Experiences of In-App Mobile Advertising

With my companies Funky Expenses Android applicaion we have two versions; An ad-supported version and a pay-for ad-free version so that users can choose how they help to fund development (yes, we do get requests for a free ad-free version, but unfortunately that won’t keep the company afloat). The ad-free version is available from Google Market (which currently has the largest user base by a long way) and AndAppStore, and the pay for version is only available in Google Market.

We’ve always used AdMob for the in-app adverts because they have an Android SDK, the adverts are not too intrusive, users seem happy with the experience (the app has an average of over 4 stars on Googles Market and the active installs count is in the low thousands and on the occasional bad day decreases by only a few users) , and we’re happy with the results (an eCPM of around 0.21 US Dollars). We looked at Googles Adsense for mobile applications, but the 3 month minimum participation and reported issues with adverts meant that it’s not something that we’re willing to use at the moment.

While the revenue generated from the adverts would not be enough to sustain the company on its’ own, the monthly income from ads is nearly twice the amount we receive from sales of the ad-free version, and so is a significant source of revenue for future development.

I suspect many people would see sales of the ad-free version as being the key to success, but there are several reasons why a larger user base for the ad-support version than the pay-for ad-free version is beneficial to our business. The main reasons are;

1) Regular income.

Once you’ve sold an Android app on Googles Market the users get free upgrades for life. This, to us, does not make long term economic sense. To continue developing Funky Expenses we need source of regular income, and that’s what the ad-supported version provides.

2) Problems in Google Markets payment system.

Google Market has, by far, the largest number of potential customers, but the payment system (based on Google Checkout) is far from ideal. For example; we can only list applications in British pounds (because we’re a UK company) and most of the customers are in the US and so are used to dealing in US Dollars.

Any salesman will tell you that when dealing with a customer you want to deal in a currency they’re comfortable with so you don’t end up with the transaction being disputed because something the customer thought would be one price actually turns out to be another once the bank has done the currency conversion.

3) Not everyone can access applications which you have to pay for on Google Market.

This means that if we focused our development and revenue generation on the pay-for version the number of potential customers would be far smaller, which is not what we want.

This then leads to the question; Why have a pay for version? Well, the answer is simple, it’s what users asked for. Some users don’t want ad banners and are willing to pay to remove them, and, to us, it’s better to have a customer on a paid version than to have lost a user.

Hopefully this post will provide some of you with some ideas for your apps, and provide others with some understanding of why in-app ads make sense. If you have any comments please post them below.

Is the iPhone 3GS a rip-off in the UK?

(If you want to tweet this please use the tag #O2iPhoneRipOff)

Is it me or is the new O2 iPhone pricing policy looking like a bit of a rip-off?

First off, there is the cost of the ‘phone. At Apples WWDC it was announced that in the US;

3G 8GB goes from $199 to $99.
3GS 16GB to be $199 (same as old price for 3G 8GB)
3GS 32GB to be $299 (same as old price for 3G 16GB).

From O2 in the UK we get price rises rather than cuts. For example on PAYG;

3G 8GB stays the same price.
3GS 16GB costs about £50 ($80) more than the old 3G 16GB
3GS 32GB costs about £150 ($240) more than the old 3G 16GB

and on Pay Monthly tariffs where you need to pay for the ‘phone you’re looking at paying an up front cost of around £55 ($88) more for a 3GS 16GB than the 3G 16GB, and £145 ($235) more for a 3GS 32GB than the 3G 16GB was!!!

This price increase is also reflected in the £45 per month tariff where the old 3G 16GB was free, now you have to pay £87.11 ($140) for the phone.

You might think that the extra costs cover materials, well, to put those prices into context, the difference in price between the 16GB and 32 GB iPhone 3GS (so you’re getting 16GB of additional storage) is more than enough to buy yourself an entire 32GB Solid State Hard disk for a PC or Laptop.

There is also a rip-off in relation to Tethering; On pay Monthly tariffs “unlimited UK data and Wifi” is included in the tariff, but if you want to tether your iPhone you have to pay £15 per month extra for a 3GB limit, so if you pay monthly you get unlimited data in your plan, but if you want to tether you suddenly get a 3GB limit, how, exactly does that make sense? (If you wondering Pay As You Go customers can’t tether, O2 just don’t offer it to PAYG customers).

