Friday, May 13, 2011

Monetizing Phone 7 Application Methods and Approaches

The past 6 months have been a bit of a wild ride with Phone 7 going live and all the apps I’ve been working on. I guess I really started preparing for phone 7 the end of 09 heading into MIX 10 where I started building apps for the upcoming release. Going from 0 to making money with WP7 apps was kind of a long journey and this is my story, the story of Simon Said and my phone 7 obsession…

For several years there was this thing in the Silverlight Insiders (Super Secret Kabal of Silverlight Universe) about the upcoming new version of the MS phone/mobile platform and many of us knew about the two device directions and in fact knew people on the teams including the super secret design stuff going on in downtown Seattle. This kind of cloak and dagger thing drew our attention but most of what we had to work with for the longest time was conjecture, rumors and half truths. Thank goodness I wasn't a mobile 6.x MVP as I feel they really got the short end of the stick but the Silverlight MVP's had an awesome deal. Coming into MIX10 we had mostly all jumped on board and were building apps albeit for the longest time there were two few prototypes phones we could get our hands on that were weak and most of us were stuck with emulator which was horribly weak as well. (Currently we are still eagerly waiting for Mango, which is promising to be more awesomeness.)

Don't get my wrong, the Silverlight/WP7 team was working over time to get the new phone platform to market so the fact we got an emulator at all was great. Of course it wasn't long before this was hacked and we started getting our fingers into the OS and post the excitement of MIX11 the Dev community started to pick up momentum around the phone and building phone 7 apps. But the question in every bodies mind was will it actually make us money? and how could we be successful?

For the longest time no one wanted to talk about this aspect of the phone much, info was few and far between and hardly anyone knew anything or would tell you anything. Now that the market is open and we have been able to see some of the dynamics in action and some of us have made some money I'm hoping to help bring at least some of my experience to others. Let's start with understanding the basic revenue models that you can do on the phone.

Model 1: The Paid App

The paid application is the most basic model for making money. 'Free App's' do not constitute a revenue model by themselves w/o a more complex model. The idea of buying something, using the existing structure put in place, in a onetime fee sort of way becomes the most basic structure. On the good side this model is the most simple, you need only have a developer account ($99 typically albeit there are MS rebates if you get 2 apps into the market place), the tools which are free and an application. Issues with getting things approved aside, it’s just not hard to do but most of the people I know have not made money at this (at least not a significant amount). For the most part given the current market conditions and the expectation of free apps that developers have themselves propagated by educating the target demographic about the evils of paid apps has made this model difficult to use. The success that I’m aware of with this model have come with primarily with strong 'existing' brands and lots of marketing (any Xbox Live game as an example).

Model 2: Paid w/ Trial

The paid with trial model for making money is really just a variation on a theme from the first model and is only marginally more successful. For the most part all the examples I refer to about Paid apps is basically this model. So unless you have a good Xbox Live company you work for… its not likely you will be successful. Now that certainly can change as the market changes which will happen as the Windows Phone market becomes increasingly viable but not yet.

Ok so how I can Make money with phone 7 apps?

Model 3: Free w/ Advertising

The idea here is that an application is free to the user but you basically sell ad impressions to an ad service such as Microsoft AdCenter and show the ad’s in your application. Companies pay firms for the said advertising. The best part of this model is that it works. A good, well designed application that is targeted to a general demographic will work great. There are problems and complexities to this model but it is achievable to the average developers... When you are talking about AdCenter it really comes down to targeted ad’s. The better you can target your app’s ads the better your 'ecpm' value is which translates into more money. 'Ecpm' basically standards for how much a 1000 impressions of an ad are worth in your application. I’ve tested this and certainly the more you can tell the AdCenter control about your user and the more your ad’s are something that would interest your users the more money you make.

Model 4: Free w/ Paid Version

This model is basically a way of driving sales to a paid application. I’ve seen this one tried but with little success so far. I suspect this might become more viable as market dynamics/conditions change allowing more sales of apps to take place as a larger market share is formed around Windows Phone 7.

Model 5: Free w/ Advert w/ Paid Version

A forth model is a variation of the third. This model really doesn’t make a lot more then model 3 but I think it is well positioned to take advantage of how the market dynamics will change over the next couple years. What this model is, is two applications like model 4 but focused on advertising in the short term as revenue stream. In this model you have an application that is free with advertising that is also driving sales of a paid version. In most cases where I’m seeing success is when the free version is very popular with a built out feature set and not handicapped but at the same time the paid version looses ad’s but also might ad some premium feature. A simple example is that I have a morse code application in the market place that is free with a rich feature set from saving code blocks, sending and transmitting code etc. The paid version of course is w/o ads but also has a complex sound transcoding feature but for the most part the feature set is the same short the one. Then the free version doesn’t expire, is a good free app but also passively drives paid sales. Have a free app tends to drive many, many more eye balls and downloads.

Model 6: In App Purchasing/Purchase Upgrades

Ok here is a six model that can be used with the others and that is in app purchases. There is no market place support for this but if you have the ability to roll your own certainly it can be done. What this means is that in your app either paid or purchased the users has the ability to pay for additional elements, features or other purchases as defined by the application. Basically if you want to write the code you can do this but there is no current support for it however there is a nifty hack I learned about primarily it looks like this… You have this app, and then you have other apps in the market place are packs or expansions for the first app. These ‘pack’s give you a code or other set value that allows you to ad more elements to the first app, maybe some under the covers sync etc. A cool idea but really a bit of a hack and could be cryptic for users.

