Always on, up close and extremely personal—the very nature of the mobile device makes it a potent marketing tool for engaging and retaining customers.
That must be why 54 per cent of Chinese consumers would rather go without TV than without their smartphones. It’s also why 100 per cent of Japanese smartphone owners use their devices to research products and services, according to Google’s recent Mobile Planet study.
Then there’s market share. According to eMarketer, the Asia-Pacific region will account for 57.7 per cent of the world’s mobile users by 2016. That is 10 times North America’s share.
Combine these high levels of usage with APAC consumers being the most receptive to mobile ads in the world, and the mobile channel is very fertile marketing ground. Yet, many mobile messaging campaigns are not yielding the kind of engagement and ROI their creators are seeking, regardless of the message types being used to deliver them. In many cases, such lack of engagement is because the messages aren’t being targeted to specific audience segments, while others are ill-targeted, excessive or not being tested for relevance.
The fact is that too many irrelevant messages will turn off consumers, leading them to opt-out of receiving notifications and marketing updates.
To truly connect with mobile customers and to boost conversions such as registrations or sales, marketers throughout the region need to analyse and follow the 'data trails' that customers create as they interact with mobile messages such as push notifications, SMS, MMS and in-app alerts.
Everything old is new again: Applying tested marketing techniques to mobile
With such heavy device usage and channel penetration, mobile marketers now have at their disposal a series of measurement techniques that can gauge the effectiveness of each message they send and yield actionable customer insights.
Action analytics—A/B split testing and retargeting, among others—can help marketers develop closer ties with their customers by identifying which messages work and, more importantly, which don’t. Yet these techniques aren’t new. Marketers have been split testing and retargeting print, television and desktop-based Web messages for decades and now these techniques are starting to make significant inroads on mobile.
A/B testing involves sending out at least two versions of one marketing message to see which yields a higher number of conversions. Marketers can use this method to know which message elements such as tone, length and urgency are driving customers to convert.
An example is the popular parenting website Kidspot, which engaged OtherLevels to boost audience engagement with its BumpWatch mobile app.
BumpWatch helps its users—expectant mums and women trying to conceive—track their pregnancies or fertility stage by stage. In carrying out the A/B tests, we split Kidspot’s push notification messages to its main audience segments targeting expectant mums in their 28th week:
A. Some important milestones for baby's brain this week. What else has changed? Let's find out!
B. Your baby can still do somersaults, but it's getting cramped in your belly, so he's starting to move into the birth position. Find out more.
Message A produced higher numbers on every metric: it achieved a 230 per cent lift rate as well as a greater percentage of users manoeuvring through the navigation screen of the app at 200 per cent lift.
We then carried out another A/B split message test to women in their 30th week of pregnancy:
A. From this week, your baby's fine lanugo hair may begin to disappear. Why? Let's find out?
B. Week 30. Congratulations! You're in your third trimester. Baby is now laying down body fat and growing hair and nails. Lovely! Find out more.
Message A not only yielded a 308 per cent lift but also a 125 per cent navigation page lift, making it the highest percentage and a clear message winner.
These tests revealed that shorter, more clinically worded push notifications ending in the words “let’s find out” were more effective across the board. After two weeks of testing, Kidspot saw an engagement boost of 87.5 per cent, eventually achieving a lift of between 200 and 300 per cent after sending the most well-received messages to wider audience groups.
Retargeting: Leave no stone unturned
After conducting a split message test, the best way to make sure that no opportunity for a conversion has been missed is to retarget those messages. Simply put, retargeting enables marketers to use what they have learned about their mobile customers to refine follow-up messages. This could be sending a more urgent call to action to customers who didn’t open the message or opened it without converting.
Marketers can retarget mobile messages again and again, using the data gathered from previous messages to place audience members in ever-more precise segments, based on common characteristics or interests, and continually make follow-up messages and offers ever more relevant and compelling.
Combined, these two analytics methods are powerful tools of engagement that can work for mobile marketers anywhere, regardless of country, language or culture. Mobile analytics solutions are flexible and can be implemented at scale, so marketers can measure hundreds of millions of messages simultaneously across different time zones and languages, but are particularly relevant for Asian marketers because of the large amount of available data and intensity of mobile use in the region.
The takeaway
Asia-Pac’s mobile dominance is evident in the fact that the region was one of the very few which saw growth in mobile phone sales in the first quarter this year. That dominance means a world of potential revenues—as long as the region’s marketers are testing their mobile messages for maximum effectiveness.
Consumers are constantly being bombarded with marketing messages. By adopting an action analytics plan that incorporates A/B split testing and retargeting, game developers, brands and publishers can deliver a powerful one-two punch of relevance that makes messages stand out, driving ROI and engagement that lasts.
Brendan O'Kane is CEO of Other Levels.
