recent research shows that for certain classes of apps, especially those that depend on deep data integration and off-line file access, the answer is more clear cut.
In the case of mobile business intelligence (BI) or mobile analytics (sometimes called mobile decision-support), a just-published Aberdeen Group study on Mobile Analytics found that given the current state-of-the-art in HTML5-based and native code-based apps, native has the upper hand.
Platform-specific or native mobile BI apps have the potential to fully exploit the hardware and user interface capabilities of a specific mobile device. As for the HTML-based approach, it will work on any platform that runs a standard-compatible browser, but will typically have more limited functionality. It turns out, based on our research, that the functional difference in these two approaches correlates with the relative success of an organization's entire mobile analytics strategy.
[ Need guidance on hiring data experts? Read 7 Insights On Hiring Data Gurus. ]
The native mobile BI app approach is on the rise; in Aberdeen's April 2013 report on mobile BI, 58% (a 53% increase over the prior year's study) of the top-performing organizations (the "Leaders") used native apps, as compared to 27% of the under-performers (the "Followers"). It also turns out that the Leaders had a sizeable advantage over Followers in a 2.3-times greater year-over-year improvement in sales revenue and customer retention, and a 70% greater improvement in operating profit.
HTML5-based Web apps' capabilities are advancing rapidly with many similarities to native apps, such as touch-centric graphical interfaces, and limited device control and local data storage. However, well-designed native apps have the potential to be more responsive, more interactive, and more capable as they can better exploit the device hardware and operating system. For example, native apps can have more robust local data storage and synchronization capabilities, can directly access a wide variety of device hardware such as cameras or other sensors, and if designed properly can have full functionality even when offline.
The study found that organizations that use only a browser-based interface to access mobile BI improved their access to timely information (sometimes called time-to-information or TTI) -- such as the latest retail inventory levels or emergency room vacancies -- by 46% year-over-year. However, companies with native BI apps improved their TTI by 66% over the same period.
It's worth noting that the use of native mobile BI apps requires more thought and planning than the browser-based approach. For example, organizations that already use browser-based BI on laptops can immediately use the same apps on any mobile device with a browser, even without optimizing them for mobile access.
With native apps, more decisions need to be made: Which platforms will be supported -- tablets and smartphones? Apple, Google and Microsoft? What about BlackBerry? Which OS versions? The browser-based decision is more straightforward. Want to go mobile? Open up your mobile browser and deploy!
That doesn't mean the browser-based approach will be very usable on the confined screen real estate of a touch-optimized smartphone or tablet. With the browser-based solutions currently available, it's also unlikely that mobile devices will be at all useful when the device is out of range of a wireless signal. This would limit their usefulness in field deployments such as manufacturing, field sales and customer service. But at least deployment is fast, and the solution can always be mobile-optimized later.
For organizations with a well-developed mobile ecosystem, our data indicates that the native approach is preferable, as current HTML5-based apps offer feature parity with native apps only up to a point. But for companies that are just starting out, or need rapid and universal deployment, the browser-based approach might make the most sense.