Testing Right: Key to Better Cell Phones

The wireless world is a mountain of confounding complexities.

Consider the bewildering, frenetic, and financially stressful situation gripping mobile handset manufacturers and service providers all around the world. Most have no choice but to indulge in a never-ending race to pack handsets with more and more functions at lower and lower retail prices—all while trying to reduce their research and development costs.

Across different and often incompatible operating systems, around countries of varying languages, for idiosyncratic personalization needs, and amid quality and brand perception pressures, these manufacturers are trying to produce handsets ranging from low-end voice-and-data models to the most sophisticated smartphones—and to get them to synchronize to carrier networks all over the world. The industry can be thought of as a 10,000-piece jigsaw puzzle.

One of the most fiendishly complex parts of this puzzle is the testing of handsets and software applications before they become available to consumers. Testing deficiencies of many types are systemic and rampant throughout the industry, which drives development costs much higher than necessary. According to a new survey from Accenture (ACN), 88% of companies surveyed don't do an adequate job of testing, which typically consumes about one-third of the product development process and has a major impact on quality, cost, and time to market.

What's worse, testing of the rapidly rising number of—and increasingly complex—mobile handset applications is becoming much more expensive. Based on insight from industry analysts and Accenture's own experience and customer research, mobile handset application testing costs grew from 3% of overall testing costs in 2004 to 20% in 2008. This number is expected to continue increasing over the next two years, to more than 30%. Handset manufacturers now spend millions of dollars testing a device, not just before market launch but even post-launch.

Nobody Wins Wireless-service providers, handset manufacturers, and consumers are caught in this lose-lose scenario. Carriers have handsets in their networks that may not work properly or reliably. Phonemakers can't sell as many units as they would like because the cost to develop each new model is higher than necessary, pushing retail prices too high for many consumers who want phones but can't afford them. And consumers are often stuck with faulty, pricey handsets that won't perform certain functions or that drop calls, thereby hurting their overall experience with mobile communication.

To overcome these obstacles—and to drive the wireless industry toward its full potential—testing processes need to be overhauled so cell phones are more reliable, easier to use, faster to market, and less expensive.

The process of launching new mobile handsets is fast-moving and often elusive. A large number of product defects—especially with software—are fixed in the last two to three months before products hit the market. Tens of thousands of test cases are executed to verify that these fixes work while not introducing new defects. And yet product managers are nervous about whether their products will encounter marketplace rejection. The challenge of cost-effectively testing mobile devices for quality extends beyond the capabilities of most companies. It's often difficult to reproduce the problems encountered on a live network in the field, which may involve isolating the software code responsible for the problem.

In the past, manufacturers had the luxury of phased introductions of new handsets. Now they're expected to launch simultaneously worldwide on nearly impossible time lines. Managing the economics and efficiencies of a global product launch differs from past experience, so it's an evolving challenge for many manufacturers and service providers. Problems that show up after products hit the market put the reputation of service providers on the line, because regardless of who is at fault, consumers don't know who's responsible—and sometimes blame the operator.

Counterproductive Approaches Of course, many manufacturers and service providers do use advanced quality control techniques, including simulation, automation, and remote testing. But some lack the ability to apply these processes productively around the world, forcing them to fix defects at a later stage by pouring on expensive resources. This approach often turns out to be counterproductive because testing costs increase in proportion to volume. Quality suffers. Brand reputations take a hit. And with markets often dependent on seasonal sales cycles, missing crucial product delivery deadlines can be devastating.

Accenture's perspective is that manufacturers and service providers need to address the problem in a more innovative, comprehensive, and systematically integrated fashion. The key is to conduct testing from a central hub. An engineer in India, for instance, can view and remotely test cell phones located in Germany or Tokyo simultaneously, eliminating the travel expenses of having to go to both places and test devices in person.

Engineers can also simultaneously test many more handsets and their features in multiple locations than by using traditional methods. The central hub approach can help drive a 30% to 50% reduction in overall testing costs, an acceleration of deliveries to market by up to 20%, and an increase in the volume of product testing by up to 25%.

There is no getting around the complexity and hypercompetitiveness of the wireless world. But remote, comprehensive, and centralized testing is one way to navigate through this challenging and fast-moving industry.

Note: Accenture will release a complete report on the mobile handset testing situation at the annual Symbian (NOK) Exchange & Exposition, Oct. 27-28 in London.

    Before it's here, it's on the Bloomberg Terminal.