Experiences Developing An Android App For Money

General topics about the Android-Platform itself.
Coding issues please to the subforum right below.

Experiences Developing An Android App For Money

Postby stoneknuckle » Sat Jul 10, 2010 5:35 pm

*The developer*
Profressional developer with over 10 years of experience primarily developing government-related apps in VB, javascript, asp, html. Believe it or not, government services apps don't always get your creative juices flowing, so I wanted to write something that was creative and see if I could make any money doing it. This was my first Android app and my first time developing in Java. I learned by poking around the SDK examples and tutorials on the web.

*Why I'm Writing This*
I'm a concrete numbers person and when I was getting started I wanted as much "real world" information as I could find from other devs. I hope those of you with your own experiences will contribute them and those that are getting started will learn from our experiences.

*The App*
A mainstream entertainment-related app that uses Flurry for analytics and AdMob for ad delivery.

*The timeframe*
The first three months of release

*The numbers*
Android Market Stats:
    *3.5 Market Rating
    *50,000 installations
    *55% of installers still have the app installed.
    *approx 1 in 200 installers have left a rating
    *approx 1 in 3 raters have commented (1 in 600 installers have commented).

Flurry Stats:
    *15,000 total "active" users each week
    *8,000 new users per week
    *OS Version distribution (most recent week): 43% 2.1-update1, 28% Android 1.6, 20% Android 1.5, 7% Android 2.1

Admob Stats:
    *Revenue Per Day: between $7-$9 (at approx 65% fill rate). But this is only a recent figure. For the vast majority of the apps release, revenue was between $2 and $3.50 per day.
    *Fill rate: around 65% but as low as 40% and high as 85%. Fill rate was much higher (95%) in the beginning and has definitely trailed off. Not sure if this is because more apps are available now, causing a glut of inventory, or what.
    *CTR: between 1% and 1.4% when ads weren't refreshed, now down to .25% when refreshing every 60 seconds.
    *Revenue Per Click: for the most part, between $.02 and $.035. However, in the past two days it's risen to an all-time high average of $.043. Not sure if that's an anomaly, if AdMob is globally showing higher-value ads, or if the app has crossed a threshold with the number of clicks it's generating and AdMob is now delivering higher-value ads specifically for this app.

General Thoughts
    *It's not a "cash cow". I don't want to think about what my hourly rate would be if I divided the hours I've put in by the total revenue. On the other hand, it's rewarding to make a little money on an app that was basically a learning tool.
    *It's taken a few months to get the bugs out. And now that they are out, the users are sticking with it, the ratings are improving. And revenue is finally on an upward trend.
    *Updating regularly (once per week) is a good idea, but only if you've actually got something to update. It keeps your app in the mind of your existing users and puts you in your category's "just in" section of the market for some period of time which attracts new users.
    *Design for the earliest version of the OS you can. After modifying a few things to support version 1.5 of the OS, I immediately increased potential user base by 20%. Also, compile with the earliest version of the SDK that you can. I wanted version 1.5 users to use my app, and set my minimum SDK version in the manifest to 1.5, but I was compiling with the 2.0 SDK. That didn't work out so well :). Compile with the SDK version that matches the minimum SDK version.
    *It's frustrating when a commenter says "this app sucks. doesn't work at all", leaves a 1 star rating, and you have no idea what they found wrong with the app. Then, the next commenter says "it's the best app available" and leaves a 5 star rating.
    *The app got a lot of new users in the first few days, then usage (and ad revenue) trailed off and flattened (10k new users the first week, 5k new users the second week, 3k new users the fifth week).
stoneknuckle
Freshman
Freshman
 
Posts: 7
Joined: Sat Jul 10, 2010 5:14 pm

Top

Re: Experiences Developing An Android App For Money

Postby OperandZombie » Sun Jul 11, 2010 11:01 pm

Thanks for this post; I've been looking for this type of info for awhile. I started tinkering with Android development recently, and am at a point where I need to decide if it's worth pursuing or not, so this was a nice read.

If you don't mind me asking, how much time did you spend initially creating the app? How many hours a week do you spend working on it for your next update?

