In the past years, mobile applications took the world by storm and already changed the way we use the internet for work or leisure. Various technologies emerged to create mobile apps and development processes start to consider mobile as first class citizens. But even though mobile already seems to be omnipresent, the future is just about to start. We're facing new generations of mobile devices like wearables or lots of mobile gadgets that make up the Internet of Things. We will be confronted with new types of user interfaces for displaying data as well as accepting commands. And we will recognize more and more companies going real mobile first. All this will influence the way we design, develop and test software in the coming years.
This InfoQ article is part of a series on the fast-changing world of Mobile technology. You can subscribe to notifications about new articles in the series here.
Mobile channels are getting more and more important - not only for companies with a focus on technology, but also for traditional businesses. Consequently, the concepts of website optimization and customer behavior analysis are being transferred to the field of mobile websites and mobile applications. The overall goal is to learn more about customer behavior and to adapt mobile applications to it. During the last year a number of companies were founded that offer app analysis beyond simple download statistics as tracked by app stores. The overall goal is to learn more about customer behavior and to adopt - if possible immediately - mobile applications to that behavior maximizing usage times, and conversion rates.
Splitforce is one of the companies dealing with in app analysis and A/B testing for mobile applications. We got in touch with Zac Aghion, co-founder and leader of strategic development at Splitforce to talk about the current state of app analysis and what the future might bring in this area.
InfoQ: Would you please tell our readers who you are and what Splitforce is?
Zac: My name is Zac Aghion, I am the CEO and co-founder of Splitforce which is a mobile, A/B testing and optimization solution for iOS, Android and Unity applications. At Splitforce, I manage day-to-day operations and set the strategic vision of our company.
InfoQ: It seems there was a lot of activity in the sector of mobile application analytics in the last year. What is the reason for this?
Zac: I think that larger companies, especially retailers but also media companies, are starting to react to the massive adoption in connected mobile devices. Larger companies, especially retailers but also media companies, are starting to react to the massive adoption in connected mobile devices. The proliferation of iOS and Android devices which started in the first half of the last decade is only now starting to really be capitalized on by these companies. Wal-Mart for example generates about 50% of their online traffic from their mobile website. Part of being able to build those mobile products in the right way is dependent on instrumenting them with the right tools - and analytics is one of those tools – allowing them to better understand how their products are performing and how they can make them better.
There are two essential questions that traditional analytics is able to answer for these mobile apps: “Who are my users?” and “What are they doing?”. There's a third question which is the last missing piece in order to understand how data can be leveraged and make better products: “How can I make the experience better? How can I make my users do what I want them to do, in a higher proportion or more often?” It's that last question that we are starting to see a lot more interest in from mobile product developers, mobile marketers and even engineers in some cases.
InfoQ: So, there was not a special event or something that triggered the interest in app analytics, but it was just a natural result of the growing adaption in consumers that led to more adaption of analytics within companies?
Zac: In parallel to the increased adoption of mobile devices among consumers there is a shift in types of analytics companies care about. So initially companies placed more of an emphasis on acquisition specifically acquiring new downloads and new users and therefore the focus of analytics was on the analytics of acquisition – so what advertising channels were performing the best, what were the clickthrough rates on advertising for an application, what were downstream, download and install rates from those different acquisition campaigns. In parallel to the growth in mobile devices, there has been a shift from purely focusing on acquisition analytics to focusing on both the analytics of acquisition and post download analytics. And reason for that is so that companies can get a better understanding of how users are behaving when using their app or interacting with their mobile app product. Post-download analytics is becoming increasingly important in order to round out the picture of what happens after a new user of a mobile app is acquired.
InfoQ: When you say that there is a shift towards post-download analytics, is pre-download or app store analytics getting less important now?
Zac: It is not a shift in the sense that there is less focus on acquisition analytics but there's a shift in growth. We are seeing a higher level of growth now in post-download analytics. Acquisition analytics are still important, but what people are looking at is a more comprehensive picture of their return on investment in acquiring a new user - especially for free applications. Gartner Research predicts that by 2017 nearly 98% of applications are going to be free downloads. If you are paying money to acquire new users for a free app, you need to make sure that they are being monetized in an effective way following the download.
InfoQ: If you look at the reason why app analysis is done, is it more for straight monetary reasons – like “I want to sell more stuff” - or is it more for customer satisfaction, “I want to provide a better experience” or is it both?
Zac: In the case of A/B testing, the goal is very different for mobile and web. Testing and optimization of websites has traditionally been focused around improving conversion rates as measured by the proportion of visitors that become customers. In the mobile context it’s not only about optimizing conversion rates, but also discovering what changes to the product users respond best to. Specifically we find that A/B testing on mobile allows companies to uncover points of friction in the mobile app experience and where they can introduce some element of delight or surprise to make the app a more enjoyable product to use.
InfoQ: If you look at the companies that are using analysis tools right now – who is using the more sophisticated methods? Is it the larger brands or is it the start-ups and innovators?
Zac: Newer companies that have mobile in their DNA are doing the best job in leveraging their data in sophisticated ways to understand their users and improve their mobile apps. For example, mobile first gaming companies and mobile first on-demand services like Uber or HandyBook have been best-in-class when using data to optimize their mobile apps.
InfoQ: You mentioned mobile first. Do you experience some mobile first movement also within the bigger brands that do not have mobile in their core business? Can we already see some signs for mobile first actually?
