In February, VisionMobile published Developer Economics Q1 2014: State of the Developer Nation, which “presents the latest trends in app development, based on our survey of over 7,000 developers” from 127 countries. I spoke with Matos Kapetanakis (Marketing Manager) and Dimitris Michalakos (Web Technology Lead) about key findings from the survey, especially related to the Open Web Platform.
IJ: What do you want to accomplish with the survey?
MK: Our Developer Economics series, now in its 7th edition, is an ongoing research project about the App economy, based on the largest developer surveys. Our goals are to track how developer trends evolve over time and to investigate the important issues of platform, monetization, and tools. Our aim is to help drive awareness of the developer ecosystem, as well as to help developers make the best choices. For this edition we surveyed 7000 developers, making this the largest developer survey to date. Developer Economics is freely available for download thanks to our partners.
IJ: What does the report say has changed in the past year?
MK: Because many of the questions (for example, on platform mindshare) have remained the same across previous editions, we were able to easily identify some trends. For example, one year ago around 50% of developers were targeting tablets. In Q3 2013 70% did. The most recent data show 83% of developers are targeting tablets. Smartphones, however, remain the preferred device: 93% of developers target smartphones first.
IJ: How do developers think about smartphones and tablets differently?
MK: It’s a matter of prioritisation. A total of 93% of developers are targeting smartphones – 72% as a primary device, that is, they first build their app for a smartphone and 15% as a secondary device. The situation is reversed for tablets. Just 12% of developers treat tablets as a first priority, while 57% treat them as companion development devices.
IJ: What did you learn about platforms?
MK: The leading players in terms of developer mindshare remain unchanged compared to other surveys: Android leads, followed by iOS and HTML5. What’s interesting to note is that Windows Phone has picked up mindshare compared to the 6-month-ago survey, rising from 21% to 25% of developers. While these figures are lower than the leading platforms’ it’s quite telling this is the first time Windows Phone managed to convert interest in the platform into actual Mindshare. The timing is not a coincidence, as we’ve seen an increase in device sales for Windows Phone, which managed to outsell iOS in 19 or so countries. What we see here is network effects, meaning that more device sales attract more developers, which in turn should help sell more devices.
IJ: What are the key strengths of the Web compared to native?
DM: Reach, meaning cross-platform and cross-screen code portability, open source licensing and a large developer community.
IJ: Your research and other reports cite “monetization” as an area where the Web needs to improve. What are some of your key monetization findings?
MK: One interesting insight relates to the percentage of developers above the “app poverty line” of 500 USD per app per month. It’s actually the minority of developers that are over this line. If you take the median revenues from respondents per platform, it is interesting to note that iOS has the largest number of developers above the “app poverty line”, followed by HTML5.
DM: I would also note that the median revenue for HTML5 app developers is quite high compared to other platforms. This suggests that contract work in large enterprises is lucrative.
MK: Also, over the past 2 to 3 surveys, we’ve seen changes in which revenue model developers prefer. The in-app purchase and freemium models have gained ground. Pay-per-download and in-app advertising have remained mostly the same. These two ways to monetize apps are the most popular models but not the most lucrative. If we ignore contract work and ecommerce, in-app purchases have a median revenue of $425 per app, per month while the figure is $150 for pay-per-download and in-app advertising. However, looking at median revenues across all revenue models, we see that it’s contract development and e-commerce that generate the highest revenues.
IJ: In March, W3C is organizing a Web Payments Workshop to discuss virtual wallets, secure elements, and so on. What would you tell the attendees about improving monetization on the Web for app developers?
DM: The Web needs a large-scale app store; Mozilla is trying to do this, for example.
IJ: Please tell me why the Web is not yet home to many app stores. In your survey you said, “HTML5 is still far off from being an app ecosystem as it lacks distribution.” Isn’t the Web a tremendous distribution platform? What’s missing?
IJ: In what other areas do we need to improve the Open Web Platform?
DM: What’s missing —which we know from other surveys— is performance, APIs, and integrated tools.
DM: On APIs: HTML5 is missing a lot of APIs available in native SDKs and there are use cases where the Web is simply inadequate. Tools like PhoneGap are important for filling the gap. And the good news is that PhoneGap follows W3C standards where possible, and gets revised every time an API becomes standard. HTML5 needs to innovate and lead the API race, by embracing new areas of developer interest such as the Internet of Things (IoT).
