Synapsy

Generally, I am a fan of aggregators – the more app stores you bring into one interface, the more reach you can get from a single deployment of your app.  This store sent a direct email to the address on file for my apps on Google Play, and a decided to take a look.

The good?  Yes, their advertisement of submitting apps in just 2-3 minutes is pretty accurate.  It didn’t really take too much time to submit.  They pulled the images and text from Google Play.

However, in this case it doesn’t seem that their reach is sufficient to merit much of my time.  The next few months showed no downloads however, and I had submitted my top two free non-game Android apps.  I don’t think a paid app would be a good idea here, since the payout threshold is 200€.

Would games fare better?  Maybe.  Am I enticed to spend any more effort on this Android market at this time?  No.  Hopefully this market will improve as they bring more stores into their fold.  For now, I don’t recommend spending much time on Synapsy.

– — – –

Hungry for info on more Android app stores?  For access to all of my reviews of Android app stores and markets, check out the Android Market Alternatives report.  It covers dozens of Android app markets in detail, with an action plan for the top ones to spend your time and effort on.

The harsh truth, however, is that outside the shiny center of the Android universe lies vast outlands, with empty app stores that have few customers and few downloads.  Some of these app stores make a great sales pitch, and some are connected to some giant and successful companies.  This makes it hard to navigate (and is one of the big reasons I keep reviewing app stores on this site).

As an example, I noticed that InMobi launched their own Android app store aggregator this week, and I was pretty excited to see what they were doing.  After all, I know that the less work it takes to release to multiple app stores, the better.  However, my enthusiasm faded when I saw that the list contained a number of app stores that are on my no-go list, and some (like Sony) show 25 separate entries since each country shows up separately.  So much for 120 app stores…

I may still submit an app and test the waters there, and will post a more detailed review.  There may be a few hidden gems in there that could make it worthwhile.

At any rate, don’t be swayed by pure numbers of app stores.  It’s high quality app stores that we’re after.

What Apkudo Developer does

With an arsenal of Android devices, Apkudo can do a Monkey test on hundreds of devices at once.  No time-consuming script writing, no complex setup.  Just upload and see the results.

To use the Apkudo Developer service, click the ‘Get started’ button and connect it to your Google account (you must be logged in).  The log-in is needed because they save your analysis data between sessions.

At that point, you upload your APK file and then click to start .  If there’s a queue, you may need to wait before analysis begins.  Otherwise, tests will run for a few minutes, showing your the results as they are completed.

That just leaves analyzing the results – if all goes well you’ll see a high number of devices under ‘succeeded’ and a smaller number under ‘failed’.  Don’t worry if a few fail – it may not be a serious issue.  For example, install timeout errors may not be important.

For each issue you get:

•     Stack trace (for crashes)
•     Monkey log
•     Logcat

At this point, dig in to the data and see if you can find any issues you need to resolve.  The analysis interface lets you quickly filter and zoom in on specific types of failures for more information.

Don’t forget to note the other side of this – you know that your app had no trouble on all of the other devices.

How Apkudo can do this

Apkudo’s main business is Apkudo Approved, where they test handsets and other Android devices for manufacturers and carriers to certify the device for a suitable user experience.

Since they work with every major OEM, that gives them access to a lot of devices.  They test each device on 6k of the top apps from Google Play in the process of approving it.  They even test the hardware – the camera in a darkroom, audio, etc.

One good thing about this work with the OEMs is that it allows Apkudo to offer the Developer service for free.  The more developers use their tools, the more they can improve the service that they provide to OEMs (the paying customers).

An interesting point about this symbiosis is that aggregate data about what APIs and features upcoming apps are using can help give the OEMs insight into future apps.  So in this way, developers can help shape future device design and usability testing!

If you’ve been reading this blog a while, you know I’m not a fan of big companies gouging the little developers (e.g. fees to test and release an app), so I like this business model.  The OEMs pay for a service that helps them make them a lot of money, and developers get some free (or inexpensive) help.

