Talkin’ ’bout a revolution in mobile analytics


Matthew Robinson, Analytics Director, London

With the number of smartphone activations now exceeding the number of babies born, and brands increasingly embracing a ‘mobile first’ strategy, the need for effective measurement of mobile behaviours has never been greater.

Yet, according to a recent Econsultancy study, fewer than 30% of businesses have a proper measurement plan with clear KPIs. Which is very sub-optimal. We can and should be doing better.

Central to the task of improving this situation is embracing long-established analytics fundamentals, and more mobile-specific metrics and business intelligence. However, it’s really only through cross-device measurement that we’ll be able to figure out the role mobile plays in prompting meaningful engagement and delivering consumer value.

Which makes it all the more welcome then that the very complex business of platform attribution has, just recently, been given a significant helping hand – thanks to Google and a new feature on its Universal Analytics measurement tool. It might not have been what Tracy Chapman was singing about all those years ago, but it’s sure to herald a revolution in the way we conduct mobile analytics.

But first of all, back to basics. Effective mobile analytics – or indeed analytics in support of any and all digital activity – needs to be rooted in an agreed strategy, be guided by business goals and be clear on what tasks the mobile experience is intended to support. To this end, determining whether KPIs need to be based on acquisition, interaction, retention and/or revenue is a function of straightforward business analysis. This being the case, failure to know what mobile KPIs to set is as much to do with not properly considering aims and objectives, as it is any uncertainty around how best to track and measure mobile.

However, while paying attention to the fundamentals is an essential first step, the next one is to acknowledge just how different are the needs of consumers when using mobile devices. With more than 80% of us using a smartphone to compare prices while out shopping, almost 50% using one to see how to get from A to B, and the FT forecasting that 28% of all search will be done on a mobile by 2015, it’s self-evident that handheld devices are perfect for surfacing timely, relevant and contextual content. And it’s crucial that this is taken into account when putting together a measurement framework and considering the scope for optimisation.

For example, in view of the rising number of searches being carried out on mobile, one very good idea would be to add some tracking to the phone number in your PPC ads. This would pull in to Google AdWords a useful set of metrics such as the number of calls placed, the number received, and the average call duration – which, if it’s of significant length, would be a good proxy for conversion. All of which is invaluable to the task of optimising mobile search ads and experimenting to find the best performing call to action.

Another area of focus that’s quite specific to mobile analytics is the one around screen resolution. Reporting on this will act to validate form factor, with any low performance numbers against screen size signifying, quite possibly, a fiddly user experience on that device type.

However, perhaps the key to addressing mobile’s specific measurement needs rests on aligning a mobile site’s function and utility to the actual experience consumers are having on their tablets and phones. Determining the pages and the sections that are seeing the most mobile traffic, and then customising the content accordingly.

This could see you promoting new or featured content much more. Or pushing current deals and sign-up offers to a greater extent. It could, of course, see you create an app. Or it might simply impel you to reduce copy, cut the number of pages, lessen the form fields, go with bigger buttons, or any number of other pro-active responses.

To that end, the whole ‘mobile first’ mind-set really isn’t about just scaling down the desktop version of a website so that it fits onto a smaller screen. It’s about ensuring the mobile destination caters for the actual intentions and behaviours of its visitor traffic, and establishes/refines this in response to what the analytics data is saying – not only by way of headline conversions, but also regards upper funnel conversions, that will very likely generate revenue and/or economic value in the future.

Of course, besides the increasing ubiquity of smartphones, there’s another thing that should be concentrating minds and motivating businesses to improve their mobile analytics, and that’s the lag in conversion rates we see on handheld devices – phones especially. According to IBM data, a quarter of visits to E-commerce sites happen on mobile devices, but only 15% of purchases.

However, while it’s advisable to take any and all steps necessary to improve the mobile experience and up KPI conversion rates, one shouldn’t lose sight of the bigger picture – which is that mobile ought to compliment other platforms, not compete with them. It’s a multi-screen world that we live in now, and people are hopping from to one to another more and more. As such, it’s important not to set mobile KPIs in isolation, but instead to take a properly cross-platform view, and attempt some attribution analysis.

By attribution, we of course mean dishing out some credit to each and every piece of device content that aids conversion. For instance, the product videos watched on a tablet, or the discount offers looked at on a smartphone, prior to the ‘actual’ conversion on a desktop computer.

Now, essential though it is to see where the customer’s cross-platform usage overlaps, properly assessing the part that mobile plays has thus far, to a large extent, been near on impossible. With no cross-device primary key acting to stitch together the many different online journeys, businesses (save those that invite their users to login) have only been able to measure the total number of customers using each individual platform. That is until now.

For at their annual I/O developer conference in early May 2013, Google announced a significant new feature to their Universal Analytics platform – Cross Device Measurement. A ground-breaking piece of utility that will permit the tracking of users across sessions and devices, and which will at a stroke enable a far better understanding of how visitors move from one device to another when accessing content.

No more routinely identifying three visits from three devices as three unique visitors. If these visitors share the same UID (User ID), Universal Analytics will now recognise them as the same consumer – as one unique visitor.

Moreover, not only will this support a measure of ‘assisted conversions, seeing to it that revenue or economic value gets more fairly apportioned across our devices, it will also enable the identification of device-specific conversion paths. So that’s insight into both the extent of mobile’s influence over the consumer when getting them to convert, and the particular devices and the particular steps to pay attention to in an effort to optimise the overall customer journey.

Of course, it could be that we’ll never see smartphones directly convert to the same extent as PCs, or even tablets. But with cross-device measurement, we will at least get to properly appreciate the supporting role they play.

For now, this new Universal Analytics feature remains in beta, and there’s no published timeline for when we can expect it to be available to all. But still, it’s beyond all doubt that it will revolutionise not just mobile analytics, but web analytics in general. And that means a greater ability to affect change, make improvements, and demonstrate effectiveness. Something that the conscientious Tracy Chapman would approve of I’m sure.

Previous article          Next article