Thanks!
OperandZombie
Freshman
Freshman
 
Posts: 3
Joined: Sun Jul 11, 2010 10:58 pm

Re: Experiences Developing An Android App For Money

Postby stoneknuckle » Mon Jul 12, 2010 11:09 pm

OperandZombie,

Glad is was helpful.

I didn't keep a tally of hours spent, and there was a two month period where I wasn't developing it at all, so it's hard to say. I'd guess initial development took approx 200 hrs. But I'd note two things: 1) This was my first time working with Java and/or the Android SDK, so there was a big initial learning curve there. 2) The app has some certain unique technical challenges that most apps wouldn't have which also added greatly to the initial development time.

After the initial release, I was still spending quite a bit of time per week adding features, fixing bugs, etc. Probably 10-20 hours per week. As things got more stable I worked fewer hours per week. Also, when I realized how little I was making per day, I wasn't as inclined to work on it every day. I'm probably down to 2-3 hours per week on development now. Almost exclusively bug fixes that are reported in the Market's comments and directly to me via email. If there was more revenue potential, I might be inclined to add some cool features, but right now I'm more interested in developing other apps.

Here are a few more thoughts on revenue/business models I've had rolling around my head. Excuse the stream of consciousness:

1) Rather than go for the knock out app that makes your target of $X dollars per day, develop 5 apps that make $X/5 dollars per day. Of course there are exceptions, but generally I'm not convinced an indie dev with limited time can develop a single homerun app. This may sound crass, but maybe spending a little time creating mediocre medium-value apps would be more profitable than spending a LOT of time developing a single high-value app. I'm not sure if this would pan out, but it definitely plays to my A.D.H.D. nature and allows me to try lots of different projects.

2) Heed the lessons of web advertising. Right now there is a little money to be made on Android advertising. But remember how advertising evolved on the web. As inventory increased (more websites displaying ads), ad rates dropped like a rock.

3) Develop a technology or architecture that can be reused. This goes back to #1 above: develop multiple apps to generate your target revenue. No need to reinvent the wheel each time. Develop some sort of platform once (game, etc.) that can be modified slightly each time to create several different apps. While the first app I wrote for Android was more of a utility, my next app is a game. The game, however, utilizes the core features of the architecture I wrote for the utility. And I don't mean they share code for accessing a DB. I mean the core engine that makes the utillity valuable also makes the game valuable. This way, I'll penetrate two completely different markets with a single core technology. Not to mention that the game itself could be modified and released as a few different apps.

4) Sell apps. When the barrier to entry is so low for creating and publishing apps (there's no review of apps before being added to Android Market), and the barrier to adding ads to those apps is also low (I can add AdMob or MobClix to any app I create), there is going to be a glut of inventory for advertising, and revenue potential per user will drop. Develop an app that is unique in your niche and sell it. It's hard to sell an app when there are one or more comparable apps offered for free, so developing something that is sufficiently differentiated from other apps is crucial.

5) If you've got free apps that are trying to generate revenue via advertising, consider using those apps as marketing tools for your paid apps (it helps alot if your paid app is of a type that would appeal to people who use your free app). The entertainment app I described in my initial post has some traffic. Depending on how well I can convert users of a free app into buyers of a paid app, it might make more cents (get it?) to only advertise my paid apps rather than showing AdMob ads. Let's say I'm making $3.00 per day via advertising on my free app (which was actually the case up until this week). And let's say I'm making $.02 cents per click (which varies each day based on AdMob's mystical algorithms, but at one point was actually the per-click amount I was making). That's 150 clicks per day. If instead of running AdMob ads I only ran ads for my own paid app, I'd need to turn 1 click out of 150 per day into a paid customer of a $2.99 app to match what I'm making on AdMob. That would approximate to a .66% conversion rate. I don't have the numbers in front of me, but I''ve looked at other apps in my niche that have a free version and a paid version just to get an idea of how well the free version converts into a paid version. I recall being pleasantly surprised. The conversion rate was much better than .66%, suggesting that given my numbers I laid out above (150 clicks per day, $.02 per click, $3.00 total per day), I might be better off using my ad inventory to advertise my paid app. Or at least use a percentage of the inventory to advertise the paid app. In any case, it's something to consider.