If you compare the price to O2s mobile broadband tethering actually works out worse value for money than the mobile broadband offering. With mobile broadband you pay the same amount for the same data allowance and you get free Wi-Fi, but with the iPhone you’ve already got free WiFi on the monthly contracts, so you’re either paying for WiFi twice or getting 3GB of data for a higher cost than they charge mobile broadband customers, which again just doesn’t make sense.

If that wasn’t bad enough; the iPhone tariffs that O2 offer their pay monthly customers have an 18 month minimum term, so if Apple stick to a 12 month release cycle pay monthly customers will always end up with a “contract buy-out” or “6 month hold-out” choice as has happened to the current iPhone 3G owners on O2.

Come on O2, sort out your pricing policy!!!

Update – cheapest tariff : To top it all for the Pay Monthly customers, if you don’t use the Wi-Fi from O2 or it’s partners, the pay monthly iPhone tariffs don’t even offer the best value for money over 18 months.

If you buy a Pay As You Go 16 GB iPhone 3G S and put it on a £19.58 per month 12 month minimum term SIM-only simplicity tariff and add the free unlimited web bolt-on you get 800 minutes talk time per month, 800 texts per month (1600 if you buy on-line), and over 18 months the whole lot costs £792.84. If you compare that to the £34.26 per month iPhone Pay Monthly tariff which offers 600 minutes talk time (200 less than the simplicity tariff), 500 texts per month (300 or 1100 less than the simplicity tariff), and unlimited web, you end up paying £801.66 over the 18 months.

Update – 12 month cost: If Apple do go for a yearly refresh cycle and you really want to upgrade, the PAYG iPhone + Simplicity tariff will cost you £674.36 over 12 months, whereas with the iPhone pay monthly contract you’ll still be locked in to paying £801.66 over 18 months.

Update – O2 trying to defend itself : Apparently, to O2, a “one month minimum term means you can turn it on and off when you like”, it’s just a shame I’d use tethering with my GPS for 1 day out every week or two which means I would pay for 1 month and 3 GigaBytes of tethered access just for 3 or 4 days of usage and about 1 Megabyte of data, and to them “iPhone tethering costs more as it uses a lot more data than traditional browsing on your iPhone itself” – Did they miss the bit of their own web site which says the iPhone tariffs already have “unlimited UK data and Wifi” as part of them???

Will the Android Market T&Cs kill off “fun” apps?

Over recent weeks there has been a lot of speculation over when Android users will be able to buy applications from Googles Market, but some of the terms and conditions in the Developer Distribution Agreement could scare many developers away before they deploy their first application.

The first thing that came to light is that purchasers will have 24 hours in which they can get a refund apparently with no questions asked. This is a good thing for purchasers, but could kill the business model for developers working on applications which only have a few hours of enjoyment because users can buy it, get their use out of it, and return it and the developer will not make a bean from it.

You may think that it’s good to enforce this, but a number of iPhone developers release cheap short-lived apps to build their funds to develop more substantial applications and many games may only be designed to entertain for a few hours and are priced accordingly.

Following on from this came the discovery that, according to Googles Market terms and conditions “…Products sold for less than $10 may be automatically charged back to the Developer”. This means that if a user tries to return a product any time after purchasing it, and the application cost less than $10, the developer could find the money taken from their Market earnings without being consulted let along getting the chance to offer some help to the user, verify the problem, or issue a fix.

Comparing this to the iPhone application store Terms and Conditions (where 90+% of apps are in the 99 cent to 10 dollar price range) shows how weighted against developers Googles terms are.

The iPhone App Store T&Cs state that purchasers can ” withdraw from your transaction without charge and without giving any reason until delivery of the Products has started”, so customers can get refunds if they’ve mis-clicked on something, but they don’t get a 24 hour trial period in which they can just return the app, and they certainly don’t get the right to return a $9.99 app at any time and get a refund without the developers involvement.

In an ideal world we would all get our apps for free, they would do what we want, and we’d all be happy, but unfortunatley developers need to earn money, and so given the choice of developing an application for  a platform where if you’re app costs less than $10 you could see users refunded with any consultation with you, or a platform where once the user has the app running the developer knows that unless there is a serious problem they will get paid, I’m pretty sure I know where many developers will be going.

Update : After posting this Google announced that pay-for apps can now be listed, so we’re going to have to see what happens.

