000-08:00

When I built Kindergarten Math, I thought there was only one market - the android market run by google. Within the last few months, I have been finding more. I must have made it to some marketing list for App developers. I am starting to get invites to post Kindergarten Math to them recently. Here are the ones I have looked at so far.

Google's Android Market
Is the only one I knew about when I started. It is low cost for a publisher- only $25/year and a publisher can publish as many apps as they want. They take a percentage for processing the order, same as all the other markets. About 1 in 20 orders on the google market fails because of some download error. And then customers are not able to re-try for 7 days. Google market has it's serious flaws as far as commerce side is concerned.

However, it's openness is admirable. Also, having put in so much into Android and making it free and freely available, google deserves to make money from it's market.

Amazon's Android Market
Amazon a serious contender as an android market, mostly because of amazon proven retail strategy. But, Android is a whole new beast. It wants freedom. If it wanted to live in a walled garden, wouldnt' it just be an Apple instead. Amazon's desire to control both it's suppliers and it's customers may not work here. We will have to see. Amazon had to backtrack and allow access to Google's Android Market on Kindle Fire.

Amazon was offering $50 back per developer for publishing to it's market back in Nov, 2011. It is also free to publish your app for the first year. Second year, it's supposed to go up to $99/year. It is all evolving too fast to predict what will happen next.

HeyZap
HeyZap contacted me only a couple of weeks after posting Kindergarten Math lite. They are a games store with a social bent. I haven't tried it ...mostly because I am not a gamer or too social. It has some concept of checking-in when you are playing a game. We can broadcast to our friends what game we like by playing it. Game developers get some free viral advertising with it. But, social check-in is not suitable for little kids. So, not for Kindergarten Math.

Android Tapp
Recently launched their market. They used to do free reviews of apps. Now they want to get a share of selling them too. I think they only do paid apps. I am unsure of market share, and if there is a point posting Kindergarten math on it.

I think I submitted LocalByUs app to it to get a review at Android Tapp. But, it wasn't reviewed. I dont' blame them.

Barnes and Noble Nook Market
Barnes and Noble is trying to be Amazon. They are trying to outdo them in control. As a developer, I can't even read their documentation unless I register first.

The Nook does not allow access to the Google's Android Market. I think B&N hope that Nook sales will drive people to their market. From my point of view, their market is driving off Nook sales. We thought about buying a Nook, but, decided against it, primarily because it is so customized and restricted. Still, it does seem to have a lot of books available on it.

I am still debating whether to post Kindergarten Math to it or not. I am not sure if it's worth the effort. But, who knows. Maybe I'll just give it a try.

Cius Market
Is Cisco's venture. They want to own the Enterprise App market. I am not sure why they invited me to put Kindergarten Math on it. Desperation? Cisco getting into it is a bit odd. I would've just thought it would make sense for them to build Enterprise Apps, not a market.

Seems like a lot of people want a piece of the android market. It makes it hard for a developer to figure out where to post the app. It may look fragmented for now. But, most likely only a few... I'd predict a maximum of 3 general purpose markets will survive. Some competition is good for publishers too.

Most of the markets come with their own Developer Kits to integrate with. Amazon has one, Nook another, Heyzap, cius, ...all have their dev kits. It is hard for a developer to integrate with all of them.

It is not developers like me who will decide the success or failure of the markets. It is customers like me that will.

Finding my team, building what I know, doing what I love

2012-01-01T11:59:00.000-08:00

Steve Jobs said he looked in the mirror every day and asked himself -
If I were to die today, would I want to be doing what I am doing today?
If the answer was No too many times in a row, he knew it was time for change.

Almost a year ago, after a year of struggling at Expedia, I finally decided it was time for a change. I could have found a job working for a different large software company, but, I am convinced it is not for me. I still love creating software, just not the politics that comes with corporate america.

So, with the support of my husband, I have been trying to launch my own projects for the last year

GardenNut.com - was a website for exchanging or selling plants from home.
LocalByUs.com - was a website for sharing what's happening hyper-locally.
LocalByUs app - was the supporting app for reading or writing posts.

Each of the above projects have now failed and I have shut them down. GardenNut.com and LocalByUs.com cost $200/mo to keep the servers running, and yielded about $5/mo in ad revenue. The local by us app had all of 21 downloads, only 6 of them active after 3 months of being on the market. The graph belows shows it clearly.

My fourth project is building educational apps for kids. The idea was my husband's. He was looking for fun and educational app for our 5 year old, and he couldn't find much that was suitable. Many of the free apps required internet connection and access to your private information. He wanted to build one that was honest and free, but, working full-time supporting us, he couldn't find the time.

In the meantime, I attended an android app hackathon. My main purpose in being there was to promote LocalByUs app that I had created. The hackathon required us to create something during the 8 hours of the hackathon. I decided to use my husband's idea and create a simple maze for the kids to play. That maze later became the numbers maze in Kindergarten Math.