Wow, that's more than I planned to say. To me, figuring out the marketing of apps is as fun as developing the app itself, it's like a game. I have no illusions of making a mint developing Android apps for a few hours each week. The $3.00 per day or $10.00 per day or whatever the total per day I might make is is really just a little score card I use to see if I'm playing "the game" any better.
stoneknuckle
Freshman
Freshman
 
Posts: 7
Joined: Sat Jul 10, 2010 5:14 pm

Re: Experiences Developing An Android App For Money

Postby OperandZombie » Tue Jul 13, 2010 7:25 am

More great info, thanks! Couldn't agree more with reusable components.

I'm in a similar boat as when you first started I think; I did software development professionally for a few years, then recently switched to a database architect position - when I was coding all day long professionally, I'd come home and not want to touch a computer. Now that code only accounts for a smaller fraction of my job, I'm much more interested in working on my side projects again. Android happens to be one.

I'm also not expecting to make bank here. I've got an idea for an app that I mostly want for personal use (to aid in my photography hobby) that will likely be my first project, and then a pair of games for followup projects (one of which currently has no similar competitors to my knowledge - hopefully things stay that way by the time I get to that app :)). I'd love it if I could pull down enough money to cover my phone bill monthly; that'd be enough for me.

What resources did you use to get started? I have a book, and I've been following the online tutorials, but I'm always down for more info.

Thanks again.
OperandZombie
Freshman
Freshman
 
Posts: 3
Joined: Sun Jul 11, 2010 10:58 pm

Re: Experiences Developing An Android App For Money

Postby stoneknuckle » Tue Jul 13, 2010 4:20 pm

I started with the SDK samples just to get my bearings in term of the platform's UI stuff. Then I just started coding, really. Any time I ran into an issue, I'd do a google search.

Interesting that you brought up reading a book. I've thought about whether I should have done that. This is my first real experience diving into a new language since, like, 1998. Back then I remember digging into VB books and it being helpful. For some reason, now I'm not as inclined to do that. There's definitely stuff that I'm missing by not reading a well-layed out explanation of the platform, but I was more interested in just getting started. I suspect I'll pick up a book in the next few months just to get a better idea of what is possible. It seems to me that when you know what you want to do, google can help you do it. When you want to know what you COULD do, having a full understanding of the platform is necessary.

That's cool that you're going to start with something you want to use yourself. Obviously, you know there's a market for it if you want it. That's exactly the same thing I did. There were a few apps that kinda offered the tool that I wanted, but none did it in as user-friendly a way as I wanted.

Another tip: Unless you're going to pay to advertise your app, the only real way to "get found" is via the Android Market. To improve your visibility in the Android Market, put the keywords that people are likely to search for in your app's title when listing it in Android Market. For example, if your app was a tool for searching for photos, you may call it "WhizBangSearch". When you list it in Android Market, call it "WhizBangSearch Photo Search Tool". Cuz even if you have the keywords (photo search tool) in the description, apps that have those keywords in the title will display above yours.
stoneknuckle
Freshman
Freshman
 
Posts: 7
Joined: Sat Jul 10, 2010 5:14 pm

Re: Experiences Developing An Android App For Money

Postby blundell » Wed Jul 14, 2010 9:39 am

Really good info,
Sticky This :-D

Oh and if you have the brain power to convert the word iPhone into AndroidPhone read this:
http://www.iphonedevsdk.com/forum/promo ... noobs.html
User avatar
blundell
Master Developer
Master Developer
 
Posts: 1610
Joined: Tue Nov 18, 2008 12:58 pm
Location: UK

Top

Re: Experiences Developing An Android App For Money

Postby stoneknuckle » Wed Jul 14, 2010 9:41 pm

Glad you found the info usefull blundell.

Really interesting link too.

I'm not sure I'd want to be writing an iphone app right now unless I had a brand name behind it cuz it seems hard to break through the noise. $50k in dev hours and 50 downloads a day of a free app? Wowww.

