SA's mobile survey future

Amien Ahmed

Sure, there are interesting challenges when it comes to mobile surveys, Among these, Ahmed lists the lack of a mobile population database as the biggest challenge facing the local market research industry. Why? Because it simply doesn’t exist.

Then there’s the fact that those who speak of the power of mobile research claim it’s quick and easy, but is it accurate? Ahmed clarifies, “For a standard national survey you’d go to Stats SA and draw your sample from that, but who is your mobile population? There is no equivalent.”

Coverage area vs coverage error

With mobile surveys we need to be cognisant of coverage area and coverage error. Coverage error impacts on coverage area as follows:

• How wide is the network coverage?
• Does everyone have a mobile phone or access to a mobile phone?
• Are you doing a SMS-based survey or a survey that requires internet connectivity?
• Can all respondents read and write?
• Are we dealing with technophobes or technophiles?

The more ‘barriers’ there are, the more we introduce coverage error or sample bias. We have to be conscious of who is ‘in’ our survey and who is ‘out’ of the survey – who is ‘participating’ and who is ‘not participating’.

Urbanisation plays another factor, with those in cities more likely to own their own mobile phone and those in rural areas more likely to share handsets and even SIM cards. Understanding of technology is another aspect, this time generational as the younger set are more technology-orientated, and find it easier to engage with, so running a mobile survey means there’s a greater likelihood that young people will participate over older people.

Reach respondents, whatever the weather

Ahmed mentions the quick turnaround time as a huge boon: it takes place ‘in the moment’, so if you’re part of a mobile panel and receive an SMS asking ‘What beverage are you drinking now”, you’re likely to reply immediately, as opposed to the conventional face-to-face survey where the interviewer comes to you days after an event.

Andrii Shevchuk © – 123RF.com

The delayed interaction means you’re more likely to give the incorrect answer as this is based on ‘recall’. Ahmed says Kantar TNS has actually proven this, having taken data of traditional surveys against mobile surveys and finding a mismatch of recall responses of significant proportions across the two, especially when it comes to product and brand recall. There’s also wide reach as you can easily get to respondents in otherwise inaccessible areas and weather conditions don’t play a role.

Simply put, you can’t avoid mobile. There’s a big drive to conduct more mobile surveys as they get to people faster, in a more economical way and allows you to use technological features not possible with conventional surveys.

Mobile surveys in the right hands will continue to grow as an important feature of the research landscape.