Android licensing system

Funky Android has made available a Android application licensing system via AndAppStore. It’s something that we feel has been missing from the Android space and we hope that it’ll help companies looking to focus on (and thus make money from) Android applications.

So if you’ve been waiting for a way to commercially license your apps feel free to give it a spin and send in your feedback.

(Declaration of interest : I am a director of Funky Android)

Google finally coming good with the Android Development Programme

It looks like the folks at Google are listening and they’ve taken a big step which will help a number of developers get at least one Android powered device to work with.

Yesterday Dan Morill announced a new version of the SDK with a link to the android blog, what he didn’t mention is that the blog post also reveals that Google have created a programme for developers which allows they to get what is basically an unlocked T-Mobile G1.

The Android Dev Phone 1 as they’re calling it will initially be available to developers in the following countries; US, UK, Germany, Japan, India, Canada, France, Taiwan, Spain, Australia, Singapore, Switzerland, Netherlands, Austria, Sweden, Finland, Poland, and Hungary. Google are looking to add to that list as and when it’s possible.

This is a great move and one worthy of praise, because it will truly kick start development in those countries by providing them with access to an “official” Android device.

The only people I can see losing out because of this will be those making ‘phones for those regions where T-Mobile aren’t providing the G1 (such as Koogan providing their Agora, and Koolu putting Android on a Openmoko Freerunner) where developers were buying these ‘phones to get their hands on some hardware running Android. It should also be made clear that Google is not trying to steal Koogan, Koolu, or anyone elses market entirely because the Android Dev Phone 1 will only be available to developers registered with Android Marketplace, and won’t be on sale to end consumers.

Now here’s hoping that the next big step is API equality for all apps to address my other big gripe with the Android strategy at the moment.

The Android Email client on the G1, possibly the worst app on an “Internet” phone ever.

The G1 has been around a few weeks, it’s billed as “The phone that’s built for the internet”, but it would appear that the G1 isn’t built for email which many people consider as being one of the most useful things on the internet for over 35 years.

The G1 has two in-built Email clients; one is for Googles’ GMail, the other is for accounts with whichever ISP you’re with. Personally after seeing several stories of problems with Google locking accounts (see here, here, and here for a few) I’d prefer to stay with my ISP, but that leaves me with an Email client which is being described by other G1 users as “the most pathetic attempt I’ve ever seen“ and “frustrating”, views which I wholeheartedly agree with.

My email account is hosted on an IMAP server, is around 300MB in size, and has a couple of hundred folders in about 4 or 5 levels. Over the years it’s been handled very well by everything from Outlook, Outlook Express, Thunderbird, Nokia and Sony Ericsson Mobile phones, Windows Mobile PDAs and PalmOS devices, but when I open my account on the G1 mail client all I see is one long list with names like “Company”, “Company\Invoices”, “Company\Invoices\Paid”. There appears to be no concept of folders within folders, just an attempt to force the GMail concept of tags onto the rest of the internet.

My experience of the email client have been so painful I actually prefer using my Sony Ericsson c702 with it’s numeric keyboard an far smaller screen just because I find it more productive. On the G1 I’ve been unable to delete an account which is no longer valid, unable to read PDF files people have sent because the G1 does not have a PDF reader, and when I try to access a mail account which doesn’t exist I don’t get an error; just get a button which says “Retry loading more messages”, which is not the best way I’ve seen of indicating a major problem.

So if you’re looking for a phone thats “built for the internet”, email is a big part of your ‘phones life, and your email isn’t GMail, you might want to think again.

—

[Notes]

Some third-party developers are working on better Email clients (k9 being one such example), my problem is that this is billed as an “internet phone” so surely should support one of the oldest uses of the internet straight out of the box.

It pays to talk…(or how I ended up buying a G1)

After talking to the in-store staff at the T-Mobile store in Maidstone, Kent, UK, I am now the owner of a G1 looking at a total cost of £360 ($580) which is about the price I would expect to pay for an unlocked G1

You may wonder why yesterday I was looking figures that were twice that amount, well, it’s a mixture of good fortune and good customer service. I explained my situation to the in-store staff and they explained they could offer me a “£20 per month for 18 months” contract or a “£30 per month for 12 months” contract (both of which have a total cost of £360), but I would have to pay more for the ‘phone.

I opted for the £30 per month tariff, thinking the ‘phone might be about £100 or £150, but when they rang up the cost of the G1 and it came up as £0… yes, free. The staff verified it with someone on the phone who agreed to let the sale go through.