One thing that excites me about the "newness" of android (ok it's a few years old but the marketing push is really just getting started for it) is the opportunity to be come the provider of "the app" in a particular niche. If you can establish yourself as the provider of the must-have app in a niche, you'll thank yourself 5 years from now.

The android market-sphere will likely go the way of the iphone appstore-sphere and eventually it'll get really, really cluttered. Even if you haven't developed a "must have" app, if you can build an app and get a sh*tton of people using it now, it could be a valuable marketing tool for promoting your other apps down the road when getting noticed is even more difficult. And each user of your app now could genereate two or three new users in the future (ever been asked by someone who just got an android phone what apps they should install?)
stoneknuckle
Freshman
Freshman
 
Posts: 7
Joined: Sat Jul 10, 2010 5:14 pm

Re: Experiences Developing An Android App For Money

Postby midasmax » Mon Jul 19, 2010 8:59 pm

Invaluable advice stoneknuckle!

I can't emphasize enough how much all the above actually matters, and I now wish I had seen this thread before I began my furore into Android development. I've recently just launched my first android app onto the market, and it's definitely one hurdle to develop the app, but marketing the app is a whole new ball game altogether.

My app has been launched for slightly over 5 days now, and I've seen number of installs per day drop from almost 500 on the first day, to approximately 30-40 new installs now. What intrigues me is what makes a game 'tick' with the users. I've seen similar applications to mine selling for $1.99, and with all due respect, don't seem very polished or complete. What's worse, it appears that they haven't been updated in a long while, yet, the numbers and ratings for those apps seem to be quite high (4.5 stars).

I'm not sure what the trend, but I doubt I'll be seeing >100 new installs per day anytime soon, so I am wondering how games can continue marketing themselves to the users. I recently made the fatal mistake of updating my app before a week has passed, and it wasn't simply a small update. I was eager to please the masses based on feedback I had received, and thought that pushing out an update with all the new features would prove to those who rated it poorly or made suggestions that we were willing to do our best to polish the app up. Always wait a week at least, as that 'Just In' category is vital especially when your app is new and has no coverage whatsoever.

I'm not sure about the life-span of apps, and would be interested to hear other dev's experiences. Do you see a spike in downloads and then a gradual decline until it stagnates? Or does it periodically increase from time to time? And if so, is there any basis/cause for those increases?

Another point about refreshing of ads. It's interesting to note that a 60s interval actually affected the click-through rate negatively. I would have thought showing the same ad would definitely be more boring. I'm using Admob by the way.

Finally, thanks once again for the insightful post.
midasmax
Junior Developer
Junior Developer
 
Posts: 15
Joined: Wed Jul 14, 2010 11:07 pm

Re: Experiences Developing An Android App For Money

Postby stoneknuckle » Tue Jul 20, 2010 9:47 pm

midasmax,

The click through rate is based on the number of ads the user is shown. Since I was refreshing the ads every 60 seconds, the number of ads being delivered went way up while the number of times users were clicking remained pretty much the same. For example, say I was delivering 1000 impressions per day and generating 10 clicks. 1000 impressions/ 10 clicks = 1% click through. With refreshing, under virtually the same usage scenario I was then delivering 4000 impressions and still getting 10 clicks. 4000 impressions/10 clicks = .25% click through.
stoneknuckle
Freshman
Freshman
 
Posts: 7
Joined: Sat Jul 10, 2010 5:14 pm

Re: Experiences Developing An Android App For Money

Postby midasmax » Tue Jul 20, 2010 10:29 pm

stoneknuckle:

I see, yes, that makes a lot of sense, thanks for clarifying that up!
midasmax
Junior Developer
Junior Developer
 
Posts: 15
Joined: Wed Jul 14, 2010 11:07 pm

Re: Experiences Developing An Android App For Money

Postby smobiler.two » Thu Jul 22, 2010 4:00 pm

Thank you for the detailed outlines. In terms of marketing in, how did you manage to get such a number of installations in a short period of time?
How to make your Android software work on iPhone, BlackBerry, Symbian, WM7
New SDK lets you develop cross-platform applications for mobile devices.
Visit http://www.uvamobiltec.com for more or follow UVA on Twitter at #uvamobiltec
smobiler.two
Freshman
Freshman
 