What’s next

Josh made it clear that Apkudo is not sitting back and relaxing – they are working on a number of new features.  He emphasizes their commitment to the developer community.

For example, they are looking into how to share data (safely) between developers, which might be very helpful (we’ve all solved problems that we know that someone else has seen before!).

In addition, they plan to roll out “Ape” soon, as a more sophisticated testing system than Monkey.  It sounds like a different take on how to categorize devices, tests, and results than I have seen from other test vendors.  That could make it easier to group similar issues to resolve them.

Conclusion

I found that it took very little time to do some extra verification that my apps can pass basic tests on a variety of devices.  Given that I deploy my apps to a number of Android app stores all over the world, this kind of testing is important.  I don’t have access to a lot of these devices!

So, I suggest that you take a look at this lightweight testing tool, and keep an eye on Apkudo for future developer tools.

I also took the opportunity to update the Android Market Alternatives Report (part of the Advanced Package in the Android Income Series) with all of this information.

App Store Updates

Google Play just made the announcement at the end of January that they will move away from paying two days after the end of the month, and will transition to paying on the 15th of the following month.  I know some developers count on that revenue in a timely manner, but keep in mind that most app stores pay a full month later, so Google is still treating developers well.

A number of app stores have closed, and I don’t believe they are coming back – for example:

• Andspot
• PocketGear
• Appsnips
• Verizon VCast
• MobileApps
• Vodaphone

A few of those closed gracefully (and even still offer developer resourced), but some just disappeared.

Another item of interest is the launch of BlackBerry 10 on January 31st.  While Android apps could be run on the PlayBook tablet already, now we can expect phones to have access to our Android apps ported to BlackBerry.  I hope to see some increased download counts in the near future.

Report updates

I updated the Android Market Alternatives Report with all of the above and scrubbed all of the obsolete and dead links due to those closed app stores.

And, as always, I’ve found a few new app stores to research and add to the list.

In addition, I updated the action plan and added information on using the open source Android Market Manager library to handle all of your cross-app linking within each app store (and to help you pass QA tests in those app stores).

With all of the above changes, not to mention numerous minor tweaks and updates, the Android Market Alternatives Report is more helpful than ever to get you on the right track to earning income from your app from the right Android app markets.  Take a look!

The promise

AdColony presents itself as a premier ad network for big-name brands to distribute their HD videos directly to devices.  That sounds like a win for developers – big dollars from big brands means a high fill rate and hopefully a great eCPM.

If you’re not cynical by this point, however, I assume you haven’t actually dealt with ad networks before.  Let’s dig a little deeper.

Ad credit bait and switch

To entice signing up, AdColony offers $10 of advertising credit. That’s just enough for a micro-campaign – maybe not enough to really boost your brand awareness or even get a lot of data to crunch, but enough to see the network in action.  If it had good results, it might be a good way to improve awareness of my app (e.g. spend my own money to continue the campaign).

So, I took some time to understand the format required to make the ad, and I made a 15 second video for my Halloween app.  A timely opportunity I thought.  I created a campaign, set it to use the $10, and uploaded my video.

Fast forward three months, and my ad is still pending review.  Halloween has come and gone.  Good job guys.  If this is your level of customer service, I’m not so sure I can expect a high fill rate if I request ads.

SDK difficulties

At the same time, I was diligently working to integrate the SDK into my app.  I integrated their SDK into my app and got as far as testing on a physical device.

After running the flashy test video ad, I was unable to continue on to the next screen – touching any of the buttons at the end of the video caused the app to hang.  If I press back before touching anything, I can exit the ad, but that’s not good.  I presume that if no one can click through, I’d be incapable of earning a cent from showing the ads.  That defeats the purpose of showing the video in the first place!

To their credit, AdColony support attempted to help me with the issue.  However, their help simply consisted of asking for a stack trace (there was no exception, so that didn’t help) and then suggesting to try using different lifecycle methods to interact with the ad.  The detailed technical information I provided (including the code I was using) was simply beyond what the technical support team was prepared to assist with.

