By Jonathan Dunn

I sat recently on a panel about mobile marketing analytics at the eMetrics Toronto conference. It was a wide-ranging discussion on marketers? use of mobile advertising, mobile Web, applications and SMS, as well as a debate on how mobile campaign success should be defined and measured.

While my fellow panel members and I were not short on opinions, it was an audience comment that struck me as particularly revealing and raised a number of important issues for brands building their mobile Web properties.

The comment went something like this:

?We?ve built a number of mobile Web sites for clients but we find that when mobile users visit the full Web version, they stay on that version of the site even when presented with an option to switch to a mobile-optimized view.?

Scary stuff if you are invested in the mobile Web space. But let us unpack this observation a bit as there is a lot we can learn here.

Knowing what you don?t
Now, it was not the right forum to ask a ton of follow-up questions and I did not get to speak to the gentleman who posed the question after the panel, so I am going to make a couple of assumptions:

? The full Web version of the site is served up as default regardless of whether the visitor comes from a computer or a mobile phone
? Most visitors tracked or referenced had devices with full Web-capable browsers such as an iPhone or Android device

There is also some important information that we just do not know:

? Who are the clients in question? What is their business, what are their products or services?
? How are the full Web sites built? Flash-heavy? Mostly HTML?
? How prominently was the ?mobile view? link displayed?
? What is the content being consumed by mobile visitors and how does that compare to a wired visitor?
? How do mobile visitor site visit times, page views per visit and bounce rates compare to wired visitors?

When I heard the remark, I proposed that this situation actually created a great testing point for him and his clients.

Instead of having the full Web version set as default for mobile browsers, use device detection and serve up the mobile-optimized version and see how many switch to the full Web version.

With this type of A/B test, you can now see how a mobile-friendly version impacts content consumption, visit times, page views and bounce rates and then bake that information back into your content strategy and site design.

If your mobile site is already well-designed with a data-driven content strategy, you should see improvement across page views and bounce rates.

What happens to your visit times will depend more on the content you are offering and the nature of your business.

Is the information snackable or response-driven like it would be for a retailer? Or, are you a publisher whose content naturally demands more sustained consumption?

Making the case
The case for having a mobile site has been well stated elsewhere and there is plenty of evidence supporting the development of a tailored mobile experience to account for unique mobile behavioral dimensions and device capabilities.

The real outcome of this exchange, for me, was a clear, broad definition of how to roadmap your mobile Web development from an analytics gathering to development input perspective.

Here is a four-step high level view:

Step 1: Use existing Web analytics to gain a view into mobile visitor devices, content preferences and usage patterns.

Step 2: Develop a content strategy based on content preferences and consumption patterns. Develop a design strategy based on device and operating system trends. Consider how content consumption relates to a consumer?s context.

Step 3: Leverage device capabilities ? e.g. GPS, accelerometer, camera and messaging ? based on content strategy and contextual relevance. Wherever possible, build in response mechanisms.

Step 4: Test the mobile version against wired Web norms and mobile content and design premises using mobile-centric analytics. If behavior fails to validate premises, adjust accordingly.

Just because something is working does not mean it is delivering maximum performance.

A streamlined mobile version of your Web site will likely do a better job at delivering against KPIs than a full Web version viewed on the device. To make sure it does, use the data you already have at your fingertips.

Jonathan Dunn is manager of mobile marketing solutions at Digital Cement, Toronto, Ontario, Canada. Reach him at .