01 Oct

Total Time Spend Using Mobile Web vs. Apps (October 2012)

Saw this today at BI website…

Source: Business Intelligence (BI)

Source: Apps More And More Important Than The Mobile Web (Business Intelligence). Original report by Nielsen.

The above is self-explanatory…

What I find ‘funny’ is that over time I have noticed that Business Intelligence’s opinion on “native vs. web on mobile” swings back-and-forth (based on the top news of the day). In this particular case, all the recent attention to “native vs. mobile web” is thanks to Mr. Zuckerberg.


30 Sep

The Eternal Debate — MoMo London Event on HTML5 v.s. Native (Sept 2012)

Native vs. Web
Image Source: Mobile App Testing Blog

Seems that I missed a very good debate. I just read a blog on MoMo London event on HTML5 v.s. Native.

Seems like an eternal debate.

Today in 2012, I am still amazed we still are debating this and have not been able to address this. This is really a ~10 year old debate, still driven by the exact same issues and pros/cons as before — centralized vs. not, cross-platform vs. not, maintainability and fragmentation, code-reusability or not, better user experience vs. not, performance, security, access to device APIs vs. not, thin vs. thick, app discovery, etc. etc. etc.

We can argue the basics are here (HTML5, CSS3 and JS-and-related frameworks), but creating great mobile webapps with great user experiences is today only possible by a few; in other words, is a niche area. (Not even Facebook was able to pull it off, right?).

The day the “common mobile developer” is able to create great mobile webapps with ease, is the day this debate will end.

Today still, “user experience” (driven by network latency, app richness, toolsets, adoption by big brands) is best maximized on mobile native. Today still we have to talk about classes of mobile applications (again driven by network, richness, performance, storage, toolsets, maintainability, security, cost of one vs. other, etc) — then decide what is better suited — native vs. web on mobile.

How much longer will it take settle out this debate? 3 years? 5 years? 10 years? Right now I say around five years — I wish I am wrong. But does it really matter?

In the meantime, successful mobile developers redefine the meaning of “full stack developers”; successful mobile developers must be “End-to-end, Cross-platform, Full Stack Developers” — this is a lot of complex ground to cover.

Related to this see “The biggest mistake we made as a company was not investing enough on native.” — Zuckerberg (2012).


11 Sep

“The biggest mistake we made as a company was not investing enough on native.” — Zuckerberg (2012)

The title of this blog is not what Zuckerberg actually said, but is what he really meant.

From TechCrunch Disrupt — Zuckerberg Shows He’s The Right Man For The Job:

“The biggest mistake we made as a company was betting too much on HTML5.” While building native apps that were bacially just a wrapper for the mobile web standard let it experiment quickly, it made the apps run way too slow. “We burnt two years.”

This validates what I have been saying for years… Don’t take me wrong, I’m a fan, user and developer of webapps. Web on mobile is big and the mobile browsers and frameworks are getting so advanced, and the mobile webapps so kickass, but for certain kinds of apps, especially rich consumer-based applications, native is the way to go today.

Consumers are about great user experiences and great quality. The “mistake” Zuckerberg refers to is not really that they bet “too much” on HTML5, but that they didn’t invest enough on native.

But that said, FB has nailed it down; recognizing the need to invest on the native apps and related infrastructure, and focusing on mobile first.

Any company going global must have a mobile first strategy — we know that for many around the world, their online experience will be on mobile.


24 Jul

Are we still talking about WORA?

A friend of mine who works at Phunware recently pointed me to their blog to read about a recent acquisition they have made; very cool, congrats.

Then a recent blog post of theirs caught my eye (as it is a topic close to my heart): The Delusion of “Write Once Run Anywhere” Mobile App, where Phunware’s CEO writes:

“…I turn my attention to another popular myth about the development and publication of mobile applications: the delusion of “write once, run everywhere” mobile applications and the fallacy of their existence.”

WORA? We still taking about WORA?


Today, talking about WORA is similar to the debate on mobile apps vs. webapps — seems it will never end. But I guess since for many mobile is still a new space, it is good someone writes about it.

True WORA will never happen. Not back in 2004 during my J2ME days when I wrote True WORA will never happen, and not today.

For certain classes of mobile applications though, there is some light at the end of the tunnel, as “close to WORA” can be achieved — for native apps with the help of cross-platform development tools, and for webapps thanks to toolkits such as WebKit, jQuery Mobile, Sencha, and so on.

But today (still) once you go media-rich, highly-sensor and context-based, it starts to break-down.

Even popular tools like Titanium and PhoneGap have limitations when trying to maximize the experience/goals, or maybe the problem is with the folks using such tools, but nevertheless, a number of companies that I have talked to recently and which have advised on the benefits of native app vs. webapp, have given up entirely and instead have explicitly asked for target-specific (Android vs. iPhone vs. mobile web) development only vs. trying to go cross-platform with the hope of reducing development costs.