After several hours of testing over a few days, I decided it was a waste of my time to investigate further.

A note on SDK version – I tested up through 1.9.11.  At the time of writing this article, I see that 1.9.12 was released just a few weeks ago.  I may test the new version to see if it fixes the bug I encountered, but the change log doesn’t list anything to give me hope.  That doesn’t move AdColony any further up my list of priorities.

Time to move on

And with that… I don’t see the value in throwing more time into a black hole.  I lost a number of hours dealing with the test ad campaign as well as attempting to get the buggy SDK to work.  I’m willing to float ad networks a little credit to validate their claims, but all signs point to this not being a good network to use.

So, I don’t suggest working with AdColony if you are an independent developer.  Large companies have the resources to burn getting a buggy SDK to work, but I certainly don’t.  There’s a limit to the number of hours smaller development house can allocate to dealing with ads before it eats away all potential income from those ads.

My general rule is that trouble integrating an ad SDK is a warning sign, and should make you run the other way.  Top-notch ad networks tend to make it easy to use their SDK.  I’ve reviewed plenty of other ad networks on this blog and in the Android Ad Network Primer book that are easy to get set up with, so unless you have time to waste I suggest avoiding AdColony.

A lot of people ask about seeing real revenue numbers, and I understand why. Seeing the actual amounts takes it from theory to real, measurable data.

It took a little longer than I expected to get it out, but I think you’ll like this special report.

I show exactly how much the app made from sales, ads, and in-app purchases (IAP).  I compare daily download counts as well as IAP conversion rates, and the end result is some very useful information to help me (and you) make decisions for future apps.

Subscribers to my mailing list should already have it, and if you aren’t a subscriber yet, just sign up now and you’ll get the report too!

Crunching the numbers has given me some valuable insight into where my time has been well spent, and where I should probably scale back my efforts.  The results surprised me a little, even though I see the raw numbers all the time.

It’s not a factor of a few percent.  I found that I’ve been spinning my wheels on some of the more difficult – but less profitable – places I could be spending my time.

I’m working on the final details now, and will send the full report out to my mailing list (the “ProjectJourneyman BrainCast”) next week. If you’re not on the list yet, sign up now in the box below!

Update – the report is out, and available to list subscribers.

The challenge

As I mentioned in my previous post, I added in-app billing to my Pumpkin Carver app, and I upgraded the paid version to make it more appealing.  Then I charted the progress on various app stores over October, when Halloween should drive the most attention to the app.

Income from ads, in-app purchases, and sales can vary greatly between the different markets.  Not all platforms even support all three sources of revenue.  This means it takes a bit of number crunching to show the whole picture for each app store.

Crunching these numbers is time consuming, so I don’t do it for all of my apps all of the time.  When I do, it provides great information to help me make decisions.

The Surprises

I knew that Nook has been a strong performer for me, but it really outshone the rest this time.  I wanted more useful information to help with my other apps though, so I kept digging.

BlackBerry has waned as a source of income this year, but Samsung has made a strong entrance.

Another very interesting result was just how well Amazon Appstore in-app purchasing performed against Google.  I’ll crunch the daily conversion rates in the report too – there’s a start difference between the two markets, and it has solidified my intent to support Amazon in-app purchases in the future.

The full details

There’s a huge pile of data beyond those few insights above, and I’m really pleased with out much information I have to work with here.

As I mentioned, I’ll be sharing (update: it’s out!) a more complete writeup and analysis of this experiment on my email list, so sign up now to be sure to receive it.  I know I’ll be making decisions based on this information, and I hope it will be helpful in your planning to maximize revenue from your Android apps.

There are lots of ways you can get delayed.  You can get busy for a while and never get back to the project.  You can lose motivation or belief that the app could succeed.  Maybe you get frustrated by unexpected challenges and progress becomes stalled.

