When you run a TV commercial, it looks great on every TV set. When you launch a new .com website, you only have to test it for a few popular PC browsers.
But when you launch a new .mobi website, it may look fantastic on the new iPhone, look terrible on the Nokia N95 and completely crash on the Blackberry Pearl.
It’s a fact of life for mobile marketers: there are thousands of permutations of mobile handset, operating system, firmware and browser (not to mention operator versions). And although they might all support the same standard, each has its own little quirks when handling and presenting content.
In short, it’s a potential nightmare that could turn a great campaign into an exercise in brand damage control as frustrated users turn away in droves.
Fortunately, a lot of progress has been made on the challenge of device diversity. mobiThinking.com caught up with Paul Nerger, dotMobi’s Vice President of Advanced Services and Applications – to summarize the issues for marketers:
What actually is device diversity?
- Put simply, the mobile universe is completely unlike the Wintel-dominated PC world. It’s even unlike that PC world when you include Mac and all of the different Linux distributions. The PC world might sound fragmented but there are literally thousands of permutations of handset, operating system and browser version that site owners have to design for and test.
To give some idea of the scale of the problem: there are currently just over a hundred handset manufacturers around the globe and each release on average about a dozen models per year – Nokia releases a lot more while Apple releases a lot fewer. Of course not all of these are used in every market, but most substantial operators support about 120 new and older devices at any single time. And most of these – certainly the latest models – go through, on average, about ten revisions every year. Let’s do the math. 100 manufacturers times 12 new devices a year times ten revisions per year that’s only 12,000 variations that you need to globally support every year!
Critically, as devices add features and functions, they get dramatically more complex too and that creates other challenges. What used to be a simple voice terminal is now a camera, PDA, games console, ringtone player, music jukebox, video camera, email client, internet browser and messaging station supporting voice, text, photo, video and IM.
All these new features are opportunities for marketers, but they’re also threats because they increase handset complexity. We, as an industry, need to make this easier.
Why does device diversity matter?
Nobody wants a mobile campaign or web page that looks great on one person’s phone to look awful, be unusable or just plain crash on another’s. That’s not only frustrating, it’s bad for your brand. If it happens once, it’s bad, unfortunately it can it happens a lot.
Detecting which handset is accessing your content is vital to anyone that needs to serve content specifically tailored to cell phones and portable devices.
This list includes marketers and agencies but also content management companies, advertising servers, mobile games developers, mobile operators… everyone trying to get great content to as many users as possible.
The device diversity problem is one of the main brakes on mobile web adoption among consumers. And it’s almost entirely avoidable.
The designers of the first-generation of mobile websites thought about the mobile web as simply a less capable version of the desktop web. They looked to exploit the new medium by simply publishing their dotcom site and then hoped for the best. The results were predictable: broken sites, disappointed users and brands besmirched with the stigma of the Error Message.
So, how should marketers deal with device diversity?
Marketers need to make their .mobi sites device-aware, so the content is served to every user in the best possible way.
Only by doing that can they exploit the capabilities of each device while taking into account its specific presentation quirks.
Some phone browsers have back buttons. Some don’t. Some handle color. Some have big screens, some small. Some have cameras, some don’t. Some can receive photo messages, some can’t.
.mobi’s tend to be multi-modal in that they use different technologies to communicate with your prospect; you might use web browsing, voice, SMS, and photo messaging all in a single campaign. But to do this, you need to understand which devices can support which technology.
Another aspect that is often overlooked is feedback. When you build your .mobi, something as simple as the price-range of the phone tells you something about the prospect that you’re talking with. This type of mobi analytics at the handset level is vital information that you should be exploiting.
How do you make a .mobi site device-aware?
You need a comprehensive database that lists all the relevant features, capabilities and limitations of every device out there and the software to detect which device your prospect is currently using based upon that database.
I know that using the software word just made it sound a lot more complex but that’s something that we are working on. We are creating and maintaining a device database that’s accurate, comprehensive and up-to-date is a massive challenge for the industry that also comes with the device detection software.
Once you’ve downloaded our device database and our software API, your developers snap it on to your servers, and they can now collect device aware business intelligence data as well as beginning to re-purpose the content on-the-fly to suite each device accessing the content.
What device databases are available?
DeviceAtlas, WURFL, UAprofs, and a few smaller ones.
The public domain databases lack quality and support. The proprietary ones are very expensive. Our investors (who want the mobile web to explode) have allowed us to create the best of both worlds: a new device database that has a company willing to stand behind it, is a superset of those other databases out there, has built-in quality checks, yet is priced very low. In fact, they’ve told us not to make money on this effort; just try to break even. They see this as a big problem that needs to be solved.
Have you got any advice for marketers on how to deal with the most pressing problems of device diversity?
Certainly mobi analytics are easy low hanging fruit. Once you know what makes and models of devices are coming to your .mobi site, you can begin to segment more effectively. But the obvious things are image sizes and font sizes. You want to make sure that your .mobi content fits right on the phone and is readable.
Any other advice for hard-pressed marketers?
Well the best advice is to keep things simple. Limit the choices that your users need to make. Focus on the content that’s relevant to a mobile user and discard the rest. Create site architectures that deploy a simple, drill down approach to ensure that your users can get to the content they want in as short a time as possible.
Remember that a site is only as good as the browser that displays it. Devices and browsers are already pretty good and they are getting better, but you can still expect to encounter hurdles like slow speeds, device incompatibility and screen size limitations.
Supporting the plethora of devices and browsers out there can be difficult and time consuming. Most service providers customise the mobile browsers they sell to their consumers, creating a large number of different deployed browsers and most have a shelf life of two years or more.
Marketers looking to run campaigns off deck – outside the walled gardens of the mobile operators – need to think about which browsers, browser versions and devices their users are using. I’ve seen some marketers beginning to target campaigns at one device type recently – the iPhone is the most common example – but most consumer brands will want to target the widest audience possible.
Making sure your content conforms to industry standards, particularly from the W3C, will help here. These are published on the W3C’s website and your development teams should know about them.
Are there ways marketers and developers can speed up getting a device database?
The main challenge in making a device-aware mobile web site is getting device data to power your site. Device data can come from many different sources including the handset manufacturer and the testing teams within the mobile operators.
Some of the larger manufacturers and mobile operators have developer portals where this information is readily available over the web. But there are over scores of different handset manufacturers and over 500 mobile operators most of whom do not have developer portals. Getting the device data is a daunting challenge for even the largest of companies.
There have been attempts to aggregate and consolidate device databases to make it easier to get device data; some are in the public domain while others are proprietary.
These public domain sources are free, but many observers claim that they lack completeness of attributes and that their accuracy is highly variable. They also lack support — there’s no single company willing to stand behind the data.
There are also some commercial sources of device data based buying devices which are then strapped to a workbench and forced to give up all their secrets. It tends to be a slow, expensive, and cumbersome process.
dotMobi has pioneered a new approach – we call it DeviceAtlas– that falls somewhere between these two extremes. It combines data from multiple sources including public domain, commercial companies, handset manufacturers, and mobile network operators. This new approach applies the thinking behind Wikipedia – the power of user-generated content, tagging and open issue resolution – to the device diversity problem.
Finally, can you provide our readers with a couple of examples of mobile sites that really handle the device diversity problem well?
Yes, there are a couple. Certainly, accuweather.mobi, morodo.mobi, and msn.mobi, and zagat.mobi are all doing a great job. Try a couple of different phones and see the difference.
Leave a Reply