Still on hold

Market researchers have been trying for a decade to harness mobile phones to obtain better data, and forays are now being made into this field as appreciation of the potential of mobile data collection grows. Mobiles have become an integral part of all our lives and have obvious advantages for accessing hard-to-reach audiences such as youth, time-poor business people and the affluent. This has enabled mobile to enhance traditional projects by adding real-time, live diary or ‘softer’ insights in record time.

But despite all this, mobile is still perceived as a nice-to-have complement to other methods. The overall value of mobile data collection projects, both SMS and WAP-based, amounts to just 0.3% of revenues, according to the 2008 Confirmit Software Survey. So why do researchers still prefer online, which has a penetration of 60%-70% in the developed countries, over mobile, whose penetration exceeds 90% in the developed countries and may be the most widely adopted method of accessing the web in developing markets?

Pls spk 2 me by mob thx
I’m not suggesting that researchers start writing questionnaires in txt spk, but for mobile we do have to be brief. In a world where both clients and researchers are guilty of a desire to get the ultimate insight into their subjects through a single study, questionnaires often run to 50-60 questions or 25-30 minutes online. The mobile medium does not fit this approach as, rightly or wrongly, mobiles are still regarded as very personal and private. Of course, there is also the issue of short attention spans on a small screen and devices that can be unwieldy for typing. This is particularly true in the case of SMS surveys, which require an exchange of text messages and thus limit survey length typically to 3-4 questions.

WAP-based research offers a little more than that. In studies conducted by Harris Interactive in collaboration with mobile phone technology provider The 3rd Degree, it was established that an 11-question banner survey on a mobile site takes on average 4 minutes 10 seconds to complete and delivers a response rate of 0.01% of site visitors who are exposed to the banner. Response rate for those who click through to the survey introduction is as high as 44%. Survey length can be extended for customers who have opted in for mobile contact – among this type of audience Harris Interactive has successfully run a 14-question survey, which takes an average of 6 minutes 5 seconds to complete and yields a response rate of 1.08% of the total SMS mailing or 73% of respondents who click through to the introduction. In comparison, a traditional online survey among an opted-in panel can be expected to achieve a response rate of 20% or higher. While an email invitation provides space and sophisticated presentation tools, the mobile invitation is usually so short that it requires an extra stage to fully introduce the study. With shorter attention spans on mobile, many respondents lose patience or interest before they even get through to answering the questions. Open-ended questions are an extra burden to respondents who are not frequent texters, and small screens mean that questions and answer options need to be shorter and preferably fit on one screen.

So the mobile medium still limits the choice of studies for which it can be used, both with regard to survey content and target audience. Over time, better screens and keyboards are bound to reduce the technical limitations, but it remains to be seen just how precious consumers are about their mobile privacy.

Is mobile really ubiquitous?
Even if a short sharp questionnaire is all you need, the next question that looms over researchers is how much of the target population can actually be reached on mobile. Leaving the ultra short and intrusive SMS surveys aside, the WAP mobile option requires browsing capability. The Nielsen Mobile Media Marketplace Report for Q1 2008 reports mobile web penetration for the UK of less than 13%.

The UK Harris Poll Omnibus of December 2008, reported that 41% of the UK population is set up for mobile browsing. However, once the focus is directed to people who actually browse on their mobiles at least once a month ( 38% of all who are set up for browsing), the pool of potential respondents dwindles to less than 16%. For comparison, the level of internet penetration at which the online mode is generally deemed viable is about 40%.

So WAP surveys remain suited to niche audiences such as young and affluent people, early adopters and tech-savvy consumers. While recognising the value of this for particular projects, researchers remain cautious in applying it when full representativeness of the wider population is important.

Does using mobile impact the findings?
Even with any potential demographic bias removed, researchers who have benchmarked mobile against online surveys report generally more positive findings among the mobile samples. A study conducted by the Danish Broadcasting Corporation¹ revealed that mobile respondents are more inclined than a matching online sample to describe a specific television entertainment programme as funny, entertaining and captivating and less so as trivial, predictable and uninteresting. In another study, reported by Ipsos Mori², recall of advertising within a TV programme was higher for the mobile sample.

These and other examples suggest that at present mobile respondents display variations in attitudes and behaviour which cannot be explained merely by demographics. Understandably, researchers are apprehensive of the potentially considerable risks in using mobile data collection while the mode effects have not yet been fully explored and accounted for.

And what about the cost of responding?
The cost to the consumer associated with mobile communications for either texting or browsing presents a barrier to the wider use of the methodology, especially as one of the core principles stipulated by Esomar and MRS is that “the rights of respondents as private individuals will be respected by market researchers and they will not be harmed or disadvantaged as the result of cooperating in a market research project”.

The main issue here is not so much the cost itself (for SMS there are free short code numbers, and WAP researchers can reimburse respondents) but consumers’ perceptions. Research into mobile browsing habits carried out by Harris Interactive reveals that about 60 percent mobile phone owners who either never browse or browse less than once a month consider it too expensive.

Until mobile browsing becomes a standard part of mobile charging structures, and is accepted in the same way the internet has been in developed markets, WAP surveys are likely to be limited to those people who are either tech-savvy or affluent enough to not worry about the data costs.

So when is the revolution happening?
It would appear that instead of taking part in a revolution, we may have to wait for the evolution of technology and consumer habits to take its course, before mobile becomes a mainstream research method. In the meantime, those of us who are rightly thrilled by the novelty and speed of mobile and its ability to reach consumers in new ways will be adding it to our more traditional toolkits.

Reference:

?1. Jacob Lyng Wieland, Using mobile phones to measure TV-broadcast quality, Danish Broadcasting Corporation in collaboration with TNS Gallup, Denmark
2. AJ Johnson (Ipsos, UK), Philip Martin (The 3rd Degree), Mobile phone interviewing — Why being engaged is a good thing!