Or, perhaps you’re the ultimate planner and are working out all of the details before you get started.  This can be the biggest sinkhole of all – you can always do more research, and forestall releasing your app yet another day.

The cost of waiting

There’s a hidden cost in waiting, however.  It’s not just missed days of your app being available to buy or download.  The problem is also that you continue to move forward without the valuable lessons that you will gain from releasing your first app.

If I had waited until I finished my great opus before releasing anything, I wouldn’t have any apps out yet.  My first app idea is still maturing, but in the mean time I have released a handful of apps of varying degrees of success.  Each one of them shapes what I realize I must do to make my big apps successful.

Also, there must be a point when you transition from planning to doing.  At a certain point, you can spend more time learning about and planning to do something than it would take to just do it.  Factor in the cost of your time, and it should be clear that you can’t just keep planning and researching.  Your time and money will be better spent building something.

The final problem is that motivation has a shelf life.  If you put your project on the back burner for too long, your enthusiasm may fade beyond recovery.  At that point, a project that you may have been able to do an amazing job with may be something you never return to finish.

What to release

Ostensibly, the only real cost of releasing your app is the $25 to become a registered developer with Google Play.  That lets you deploy as many Android apps as you please, with no renewal costs.  That’s downright cheap compared to the amount of your time you have likely spent building or planning your apps.

Of course, rushing to release your big idea could make it less successful.  I’m not suggesting that you do a rush job on your best idea.  I suggest that you pick an idea that you’re exited to do, but is as simple as possible.  The point is to finish this app and go through the full app life-cycle.

For me, getting that first Android app released into the wild and getting downloads was all the motivation I needed to take things things to the next level.

While the Nook test team will verify that your app is compliant, you should probably take a more proactive role if you want your apps to be available for the new Nook HD devices rolling out very soon.

First of all, the new devices are running ICS (Ice Cream Sandwich – Android 4.0).  This brings in a number of more modern features of Android, including widgets.  In addition, Nook reversed a few of their customizations used for the Nook Color and Nook Tablet, following the standards a little better. Unfortunately, that breaks some apps that have been customized for Nook already.

Reverting Nook Color and Nook Tablet customizations

Things would be easier if I hadn’t made customizations to handle the previous two generations of Nook devices.  Don’t get me wrong, though, I still want to support all versions of Nook to maximize my exposure.  Supporting older devices is always a business decision, but I suspect that the older Nook tablets may hang around for a while.

Two changes that stand out to me are the location of the external storage (/mnt/media is no longer accessible), and the wallpaper mechanism is now the standard one used by Android.  If you use either of these currently on Nook, they won’t work with the new devices.

Beyond that, the higher resolutions need to be handled.  Since the Nook devices use a little screen space for the on-screen menu bar, you will want to verify your layouts on the new Nook devices if possible, even if you already have done testing on ICS devices.  If you can’t do that, at the very least load up an emulator for the Nook HD and ensure the app works flawlessly there.

Read here for more details on the new developer requirements for the Nook HD.  The Nook team didn’t really emphasize the details of the changes very well – I guess they figured we’d all scour the documentation immediately.  I personally skimmed the changes and figured I was close (I already supported higher resolutions and Android 4.x).  The problem for me lies in the Nook-specific changes.

Determining which Nook device you have

While I previously used my market flag (since I created a Nook-specific .apk) to enable the Nook-specific code, I need to make updates.  I now need to know which device line the app is running on before deciding which code to execute.

The Nook team now recommends using the Build.MODEL constant to decide which device you are on.  They haven’t released the PRODUCT field values for the HD tablets yet, so stick to the MODEL name for now.

It’s not a difficult check to perform.  In fact, I just updated the Android Market Manager library to add methods to check if you are on a known Nook device, as well as which Nook product line (e.g. Nook HD or HD+ vs. Nook Color or Tablet).  Go ahead and grab the code for your Nook apps.

Avoiding problems with Nook Updates

I was in the process of submitting an update right as the HD roll-out was happening, so I hit a few snags with deployment.