Posts: 2
Joined: Tue Jul 13, 2010 3:50 pm

Re: Experiences Developing An Android App For Money

Postby stoneknuckle » Mon Jul 26, 2010 4:49 pm

smobiler.two,

I didn't do any marketing outside of the Android Market (no advertising to drive traffic to my app, for example). Within the Market, I made sure that the keywords that I thought were the most likely to be searched by users were included in my app title and in my description. That's pretty much it. Also, I make sure to update about once a week -- that way the app shows up in the highly-visible "just in" category of the Market.

I didn't realize that was a high number of installs. I'd be curious to know what installation rates other developers are seeing.
stoneknuckle
Freshman
Freshman
 
Posts: 7
Joined: Sat Jul 10, 2010 5:14 pm

Re: Experiences Developing An Android App For Money

Postby midasmax » Mon Jul 26, 2010 7:18 pm

Based on various people whom I've spoken with, different ideas have been thrown about with regards to getting in the 'just in' category.

stoneknuckle: Do you know if 1 week is a big enough buffer to get into the 'Just In' category? One person I've spoken to suggested at least 2 weeks, which imo seems a little too long, especially since I actually have legitimate features/updates/bug fixes which I want to work into the game as soon as possible. At the same time, I want to ensure it gets maximum exposure.

Also, supposing you uploading on day 1, and then do an update on day 3. Obviously, there is little chance it will appear in the 'Just In' category. However, should the next update be done on ~Day 8 (1+7) or ~Day 10 (3+7)?

Thanks very much once again, stoneknuckle. I can't iterate how useful all this information has been, and I sincerely appreciate you being so open on your numbers/evaluation and results. Just to give an indicator of two apps that I've had the chance to oversee:

1st app was a simple application/tool, kinda niche as well. It say about 100 downloads on the first day, and then subsequently 5-10 installs a day, eventually tailing to about <200 downloads with a active % of about 40-50%.

2nd app was a simple game. It had about 800 downloads on the first day, then about 30-40 new downloads every day and about 50-60% active downloads.
midasmax
Junior Developer
Junior Developer
 
Posts: 15
Joined: Wed Jul 14, 2010 11:07 pm

Re: Experiences Developing An Android App For Money

Postby stoneknuckle » Tue Jul 27, 2010 5:41 pm

midasmax,
My experience has been that a week is indeed enough time. For the past several weeks I've released updates on the same day each week and I can confirm that the app shows up in the Just In section.

As for timing your releases, my suggestion would be to keep your existing customers happy, so release any necessary bugfix updates as soon as possible (not waiting a full week, etc). Once your app becomes more stable, then switch to a weekly schedule. From my own experience, I was releasing updates 2-3 times per week for the first few weeks as I fixed reported bugs. But you don't want to release TOO often cuz you risk annoying existing customers ("why in the hell do I get an update notification for this app every other friggin' day???"). Soooo, to sum up my own thoughts about this: Bugfix release: asap. Feature release: once a week.
stoneknuckle
Freshman
Freshman
 
Posts: 7
Joined: Sat Jul 10, 2010 5:14 pm

Re: Experiences Developing An Android App For Money

Postby midasmax » Fri Jul 30, 2010 12:08 am

stoneknuckle

Thanks once again for the advice! :) It seems that for a developer starting out with little or no marketing budget, that the Just In category is the only way to gain maximum visibility. It's a shame that several times, it gets clouded by spam-apps. These developers release 4-5 games at a time, each consisting of nothing more than rehashing the same genre (i.e. puzzle slide game to unlock a wallpaper).
midasmax
Junior Developer
Junior Developer
 
Posts: 15
Joined: Wed Jul 14, 2010 11:07 pm

Top
Next

Return to General

Who is online

Users browsing this forum: No registered users and 2 guests