So now I have a G1 at a price I was happy to pay, on a contract that won’t leave me stuck with an also-ran if new Android phones appear and gain market share in the next year, and added to that I am seeing reasonably OK 2G reception, it looks like most of my problems with the purchase have been resolved.

First Android security flaw to be publicly reported…

The New York times is reporting a security flaw in the G1, and to me it’s confirmation that Google left opening up the source to public far too late by doing it the day before users could go out and buy hardware with it installed.

If the source code had been available a few months (or even weeks) ago there would have been time for the collective eyeballs of the development community to take a look for problems like the reported one (which was found after a few days) and fix them before the device reached the hands of consumers.

So where are we now?, well, there are several thousand devices in the hands of non-technical users all of which have have a security flaw which could allow a third party to steal user-names and passwords for websites that those consumers visit… definitely not good for Android.

I can’t see this being the last problem found, and given that the first update was announced almost as users were getting their phones and it didn’t fix this problem, I can see a few more updates to come in the next weeks and months

The fact that the reported came to light so quickly shows the strengths of the open source process, but it also shows why the “Release Early, Release Often” strategy is the best way to go every time, and why companies such as Google shouldn’t keep the source code closed until the very last minute.

Expect to pay when you setup your pitch in Googles Marketplace

If you want to have your application listed in Androids Marketplace the first thing you have to do is pay them $25. That’s right, even if you’re developing a free application for their open platform to list on their “user-driven” marketplace, it’s going to cost you a $25 entry fee.

Do you have an app you’d like to to sell?, well, if you make a sale via the Marketplace, that’ll be another 30% (yes, thirty percent) of your sales going to them. The post on the developers blog says the 30% “goes to carriers and billing settlement fees-Google does not take a percentage”, but it was Google who made this arrangement, so they shouldn’t expect to hide from developers comments by pointing the finger at the other guys.

Now lets compare this to some of the other options; AndAppStore.com take 0%, the money goes straight to the developers PayPal account. SlideME say 3 to 5% based on the processing charge, and if the developer takes payments direct then it’ll all depend on they payment mechanism, but they shouldn’t expect to pay more than 10%.

Given that Marketplace already has advantages over competitors due to being shipped with the ‘phone it would suggest how Google persuaded carriers to take on the Android OS; namely giving a cut of all application sales to them. If you want to imagine how enticing that might be then think about being able to sell your house and get a cut from all the furniture the new occupant buys, or sell your computer and get a cut from all of the software the new owner pays for, then you’ll start to see why I think we may have just found the sweetener which made T-Mobile start singing the open song….

Android; The not-so-open open platform

If you’ve taken an interest in Googles Android platform before you’ll be familiar with phrases such as “Android is the first truly open and comprehensive platform for mobile devices” from when the Android and Open Handset Alliance initiative was started, and if you’ve attended talks by some of the Evangelists you’ll have been told that you can replace any part of the ‘phones software, but, this morning, it would appear that you won’t be on a level playing field.

It is true that the Android source code is available to everyone, and it is true that you can write apps which provide the functionality of the contacts app, dialler, Marketplace, or whatever, but what hasn’t come to light until now is that you may not be able to get your application to run on an Android device because some functionality will only be provided to applications signed with a code signing certificate which is specific to each platform.

This basically gives Google and the ‘phone manufacturers the ability to lock down functionality and make it only available to their applications. You might see this as a slightly out-there scenario, but there are already applications on the G1 which are using functionality which will Google have no intention of making available to third-party developers giving it an unfair advantage over any competing app.

We’re not talking about marginal functionality which will be rarely used (such as the ability to change the operating system and hardware configuration), we’re looking at things like Googles Marketplace offering seamless upgrades but any third party updating software being prevented from installing new versions of an application (and even if it could it would be forced to put the user through the permission granting system for every single update). In another post it’s mentioned that dialling emergency services isn’t available to third party applications, so if you were thinking of writing a trendy replacement for the in-built dialling application, well, you’ll just need to make sure your users know how to revert to the “approved” one shipped with the phone if they need the police, ambulance, or fire services.

So remember, although Apple may reject third-party apps because they duplicate functionality, if you’re developing for Android you may find that although you can write an app which offers a better dialler, browser, or whatever, you may find it won’t run on a device purely because of the way the platform has been designed.