Case Study: Of a free app?

I have this one application in the market called 'Simon Said' that is now in the top 100 most downloads apps. This started off some years back with a demo project called 'Silverlight Simon' to show how to build composite WPF/Silverlight assets and controls. 'Silverlight Simon' being the case in point. Over the course of several years and 2 mvps 'Silverlight Simon' was in Silverlight, WPF, Surface, Azure, OOB and then the phone came along so I took the existing Simon game control and put it into a Phone 7 application shell and submitted it to the market place just to see that it could be done. At the time I didn’t really think it would be big but I did add the AdCenter control as I was trying all the basic models I could think of. ~3.9 million impression’s later and ~150,000+ downloads I’ve learned a lot about phone 7.

At first people would download the app and get upset that the animations lagged or the timing was slow but I was just focused on testing so Ignored this as the app description talked about this only being for testing and demo purposes but I kept getting issues and less than stellar ratings. I finally realized that there were tens of thousands of people downloading the app so I finally capitulated to building a new version optimized for the phone.

Now the Simon Says is basically a picture of Simon, 4 paths and a text box plus an app bar and the ad control and some code. The highly complex XAML that made of Simon needed to go and that turned out to be 90% of the problem. Most of the other features are implemented as separate views and a combination of a good design, well built app, a degree of familiarity and the AdCenter Ad control and the app being free has continued to drive the apps popularity as one of the top apps in the market place. Of course some guerrilla marketing helps too, a few blog posts, videos etc and its all good but as of late, if you’re a one man or women shop, free with advertising is the way to go.

Thursday, May 12, 2011

A Letter to Some Listing to Silverlight Haters...


I guess up front this topic is one of my hot buttons but I’ll try to be as even as possible in my explanation. But in fairness I do have a vested interest in Silverlight as a Silverlight MVP for the past two years, a published author and professional speaker as well as being one of the Silverlight insiders at Microsoft… besides having a vested interest it also puts me in the know as it were.

First WPF or ‘Windows Presentation Foundation’ to say is dead would be on par as saying Windows is dead as WPF is the primary method for building native applications on Windows. VSTS is in WPF, most other ms products are written in either WPF now or are C++ written in VSTS which is in WPF. That notwithstanding, there hasn’t been as much excitement about WPF as most of the industry talk has moved on to other things, even the last version of WPF was more or less a support release. Where all the talk in the ms world is as of late around Kinect (oh but all of these apps are WPF or C++), Phone 7 (which is XNA/Silverlight) or HTML5 (which the browser is written in WPF on windows) etc. WPF is therefore for native Windows Only applications. WPF primarily consists of XAML markup and C# code.

Silverlight which is also XAML markup and C# code, is a cross-platform lite version of WPF (Silverlight was first called WPF/e), it is cleaner with lighter overhead and runs on Mac, Linux, phone7, Symbian, CE, windows etc.

Last year ms decided that the HTML standard moving to 5 was a good thing and that they would support it for cross browser applications but that was never what Silverlight was about. During PDC last year (MS Professional Developer Conference) a former ms employee that at the time was senior vice president… Bob ‘what’s his name’ was interviewed on TV and made some less then well thought out comments about Silverlight which started this silly-ness about Silverlight being dead and ms has been doing damage control ever since. In Dec they announced the feature set for Silverlight 5 coming out later this year and demo’d features like pinvoke, native 3d and hardware acceleration even on the mac and XNA support built in. Silverlight is also the primary method for building apps for Phone 7 which ms is basically bet the company on. They will make it work which means supporting Silverlight for the foreseeable future. Silverlight is for building immersive cross platform user experiences and high end video, LOB apps that are cross platform etc. Developing Silverlight is for the most part the same as WPF, if you learn one, you can do the other easily.

One last point is that I work for an interactive design agency of almost 200 and we are primarily an Adobe shop. We do things like the Nike Touch wall or the Jordan retail kiosk, or the Microsoft Software Kiosk in the new MS Retail Store’s etc. Our clients don’t care about technology and projects should generally never be about the technology but the best way to deliver the Best Possible User Experience to the target demographic. I believe in this more than anything and is also way I helped start ‘Interact Seattle’ – designer developer interaction group. We meaning ‘[wire] stone’ deliver most of our work in WPF or Silverlight as our time to market is double our competition over say Flash which used to be our primary delivery model for interactive designs. Silverlight allows us our dev teams to build almost directly against illustrator files so that we are able to not have a design team or a dev team but a single team working together. In short as mentioned it cuts time to market in half once the dev’s and designers learn to work together and the designers can continue to use their tools they are used to. Therefore for business reasons we continue to use Silverlight…

I hope that helps :) please feel free to ask any questions and if you are in Seattle feel free to come by our office.


David J Kelley
Principal UX Architect, MVP
[wire]stone | Blog/Site | {Interact} | SeattleSlug
twit: @DavidJKelley