IJ: And tools?
DM: The HTML5 toolchain is broken. Developers write code on vi or Sublime and debug in the browser. Each browser has its own debugging environment. And even though they generally look alike it’s the tiny differences that lead to death by paper cut. Developers tell us they prefer Chrome and Firefox tools, but also that they are complex for some activities such as for memory profiling. Developers like Yahoo’s YSlow because it is easy to use. So, yes, there are HTML5 tools, but they may not be easy enough to use. In How can HTML5 compete with Native? we recommend a unified API for development tools to connect with browsers for debugging. We claim innovation would fuel competition and produce the HTML5 tools developers much need.
MK: There’s a related 2012 post on what the HTML5 ecosystem needs (Mozilla Boot2Gecko: can the new HTML5 champion succeed where webOS failed?).
IJ: So the big message is “we need to look at the ecosystem as a whole”?
DM: Apple and Google are working hard to maintain their dominance. They have created a whole ecosystem, not just a technology stack. To change the current duopoly you have to convince developers and OEM manufacturers. You have to bring together a complete ecosystem recipe: technology, developers, monetization (e.g., micropayments, ad networks), distribution, and discovery.
DM: You need the full set of ingredients for the Web. A Consortium of stakeholders could potentially create them. Single players will have a hard time today challenging the duopoly. HTML5 needs to grow into a full platform. It needs a champion.
MK: Or, you erode some control points. For example, Kindle devices are inexpensive; Amazon doesn’t make money from device sales, they care about content consumption. That’s how Amazon inserted itself into the Android ecosystem, by eroding the app discovery, monetization and distribution control points.
IJ: I read in the survey that HTML5 mindshare is stronger in some parts of the world. To what do you attribute this?
DM: Android and iOS devices are less available or more expensive, so that makes HTML5 more popular. But this is changing, and Android is taking the lead.
MK: Note that regional production and regional consumption are different. Apps can be produced in any part of the world for any market. But in South America, a good percentage of app sales also happen outside regional borders. The ecosystem is more global than regional.
DM: HTML5 is established in some regions. I believe it could maintain that lead in some places.
IJ: Your survey focuses on smartphones and tablets. Is there app momentum on other devices?
MK: The survey shows that, despite the buzz, only a small percentage of developers are writing apps for TVs, M2M devices, watches, and e-readers. This could change in the future. A lot of smart TVs have numerous apps, but the cross screen experience is not there yet. As we said a moment ago, the technology capabilities may be there, but it’s still a matter of pulling everything together (devices, store, etc). It’s important to create a unified experience to drive manufacturers and vendors into a complete ecosystem.
MK: Overall, the fact that HTML5 is the third platform despite shortcomings is very telling about the value of HTML5. We believe the future of HTML lies beyond the browser.
DM: Keep in mind that a few days ago Google announced the ability to create Android apps with HTML5, using Apache Cordova, the same technology that powers PhoneGap. That should say something for the future of HTML.
IJ: Do you mean you see browsers going away?
DM: No, definitely not. Browsers are here to stay. Pick any modern platform and you will see there’s a browser in it. Even the Wii game console has a browser.
IJ: Are in-browser apps going away?
DM: Again, no.
IJ: So if “the future of HTML lies beyond the browser,” do you mean that you see FirefoxOS, Tizen, etc. flourishing?
DM: Not necessarily. Firefox OS might flourish, even though it is picking up a fight against giants. Tizen has not even launched a device yet. It would not surprise me if the champion of HTML turned out to be Google or Microsoft. As we discussed, Google already allows the creation of Android apps with HTM5. And Microsoft seems to be heading in that direction with Windows Phone 8.1.
DM: HTML is a mature technology with a large developer community. It must continue to grow in the way we’ve described (more capability, better performance, more complete ecosystem) and one of the ways it will grow is through out-of-browser tools like PhoneGap. Tools will play a key role in making HTML competitive with native SDKs.
IJ: Thank you both for your time!
To readers: I also recommend a slide deck with key findings of the Vision Mobile report. as well as W3C’s own Standards for Web Applications on Mobile: current state and roadmap.