Thursday, January 29, 2015

What makes a better ad for your app?

Do you advertise your app? Have you integrated cross promotion of your app with AdDuplex?
How did you come up with the copy you use in your ad? Have you tried testing different versions of your ads to see which is more effective?

This is just what I've been doing with a number of developers recently. We've run two versions of an ad for one of each of their apps to see which is more effective and what we can learn from that.

I'll post the full results of the experiments soon but for now lets try a little test. Consider these two options.

Option 1
Option 2

Which option would you choose?
Which option do you think more people would click on?

Let me know what you think, in the comments or on twitter, and I'll tell you which did better soon.

A bleak outlook for Windows Phone?

Tomi is the definitive expert in the mobile industry.
Here's what he recently tweeted about the outlook for Windows Phone.

Draw your own conclusions.
Make your own plans.


Wednesday, January 28, 2015

What developers should take from the monthly AdDuplex reports

As they've been doing since 2013, this week AdDuplex released their monthly report showing some statistics on Windows Phone device usage.

Here's the most important thing it contains for developers:

Most people (more than half) who use a Windows Phone use a lower spec device.

However, in my experience:

Most developers who use a Windows Phone have a higher spec device.
And, most developers only have one device

There's a disparity here. Does it matter that developers are using devices that are, potentially, very different from the ones their users are using? I think it does.

To ensure an app works well on a lower spec device (or one with a smaller screen - if you have one with a large screen) it's important to test on such a device.
The emulators are good but real devices are better.
If you don't have a low spec device, find someone who does and ask them to let you borrow it.
Alternatively, and fortunately, the lower spec devices are cheap. If you're building apps to make money then it's a small investment to buy one.

Very few developers have access to a wide range of devices. A table like can be seen below is not the norm. Don't be disheartened though. For the most part, as a developer, you don't need this many devices to test on.
My suggestion is that if you're getting devices for testing and you can only afford two, you should get one with the highest spec and one with the lowest spec available. A Lumia 930 and a 520 are probably the most appropriate options right now. If you can get more than this, then I'd consider something with a larger screen, like the 1320 or 1520.

How many devices do you test on?

Tuesday, January 27, 2015

Why you should always set an end date on your advertising campaign

If you've ever advertised on AdDuplex you've hopefully had an email that contains the following:
Thank you for advertising on AdDuplex!

Your AdDuplex campaign has come to an end. Please visit the client area to extend the campaign or start a new one.

Why is this important?
It's important because you'll get an email like this when you reach the end date of your advertising campaign.

So, why is having an end date important?
If you don't set an end date for your campaign it will run until you run out of credit. (If you later top up your credit without disabling the campaign it will start again automatically.)

The real reason though, is so you don't forget about your campaign.

You should not just start your campaign and forget about it, or ignore it.
At the very least, setting a short end date will mean you get an email telling you the campaign has finished. When you receive this you can go and look at the statistics of the campaign, see how successful it was, make appropriate adjustments and then restart it.
Of course you can analyse and adjust before the campaign ends but I find that having an end date set acts as a useful buffer in case something comes up and I forget to check.

I'm guessing that you'd consider yourself a developer more than an advertising expert. As such I think it's fair and reasonable to assume that the copy you write for your advert(s) won't automatically be the best converting copy it could possibly be. As such you'll need to experiment to find what works best for you.

Don't just create your campaign and leave it. Experiment with different copy. Try different  value propositions and calls to action. Try things out. Measure what works and evolve your adverts.
You should be able to improve your CTR with relatively little work.

Some more on this soon....

Wednesday, January 21, 2015

Create Android apps with Java in Visual Studio - my logic

I just posted this.
Here's my logic:

  • Microsoft want to get people building Android apps to build Windows apps. Getting them on the OS and using the tooling is a good start.
  • If building in Visual Studio, on Windows, it would be a smaller step for Microsoft to make to get them to build for Windows. 
  • If VS was extended to build Android apps in Java, They'd be one step closer to enabling building Windows apps in Java.
  • Visual Studio is a shell that can be "easily" extended to support different languages, etc.
  • VS 2015 already includes an Android emulator. - Did they really just add it for HTML5 based apps? - We know how popular those are ;)
  • VS tools are widely considered as being the best IDE in the industry. Tooling for building Android as is widely considered as being some of the worst. I think it would be an appealing proposition for developers building Android apps.

What about Xamarin (and their partnership with Microsoft?
I don't see an issue. They're both targeting different audiences. Xamarin are appealing to those who already have .Net skills and helping them build for Android and iOS. This would be about getting people coding in Java to build for Windows.

And of course, I wouldn't speculate on anything like this if I knew even a tiny fraction of it to be true or related to any NDA I may or may not have signed.

If any of the above turns out to be fact then it's because of my knowledge, insight and, most importantly, luck!

Your thoughts?...

Create Android apps with Java in Visual Studio?