I suggest releasing your app updates separately before the HD update, if you have them ready and want to speed up release.  In my case, I want to get the new Pumpkin Carver Pro update (which coincidentally added HD to the name, but wasn’t fully tested for the Nook HD devices) available for all of October.  I hit a few rejections because I tried to target the HD devices and the UK as well, all in one update.

If your app hasn’t been reviewed and approved yet for the HD platforms (they did an initial review of all apps and sent an email to developers), don’t set a price for the UK in your next update – your app will be rejected.  You probably shouldn’t check the box for the HD devices either, as this may slow down this release.

On the other hand, if your primary purpose is to update for the HD devices, by all means take care of it all in one go.  Just make sure you address any concerns noted in your HD test email.

Getting ready for HD

Now that you can detect if you’re on one of the HD devices, and you know about the changes you need to make, get your apps ready!  The new devices will likely bring a lot of downloads, and you want to be there for that.  Plus, the new UK market will only sell your app if it is for the Nook HD or Nook HD+!

This year, however, I decided to try a few  new things.  I’m always about trying new tactics, especially since there are so many things to explore in the world of Android app development.  Hopefully these new experiments will improve income from this line of apps.

In-app purchasing

I have implemented IAP, or In-App Purchasing before, but not in my own personal apps.  I decided this was a great opportunity to try this out.

And of course, since I never just focus on Google Play, I implemented both the Google Play and Amazon IAP libraries. This will be a good opportunity to see how the two markets compare for in-app purchases.

In fact, the relatively small user base of these apps (and the seasonal appeal of Halloween) may not be large enough to bring in great results from in-app purchasing.  It will, however, give me insight into how Amazon and Google in-app purchases work, and tell me where to focus in the future.

I considered also doing the Samsung Apps IAP as well since that market has been treating me very well.  However, their library required a few extra permissions (such as SMS billing to support other payment methods) which means I would need to create a separate project to deploy the app there.  I work hard to minimize the number of projects I need to create for one-off deployment targets.  I’ll see if the high traffic from Samsung will justify the effort.

How did I handle in-app billing for both markets?  I implemented both billing solutions in the same project, and used the same techniqe as the Android Market Manager to easily build two separate APKs for Amazon and Google.  Each APK will execute the appropriate billing code when the user initiates a purchase.  It wouldn’t be too hard to add additional billing libraries, but the extra permissions are where things get sticky.

Going High-Definition to target newer Android tablets

I also updated the paid version of Pumpkin Carver to be high-definition.

With new content and high-resolution images for all of the pumpkins, the newly retitled Pumpkin Carver Pro HD rings in around four times as large of a download as the previous version.  I was a bit hesitant to increase the app size, since many believe that larger apps get less downloads.  However, even at almost 4MB this app isn’t really in the big leagues for download size.  Plus, the data on decreased downloads was for free games, and we know that paid apps are perceived differently.

The HD (High-definition) angle was already in the works, but seeing the launch of the Kindle Fire HD (and the larger Kindle Fire HD 8.9″) as well as the upcoming launch of the Nook HD and the Nook HD+ has convinced me that these commercial giants have decided that 2012 is the year of High-Definition.  And if consumers have an “HD” Android device, they’ll want content to match.

By the way, if you didn’t get a Nook or a Kindle Fire previously to test your apps with, now is a good time to get one – at the same prices the old ones launched for.  I already have an older tablet in each line, but I should update soon because emulating the extremely high resolution of the new HD devices in the emulator will be a PAIN.

Also note that Pumpkin Carver Pro on Nook has done better than on any of the other platforms.  I would welcome a change on Google Play, but for now the free app is the stronger performer there by far.  I still release both versions to most platforms, but put less attention to poorly performing platforms.

Enjoy Halloween

Hopefully I’ll have positive information to report here after October.  Also, if you have any related experiences or suggestions, share them with us!

And if you have Halloween updates for your Android (or other mobile) apps, get them out now!