My daughter loved the maze I had created. She asked me to create some more. And I did. Along with counting, sequence, addition and subtraction exercises for her to play. In the meantime, I got a software development contract offer that, not having earned for a few months, I could not refuse. I published Kindergarten Math Free before starting work.

Unlike the projects I did before, I built Kindergarten Math for my kids. I wanted to share my love of math and logic with my daughters. In creating Kindergarten Math, I found the team I work best with - my family.

Amazon vs Google Android market for a Publisher

2011-12-27T15:23:00.000-08:00

Download Count
Amazon android market has proven to be better for me in terms of daily download counts for the last few days. It was surprising since I thought less people have amazon market app on their machines. Amazon doesn't break down the device type, so, I cannot say if this is entirely because of kindle fire sales, or other devices as well.

Amazon has a specific category for kids. So, it is easier to find kid's apps. Amazon also has a review process, which results in better quality apps on their market. Google's Android market, on the other hand, has less restrictions on apps for children. So, there are kids apps on Google's Android market that require permission to read private information and connect to the internet.

But, as a publisher, what matters is who does better with conversion and sales. I will post first month of paid app sales for Kindergarten Math on Google Market and Kindergarten Math on Amazon Market in a few weeks.

Statistics and Errors
For a publisher, it is invaluable to be able to see where their market is and how people who have downloaded the app are doing. Google is the clear winner here. Google breaks down the downloads by country, by device type, and operating system version. It helps a lot in terms of planning for next steps - which language to translate it to, which hardware to customize it for etc. Amazon provides no such data yet.

Google also provides detailed information about crashes. It gives an idea of the quality of the experience the user's are seeing. The information is detailed enough to be able to fix any issues in subsequent releases.

Publishing Ease
Google market is the clear winner in ease of publishing. The publisher just has to upload the app, and it's published. Newer versions are easier to publish too. The publisher can provide the description. The publisher can unpublish the app at any time. Publisher sets the price. In Amazon's case, amazon edits the description. Amazon also reviews the app when it is first published and every new version of the app too. New version takes another week or two to be reviewed and approved. Amazon can also discount your app at any time. Amazon won't allow unpublishing the app unless the app is being discontinued. So, as a publisher, once published on Amazon, you are stuck with it whether you like it or not.

First month of publishing a kids app on Amazon android market

2011-12-27T15:03:00.000-08:00

Kindergarten Math Free was published on the amazon market at the end of November.
Amazon's market is a bit different from the android market. It takes longer to publish the app on the market. It has to be reviewed and approved by Amazon, which can take a week or two. I happened to publish it right around the time Kindle Fire came out as well. Kindle Fire has android based OS, so, it can run apps built for android.

Unlike the google android market, the amazon android market was very slow uptake initially. Then, around 12th of the month or so, it started seeing 30+ installs a day. My theory is that the app got approved for kindle later, or kindle activations kicked in around that time. Installs per day was rising slow and steadily, first to 50+, then 80+, then 120+. Last two days, it suddenly jumped to about 300+ installs. That is considerably more than the google android market downloads for the same day.

My theory is that is a lot of kindle fire christmas presents. People are now loading their presents up with apps. The amazon market also has a dedicated category for kids. It helps people find quality content for their kids more easily. Their recommendations engine also kicks in more-and-more as people download the app. More people download it, then, it shows more often in people-who-downloaded-this-also-downloaded. That drives more sales.

I don't expect the exponential part of the curve to last beyond the holiday season. Specially, because there are some negative reports on Kindle fire.

First month results of publishing a Kids app on Android Market

2011-12-19T10:56:00.000-08:00

I published Kindergarten Math Lite to the google android market on Nov 14th, 2011. It started out well. Going to the 50-100 downloads point within a day. The number of downloads and the ranking kept rising, slowly. The number of total installs is in the android market is at 4000 right now. I expect it to cross into the coveted 5000-10000 category in roughly a week's time. It became the first app for keyword kindergarten a week ago. There is about 100-200 downloads a day at this point. Below is the graph of the downloads for the first month.

The number of downloads rise during the weekends, and falls during the week. The peaks in the above graph are weekends and the valleys are mid-week. Perhaps, that's because it's a kids app. Parents are looking for ways to keep the kids busy and learning during the weekend.

The chart of active installs looks like a rising curve mirroring the total installs. The curve is beginning to flatten at the end. This could be because people who downloaded it a month ago are done playing - there is only 25 exercises in the free version. Those people are now starting to uninstall the app, affecting the active numbers.

Most of the downloads are in US. Next is Germany, then Canada. It is english app, so that explains the US and Canada uptake. The German uptake is probably because of the name of the app - Kindergarten is a German word, so, I expect it is a likelier search in Germany.

Overall, it is exceeding my expectations. I will post the amazon market uptake next.