On-line research is no longer the cheap and fast alternative to conventional data collection, but is now defining how to research consumers.
By Piers Lee – Senior Consultant & Managing Director of BDRC Asia
The Asia Research annual review of the on-line research business is a deep dive among on-line industry experts, designed to inform industry stakeholders of key developments in the on-line research business. This leader is the review of 2011 and what to look out for in 2012.
As part of this review, Asia Research undertook depth interviews with the heads of some of the major panel and on-line technology companies including Lightspeed Research (Terry Wiley), SSI (Jason Buchanan), GMI (Ludovic Milet), Toluna (James Rogers), Research Now (James Burge), and i-Link (Chris Rowen).
The most notable aspect of the 2011 review compared to previous years is how these companies talk less about the practicalities of doing on-line surveys in Asia, but instead how on-line research is now determining the type and style of research that is being implemented, and the changes that technology is having on the overall research business.
In a sense the writing is on the wall for traditional research – respondents are less accessible, less co-operative, and less patient. The 20+ minute telephone interview (which is still the mainstay of most CATI projects), is no longer practical with the type of audiences brand owners need to reach out to such as the youth market, affluent groups, and business decision-makers. Ludovic Milet from GMI comments “We can’t do brand health tracking in 10 years times in the same way as we are doing it today.”
On-line research is leaving home
The hottest topic in this year’s review is mobile technology. Research is about how to access consumers, and in many of the developed markets the smart phone has almost become a permanent consumer appendage. Ludovic Milet from GMI comments that consumers “wake up and go to bed with their smart phone”. Its closeness to the consumer therefore lends itself well things like diary based studies.
Research Now has launched its smartphone App panel in the UK and Australia. Here panelists can participate in surveys any time, any place, and this is particularly useful for more mobile groups such as the youth market. James Burge from Research Now says smartphone based research could account for up to 20-30% of their business in 3 year’s time. This method is particularly useful for time / location sensitive studies such as media.
James Rogers from Toluna comments that the technology for mobile research has been there for about 5 years, but the reason it is not yet taking off like it should do is that surveys are still too long. With mobile research the questionnaire needs to be very short, and there are limitations because of the small screen.
The new technology is being designed around consumer lifestyles but the market research agency, probably under pressure from the client, is not adapting the surveys to fit this new landscape of research. Research buyers within corporations have to meet the expectations and information requirements of a multitude of their internal clients all trying to get their slice of information within limited budgets – hence the 30 minute survey. The answer is for all stakeholders to align to the new realities of 21st Century consumers. The definition of a ‘long survey’ has fallen from 35 minutes to 15 minutes and panel companies need to work with the market research agency to educate clients on this. Shorter interviewers will mean more accessibility, and more research – literally “less will be more”.
Some of the old arguments over the ‘representativeness’ of on-line panels in Asia are starting to fade as more and more consumers now get access to the Internet.
Some of the key developments in on-line research over the last year have been ‘river sampling’. The technique has both it plaudits and critics.
Terry Wiley from Lightspeed comments that the increasing demand for respondents has resulted in dynamic sampling and on-line intercepts via the likes of banner advertising, but he argues that there is not enough transparency regarding the use of these methods due to potential data quality concerns. “People (through these methods) are not engaged and as a result click through rates and quality of responses is much worse because people are primarily motivated by an instant reward”.
River sampling can be particularly problematic for brand awareness surveys. Respondents are drawn in from specific websites where they might have had heavy exposure to a particular brand or set of brands. This will lead to highly biased brand awareness data.
The counter argument is that traditional email invitation to panelists is simply not sustainable. James Rogers from Toluna comments that email as a medium is far too overused, and you now need to rely more on community based panels and “embracing the essence social media”. This can mean creating an open dialogue between panelists and then routing them to surveys at appropriate intervals.
Jason from SSI comments that to help ensure representative web-sourced samples, respondents are first drawn into our sampling system and screened prior to taking part in the survey, which overcomes the problems associated with ‘river sampling’. But he points out that the threat to the on-line research business is that methods are being chosen mainly on economics rather than what is the optimal survey solution.
Companies still have to consider mixed methods, comments Buchanan, on-line is part of the mix of survey methodologies and not the be all and end all to research. The SSI merger with Opinionology has helped provide the company a wider range of data collection methods. We believe the optimal mix for data collection from a pan-Asia perspective includes online, offline, mobile and text message-based methods – the industry should accommodate for those countries that have slower Internet speeds and lower smart phone penetration. This can also be more convenient for some respondents, but again for this to work the surveys need to be much buy prednisone over counter shorter than those market research agencies are currently designing.