At the same time, some of companies do realize that while mobile webapps won’t able to deliver (today) the level of functionality and user-experience that is as rich as native apps do, they do understand the trade-offs where the functionality and experience of mobile webapps is sufficient-enough for their goals, their customers, across mobile platforms, with respect to their development budget and expectations.

So yes, for certain application classes WORA is possible today, while true WORA across all classes of applications is not.

Related to this see: Mobility in 2011: Mobile Apps, Webapps and Tipping Points


02 Jan

Mobile Apps in 2009: Local/Native, Mobile Web, App Stores

Happy 2009 New Year to all my readers. My first post of the year is about mobile applications: local/native vs. web, and app-stores.

App-stores have been shifting the balance on application development and distribution (back) towards local/native applications. I don’t mean to undermine mobile web which will continue to be very important and very large for access of information of type “web content”. But the reason for this is the same reason I’ve been preaching for a long time: the ability to deliver/maximize “application richness, functionality and experiences” – which (today) maximizing these is only possible via local/native applications. This is the same topic took Ajit and myself into a debate at JavaOne’s keynote a couple of years; that was fun.

What I’m describing above can be seen on the iPhone which is the best mobile web handset today, and which redefined and raised the bar on mobile web applications, yet the really cool applications are local/native, and more importantly, developers of local/native applications seem to be the ones who are generating (receiving) the most revenue – after all, it is about making money. And this trend will continue… I do believe that in the future mobile web will be able to match the richness and functionality of local/native applications, once the proper APIs and functionality are put in place and become standard, and that today a happy medium are local/native applications that consume mobile web content; i.e. hybrid apps providing the best of both worlds.

And when combining the above with App-stores, which provide for the application discovery and revenue streams for developers, the market place becomes very attractive and thus active.

But to be successful, app-stores must exhibit certain characteristics:

  • From the end-user perspective: the app store must be seamless and well integrated into the user experience. Downloaded applications must work. There must be a good selection of high-quality apps. Integrated checkout/payment is easy and straightforward and secure. The end-user is in control, including influencing how applications will perform on the market (via feedback that influences ranking).
  • From the developer perspective: there must be a low cost and barriers to entry and distribution. Must provide application visibility (see below). Good revenue model. Provide feedback back to developers for improvement.
  • Application visibility: the app store must provide the means for good application visibility. Already established applications are ranked appropriately based on user feedback, while new applications (including new versions) go into a different bucket that allows them to be visible regardless of ranking (perhaps for a period of time).
  • As Ajit writes, must provide a true ecosystem (that benefits everyone: developers, network providers, the end-user, and so on).

Part of the above is why the iPhone has been successful. And is also the reason that I expect the Android app store to do well once it starts paying back to developers — it is just then when the Apple and Android stores can be compared.

There are app stores for Java ME, for example GetJar. The problem faced by GetJar is that there are things that are out of their control, such as cost and barriers to entry (due to fragmentation and certifications and fees and it is just a pain-in-the-neck to deal with network operators issues in general), and not being well integrated (seamless integration) into the overall user experience.

For years I’ve been attributing the lack of integrated solution for Java ME that works (per the above) for application discovery and download and revenue share, as one of the top the reasons why Java ME has failed to maximize its opportunity. As a member of MIDP expert group, this is an important lessons learned for me, that sometimes you do need to include such functionality into the platform vs. expecting 3rd parties to solve the problem; this seamless solution is still needed for Java ME…

On Goodbye 2008 and welcome 2009 and some predictions on mobility I made some predictions on app stores, which I will repeat here:

  • Google will introduce a checkout process for its app store, and developers wanting to make money will notice; the Google app store will explode with a large number of applications.
  • App stores will continue to have its huge effect on mobile apps and distribution. Due to the revenue and fast distribution models offered by iPhone and soon Android app stores, developers will first target such local applications (vs. mobile web). An even larger number of local/native applications will be created and distributed via app stores for Android and iPhone.
  • The BlackBerry app store will be somewhat successful.
  • Someone will introduce an app store for mobile web that goes beyond an application catalog. dotMobi will take leadership by going beyond an application catalog but also providing an associated business/revenue model.

Last but not least, and related to this topic, check out Paul Golding (Wireless Wanders) on his video blog entry Mobile 2008/9 all about App Stores where he discusses and provides his insight on why app stores have been important in 2008 and how important they will be in 2009… right on.

Also related to this see Ajit on Mobile Web Megatrends event – Making money from Appstores – Singapore – April 27 and 28, as well as his related thread on ForumOxford.