This is a two-part series on iPhone analytics explaining how analytics works and what it can do for a developer who wants to integrate it into their iPhone application. Part 1 covers analytics basics and explains what information a developer can get and how it might be useful. Part 2 covers the important issue of privacy and what a developer needs to be aware of when collecting information from users.
IntroductionYou may have heard a lot about iPhone analytics but were unsure exactly what it was and what it could do for you as an entrepreneur. In this post I’ll explain how a developer can integrate analytics code into their iPhone application and the reasons why they would do so. I’ll also cover some scenarios where analytical tools can provide a developer valuable information that will enable them to plan long-term strategy and make critical business decisions.
First of all, I’d like to begin with a little background. A while back I interviewed Pinch Media CEO Greg Yardley about Pinch Media’s analytics platform and how it helps developers. Therefore, a lot of my information is based on Pinch Media’s tools, and I’d especially like to thank Greg for spending the time talking with me about the nitty gritty of iPhone development and analytics. Developers looking for more information can find it at
www.pinchmedia.com.
What is analytics?
Basically, its code that you insert into your iPhone application that collects information or tracks actions a user would take, much like a web analytics suite tracks hits and clicks. The application stores the data locally and then transmits it back to the analytics company where it is aggregated and annonymized. A developer can view the aggregated data, often using charts and tools developed by the company providing the code. The amount of data transmitted is small and Pinch Media CEO Greg Yardley says their software determines signal strength before needlessly trying to send data. If the software determines the signal is not strong enough, it locally records the data and sends it at the next available opportunity. A developer can add code to track user actions, location, events, and time between events, among other things.
What can it do and how will that help me?
The Basics: Estimating Sales and Usage Stats
At the very least, implementation of an analytics suite can help a developer estimate sales and track the number of users for each application they develop. While Apple now releases daily download stats, an analytics suite may help a developer track an application’s use more closely on launch day or at other important milestones. Some developers may be fine with just knowing their daily stats, as they will be paid by the download. For others, however, actual use or frequency of use may be more important than downloads. Pinch Media’s suite sends data back once per application use if the data connection is strong enough, and the statistics are updated hourly, allowing a developer to track usage over time.
Stats by Location
If the user opts in, an analytics suite can also track a user’s location, however this information is never reported individually but in the aggregate form. That means that while a developer can track how many users an application has in each part of the country or world, they cannot track where a user was at a specific time. Developers can get a report of region information allowing them to determine how many users an application has in each part of the world. Pinch Media’s software can generally estimate which city a user is in or nearby.
Location tracking, coupled with the features above, can determine how frequently users in each region are using an application. A developer could use this information to determine where to run an ad campaign or where to concentrate resources. For example, a developer of a restaurant review app might notice that it has gained a lot of users in one or two particular cities and could add new content accordingly. Without this information, the developer would not know where users are concentrated and might spend time updating content that won’t be needed.
A developer could also estimate the number of users an application has as well as the number of time an application is opened. A developer armed with this information can measure how “sticky” their application is by estimating the number of uses per user. Does your app have power users opening it several times a day? Or just casual users opening it once a month? High average usage per user can be useful information to developers looking to monetize and determine long-term strategy.
Event Tracking
Another great use for analytics is to track designated events. A developer can keep a count of how many times an event takes place, the time between two events, or even a series of events. A developer can track how many times a button is clicked and the elapsed time between two events. One purpose would be to track how long it takes a user to complete a task or how long they keep a specific part of an application open. Developers can determine which part of their applications are used frequently and which features might hardly be used at all. This might be extremely important for developers of enterprise applications because knowing this information ahead of time can allow a developer of a large application to save hours and development costs. A developer can choose to build out or improve the features that are being used the most and could even have a critical advantage over competitors by knowing ahead of time which features appear to be critical and those that are being ignored.
Game designers will find event tracking especially useful, as it can help them balance game play and determine proper difficulty levels. For example, if users commonly struggle to trigger the event of ending level 3, a designer will know that level 3 might be too hard. On the other hand, if they streak though level 3, a developer will know to make it harder. Even better, if a level that should normally be difficult suddenly starts showing drastically reduced times, a designer might suspect that an exploit has been found that they will need to take care of. Using event tracking properly would be useful in multiplayer/online games where making sure the gameplay is fair could be critical to adoption rates.
User Interface Testing and Improvements
One incredible feature of event tracking that came up in my interview with Pinch Media CEO Greg Yardley was using a user’s actions to improve and test user interfaces. For example, you could track how many times a button was clicked on or how often a settings page was accessed. Are users constantly changing the same setting over and over again? Maybe that option should be featured more prominently. You can also track how long it takes people to get through certain parts of the interface, like a form, or how long it takes them to complete a task or get through a set-up process.
While it’s not advisable to change the interface on your users all the time, information gathered over time can provide the impetus for a refresh that better reflects how users are actually using your app. Using ad-hoc distribution, developers can use event testing with a Beta Testing group to refine their UI and user experience before putting the application in front of actual customers.
What revenue stream should you be using?
Knowing the proper usage of your application could be important in determining a proper way to price and distribute your app. For example, an application that is opened and used a few times a month might fare better as a paid download, while an application that is used many times a day could benefit more from using an advertising model and being distributed for free. Developers looking to test the waters with new models might first want to institute an analytics suite to accurately gauge how their application is being used. It’s not too out of line to say that actual use of an application can differ from a developer’s perceived use, and double checking your numbers before giving away your application is just a smart move.
Great article. Analytics are going to become more and more crucial as new business and revenue models develop around iPhone apps.
Posted by: DJ Burdick | August 28, 2008 at 09:24 PM