It seems there's a lot of wild rumours and speculation about what's coming in Windows 10 and how Android apps may or may not be supported.

I think the adding of support for writing Java to create Android apps inside Visual Studio is actually a lot more likely than some of the things I've heard and I could probably make a good business case for it too.

It should go without saying but for the avoidance of doubt. Of course I don't really know anything about what may or may not be coming. This is just my speculation. If it turns out to be true you can be as surprised as me. ;)

I'm looking forward to announcements .... but not expecting to hear anything related to actually building apps for Windows 10 until build.

Update: my logic

Monday, January 19, 2015

Windows (Phone) app analytics

I was recently asked what the "best" analytics solution out there is for universal apps.

Unfortunately I haven't reviewed what options are available recently. In the work I do for other people they already have made the decision about what analytics solution they want to incorporate.
There are lots options available and what you'll want to use will depend on the actual analytics you want to gather and what you want to do with them.

Here, in no particular order, are the 3rd party app providers I'm aware of:

There are a couple of other options too:

If you're using an ad solution that guarantees 100% fill rate then the number of ads served can be a good indication of the number of people using the app. It's the most basic level of analytics there is, but it comes for free. ;)

The last option (and the one I'm integrating into what I'm building in my rare moments of free time) is to integrate the analytics into the back-end of the application. Doing this is more work but has a few advantages:

  • No extra requests need to be made by the app
  • No managing a 3rd party component and service
  • Able to get exactly (and only) the information wanted 
  • Much easier to integrate with CRM systems
  • No need to reconcile data from a 3rd party with own data

Thursday, January 15, 2015

The secret of mobile app/game success

When it comes to mobile developers and achieving success many don't do anything and are surprised when they don't become overnight millionaires.
They also often don't have a good answer when asked what they've done to try and achieve the success they're after.

Here's the secret to success:

Test. Iterate. Optimise.

Experiment, change things, improve.

It sounds simplistic but it takes time and discipline.
If it was easy, everyone would do it.

Wednesday, January 14, 2015

WinApps London (& WPUG) 2014 in review

I'm not great at keeping stats about the user group I organise but I just had an email from Eventbrite with some summary stats and after a quick check on them I thought they would be good to share.

In 2014 we had 12 meetings with an average attendance of almost 60 people.
Even more encouraging is that there was a general trend towards an increase in numbers at each meeting.
As it feels that, generally, there's been less interest in building apps for Windows and Windows Phone in the last few months, that we've been growing in numbers during this period is something that I'm very encouraged about.

If you want to know more about our past meetings you'll find them at You'll also see details of what's coming in the next few months on

Tuesday, January 06, 2015

We've reached the point where mobile developers shouldn't have to think about battery usage any more

I realize this is potentially controversial but hear me out.

Ignoring the devices with massive batteries that can last days.

  • Batteries in devices are better than they used to be. 
  • People have learnt when they can't use devices all day without consequence to the battery.
  • Because of this people have learnt to manage their usage so that they get through the day without running out of power. After all its them that suffer if the battery is drained.
  • People do this through a combination of managing their usage, recharging during the day and disabling radios and background tasks that they don't want or need.

So, if people have learnt to use their devices so that battery life isn't an issue, I think we're at a point where developers don't need to think about optimising their code to minimize power consumption.

Yes, some apps require intense processing or network activity that requires extra power consumption but this shouldn't be an extended event.
Also, line of business apps can have extra, special requirements for preserving battery life and ensuring usability. They're a hopefully obvious exception.

Some games are particularly bad at this and this should clearly be avoided. If I play a game for half an hour and it uses up half of the power in the battery (naming no names) then something is definitely wrong.

Of course this all comes with a disclaimer. I accept no responsibility for bad reviews or other negative consequences if your app drains a person's battery.

If you're an app developer, do you think about the battery consumption of your apps?

(Hello those who came here at the prompting of @DVLUP)
This was also posted on DZone where discussion in the comments lead me to try and clarify some of the above.
My point is that there is probably little business value in spending time optimizing your code to be battery efficient. This makes two assumptions.
1. You aren't doing anything silly in your code.
2. You are managing resources that might drain the battery quickly appropriately.

Monday, January 05, 2015

Reflections on how the devices we have can affect what we do or think

Upon realizing that if I could do something on a PC in seconds but it takes lots longer on a phone and that's all I have with me I often just don't bother doing it at all.

5 things then:
  1. Could this be an indication of an opportunity to make something better?
  2. Could just be a fundamental limitation of small, occasionally connected devices - can, and how do we live with this?
  3. Is this an opportunity for something that can capture context, intent and motivation for "when I'm next at my PC"
  4. Why am I happy letting my device dictate what I'll do? Is it just a trade off of effort and reward or am I at risk of getting in a position where the device available becomes the overriding decision making factor?
  5. If you want someone to do something you need to make it easy for them to do that, regardless of the type of device they may be using.

Any thoughts?