Zac: There are some larger companies, which are being challenged by start-ups or newer companies in mobile because they are behind in terms of taking a mobile first approach to their business. For example companies like Grubhub that have technology in their culture have been faster to move on mobile and have seen more success than larger but older companies like Burger King or McDonalds. Indeed, food and beverage as an industry is one that is experiencing a lot of pressure to adopt a mobile first culture. Domino’s and Chipotle have been particularly successful and have top-ranking apps in the food category of the app store.
Another industry where mobile makes a lot of sense is in hospitality and travel. Start-ups like Hoteltonight or Hipmunk that have had a strong mobile focus have reaped their rewards of a consumer shift to mobile. Many hotel chains are now starting to take mobile seriously because they do not want to lose market share. The Marriot group for example just launched a new app offering mobile check-in and various other ancillaries services.
InfoQ: I was also talking to someone from an Australian bank and they told me that they also saw a change in statistics when they checked the server logs. Usually banks had peaks around lunchtime and in the early evening when people were checking their bank accounts and maybe doing some transactions. But now there no peaks any more, but kind of constant activity from 18:00 on because people are sitting in front of the television and they are using their tablets to do some banking during commercial breaks. It seems that we might experience this mobile first approach in much more areas in the future.
Zac: Mobile banking is another big use case for mobile app analytics because it helps banking service providers understand what tasks their customers want to accomplish from their mobile device. If banks are able to provide mobile products and also answer any questions their customers might have about their account banks have an opportunity to reduce the costs of their traditional customer support and account servicing operations. In the future, banking customers will require that their bank provide them with a seamless and easy way to do what they need to do from their mobile phone or tablet. For big banks the mobile opportunity is twofold – increase customer satisfaction and reduce operational cost.
InfoQ: Yes, that is right. Well, let’s talk a little bit about Splitforce and the services Splitforce offers. You launched new services a few weeks ago. One thing that is really cool is your auto-optimization feature. Would you like to tell us a little bit about this?
Zac: While A/B testing is great, we've found that it is still a very manual process today where it doesn't need to be. Mostly, we've found that many professionals that are managing A/B tests still spend a lot of time designing their experiment, and then after launch tweaking traffic and monitoring the incoming data until a statistically significant result is achieve.
Auto-Optimization is our way of making testing easier. We do this by dynamically adjusting the probability that variations in an experiment are displayed based on their current level of performance. So, in a traditional A/B test with two variations you may have the probabilities fixed at 50/50 for A and B throughout the test. With Auto-Optimization, an algorithm which evaluates the performance of A and B in real-time will dynamically adjust their probabilities to 51-49, 52-48, 53-47, etc. so that the winning variation floats to the top and the worse-performing variation gets pruned away.
The benefits of Auto-Optimization are 1) allows our customers to maintain a higher average level of performance throughout the testing phase; 2) drives faster test results; 3) reduces the time and costs associated with designing and administering a testing program.
In addition to Auto-Optimization, we've also released support for segmentation which allows companies to target tests to specific groups of users. So now, our customers can simultaneously test and measure what works better for old vs. new users, French vs. German users, users who visit during the day vs. at night, etc. We've found that not all users are the same, and so it doesn't make sense to treat them as such. Targeting peels open a whole new layer of the optimization onion. In this way, we've really taken a step towards becoming a platform for personalizing mobile app experiences.
InfoQ: Besides iOS and Android you also have support for Unity, which I know from gaming. Is gaming the main focus of Unity or can it be found in other areas, too?
Zac: The primary use case for Unity today has been in gaming and especially for games that make use of 3D graphics. Unity is great because it is a tool that developers can use to create apps on several different platforms. A Unity app is written in either Unity Script or C# and then compiles into an application that can be run on iOS, Android and even .Net. So, people have even been using it to build games that run in a web browser for example for a Facebook game. From what I understand, it has also been picked up by other industries outside of gaming, in architecture, prototyping and 3D modeling.
InfoQ: If you look a little bit into the future … you are doing sophisticated analytics right now, but what is your vision for the next two or three years. Will we have intelligent software that adapts itself, where in-app analytics is highly integrated with business logic? Is this where we are heading?
Zac: I believe that in the future marketers and analytics professionals will evaluate their choices in analytics tools based on how much it allows them to automate the use of data to make decisions. The assumption is that advanced statistics and affordable computing resources allow us to do very powerful things with data. But few people have the knowledge required to implement these practices on their own. Our vision is to give everyone the opportunity to leverage their data and make automated decisions based on that data.
InfoQ: I am really looking forward to what new features will come with Splitforce, how our mobile apps will be optimized in the near future. It was really interesting to hear your thoughts about this whole topic and I have to thank you very much.
Zac: Thanks, Ralph. This was fun.
About the Interviewee
Zac Aghion leads Splitforce’s strategic development and manages day-to-day operations. He previously managed a team of six as the head of marketing at the China Economic Review, and has been analyzing customer data to optimize digital businesses since 2009.
In the past years, mobile applications took the world by storm and already changed the way we use the internet for work or leisure. Various technologies emerged to create mobile apps and development processes start to consider mobile as first class citizens. But even though mobile already seems to be omnipresent, the future is just about to start. We're facing new generations of mobile devices like wearables or lots of mobile gadgets that make up the Internet of Things. We will be confronted with new types of user interfaces for displaying data as well as accepting commands. And we will recognize more and more companies going real mobile first. All this will influence the way we design, develop and test software in the coming years.
This InfoQ article is part of a series on the fast-changing world of Mobile technology. You can subscribe to notifications about new articles in the series here.