Change for better
Common feedback across all these stakeholders is how the industry is seriously addressing panel quality issues. Terry Wiley comments that steps to ensure on-line panel quality will be one of the main factors that differentiate the panel companies in the future. Work done by industry bodies such as CASRO and the ARF will come to Asia. Clients such as Samsung, P&G and Microsoft are also driving the quality issue. “Brands need to be aware and have an thorough understanding regarding their data suppliers as methods such as river sampling can be dangerous, and poor panel quality could seriously affect the industry.”
Social media is also viewed as the next big thing in on-line research. There are several applications for research in this medium including short, quick surveys, e.g. for polling and observing chatter. BuzzMetrix and Radian 6 are companies operating in these areas. With social media based research, the incentives need to be different as compared to normal on-line research, e.g. respondents are given credits that they use to buy Apps.
But developments in on-line research are not just limited to sampling and the devices through which consumers complete surveys.
Chris Rowen from i-Link, an on-line fieldwork company, says they are developing new tools for combining on-line qualitative and quantitative methods by branching people off into qualitative research at appropriate junctures. “We are also improving the capabilities in question methods with smart phones. We are also seeing mobile technology being used for traditional in-person research, e.g. at venues as a form of Computer Assisted Personal Interviewing (CAPI)”. Chris also says “the situation in Asia is unique in so far as consumers are bypassing the PC, with their main on-line access being through their mobile. Currently there are around 50 different question methods in PC-based on-line research, but only 12 in mobile and this has to increase. And of course surveys will need to become shorter and more targeted”.
Jason Buchanan comments that technology will continue to move into market research. While some of this technology is about reducing the more labour intensive aspects of the business, others will unlock more insight including facial recognition technology, location based research (e.g. through GPS), and smart phone based research.
Still, the one area that remains mostly off site to on-line research is in B2B research. While some of this does take place (GMI report about 5% of their surveys are B2B), these tend to be highly focused samples – sometimes undertaken at higher unit costs that traditional phone based research. There are still issues about access to more senior business decision makers, e.g. the C-Level, and it will be a very long time before one can get a representative sample of businesses, e.g. appropriately sampled by industry sectors and size of business. In the US, there are panels of sole traders, e.g. gardeners, interior decorators, etc, but in Asia one is unlikely to get enough sole traders within one specific industry – ‘the more you drill down, the more difficult it gets’.
Terry Wiley says “quality B2B panels in Asia are far off. The home business is not proper B2B. LinkedIn tried to set up B2B research with their members 2 years ago, but have since pulled out.”
But Ludovic Milet from GMI points out the on-line panels, while not getting you specific samples by business type, will give you access to business people as a general category and with it higher income groups. Those with annual incomes of USD80k+ are generally accessible in sufficient samples across many markets in Asia Pacific.
James Rogers from Toluna says that opportunities in B2B on-line research exist in partnering publishing houses and also with organizations such as PWC. Indeed the Asia Research publication is very successful in undertaking B2B surveys, albeit limited to those who work in the market research industry!
The conspiracy theory
Many research conferences have highlighted the increasing roll of the on-line panel company in end clients’ repertoire of research suppliers. James Burge comments that clients with in-house research departments have been buying fieldwork directly for many years, so there is no real change here, i.e. they are still buying fieldwork but now on-line fieldwork, though there has been some increase in in-house survey programming in client organizations. Also, when end-clients go directly to panel companies, it frees up more of clients’ budget to spend on more strategic insight-based research with the market research agency.
The roll of the market research agency is to provide the value add, and if the client doesn’t need it (e.g. by virtue of having their own in-house research department), or the research agency cannot deliver it, then the agency will not be used.
While there has been this increase in end-clients going directly to panel companies, the ‘conspiracy theory’ (being the view that panel companies are looking to undermine the research agency) does not hold much water. Still 90% of business for on-line panel companies is obtained from the market research agency. Some agencies are such big buyers of panel-based research that they have set up their own centralized buying departments, partly as a means of routing sales calls away from the researcher to the buyer.
Ludovic Milet from GMI observes that the clients who buy directly from panel companies are FMCG companies, banks, and media companies. Some of these are characterized as having their own internal research departments, but some such as media clients are often looking for very fast turnaround research, often for ‘headlines’ rather than for detailed research.
The panel companies admit that survey research will still need to represent people who are willing to take part in surveys but don’t want to commit to being part of a panel – so traditional fieldwork is not yet dead. But what about the people who are not willing to take part in any type survey? Maybe just watch them – observation research could be the next big thing!