A survey conducted by change-management consultancy Cambiar in 2005 predicted that global online research spend would triple to £2.2 billion by 2008, with between a quarter and a third of all survey research conducted online by 2010.
Whatever else might have happened in the past two years, we certainly haven't seen a slow-down in the development of this sector. With surging broadband penetration and an increasingly representative online population, every leading research company has now shifted at least some of its work on to the web.
"Market research has been growing at three per cent or four per cent annually but, within that, there has been a massive growth in online activity during the past five years," says Peter Thompson, commercial director for Experian's marketing solutions division. "A lot of the big names traditionally had people going around knocking on doors or sitting in call centres, and they are moving it on to the net because of the economies."
Take the example of the online BT customer satisfaction survey, which is conducted by eDigital Research. "We monitor that daily across all their principal interactions and that automatically gets fed back to them on a daily basis, on a weekly basis and on a monthly basis," says eDigital Research director, Chris Russell. "That is research they could not do before, certainly not at the same cost and within the same time frames."
Cost effective
Apart from the cost savings, there are numerous other significant advantages to the internet where data-gathering is concerned. The spectacular reach of an online survey, for instance, puts most face-to-face surveys firmly in their place, giving the ability to capture information from right across a market, rather than a handful of physical locations.
In a similar way, where a qualitative focus group requires a group of specialists, who might be difficult to gather together in a single room, the internet brings its celebrated ability to drill down into communities of interest.
"The internet has become the natural place for hard-to-find special-interest groups and younger demographic groups to converse with each other about products and lifestyles and brands," says Laura Roberts, founding partner at brand and consumer insight consultancy, Sparkler. "That's why it represents a huge opportunity for organisations to talk to consumers directly."
There are other opportunities too. The internet can present panelists with supporting material much more easily than many physical methodologies could manage, while research into the use of relatively interesting mechanics, such as sliders, even suggests that people tend to be more discriminating in their answers when the application is appealing.
Tried and trusted?
"People tend to forget the drawbacks of methods that have been around for some time," says Nick Sparrow, managing director of social and political polling company ICM and founder of online research specialist NewVista. "Actually, the old methods weren't necessarily that clever, so in some respects, online offers some significant advantages over traditional methods."
Not every effect of the change has been totally positive: Esomar last month indicated that global research revenues were static in a market which grew five per cent by volume, largely due to the ongoing adoption of lower-priced online research.
What's more, cheaper clearly doesn't always mean better, and certainly not in the sensitively calibrated world of market research. Even as brands revel in the responsiveness, the flexibility and the favourable cost of online research, doubts hover over the web's fitness for certain quantitative and qualitative assignments, as well as the rigour of some companies' methodologies, particularly among the industry's old hands.
WPP's Lightspeed Research has an online panel of 200,000 respondents in the UK and has gone as far as to initiate a project entitled 'Research on Research' to investigate the reality of online panel research methodologies, promote best practice and ensure high-quality, reliable and accurate responses.
The results have been hugely instructive, says Lightspeed chief executive, David Day.
"It highlights how apparently minor differences, such as the way respondents are invited to surveys, the incentives that are offered, and even the day of the week and time of day impact on the reliability of the results," he explains.
Russell also cautions against the assumption that all market research methodologies produce the same results. "A retailer's email survey that drops in immediately after a purchase has been made may give a different result from the one you would get from someone phoning two days later," he says. "The immediacy of online can be a benefit, but you need to be aware of the difference."
Sparrow believes that for all the strengths ?of the online platform for research, it is important not to overstate its advantages. "It is quick, but it is probably no quicker than telephone research," he says. "When it comes to cost-effectiveness, it certainly is cost-effective if you pay your panel members nothing and don't worry at all about quality control."
There are sound reasons for practitioners to tread carefully in this area. In the US, where the online research market is significantly more evolved, statistics recently emerged that suggested that 34 per cent of all questionnaires submitted to the top ten online panel providers had probably been completed by only one per cent of all panel members. If some people were always more likely to answer questionnaires than others, even in the offline days, they certainly never had the opportunity to be that prolific before.
A genuinely significant strength of online research is in the removal, in most cases, of the interviewer, who was known to be capable of skewing or even falsifying responses either face to face or over the phone. But equally, what has replaced the interviewer isn't necessarily a fail-safe mechanism.
"The quality-control measures the industry had in place were all directed at making sure the interview did his or her job properly, and when it moved online, a lot of people said: 'Oh, there'll be no more quality-control issues, because we haven't got an interviewer'," says Sparrow. "But what happened was that the respondents themselves replaced the interviewer in the pay of the market research company."
The kind of payment the market research industry offers its online panelists varies from promotional incentives to sub- minimum-wage amounts, to somewhat larger cash sums - NewVista pays £12 an hour and Sparrow believes the company leads the market in that respect.
Experian's Canvasse Opinion survey draws on an online panel of 60,000, who are offered modest incentives or rewards for completing certain surveys, though Thompson says the value of the gifts is kept relatively low. "It is very important that the incentives are a thank-you for doing it rather than a reason for doing it," he says.
Monitor respondents
Though fraudulent questionnaires are fairly common - certainly if the US experience is any guide - technology offers ways to clamp down on offenders. One is to physically prevent respondents from answering questions in less than a specified amount of time. "That doesn't ensure that they pay attention, but it does stop them from sitting on a button and racing through as fast as possible," says Sparrow.
There is software available that looks for suspicious patterns, while basic check questions at the beginning and end of surveys are another way of catching people who are not paying attention. Likewise, where the person is a returning member of a panel, repeating basic demographic questions and comparing the answers can provide a further tell-tale sign when things are not quite right.
For all the care that needs to go into online research before the data it produces can truly be trusted, there is no doubt in some quarters that the online platform has brought a much-needed breath of fresh air has to the marketplace.
"Research response rates have long been in decline for face-to-face, because of security issues," says Rob Seebold, director at research outfit Qmedia, a division of Leeds-based marketing agency Quaestor. "Rates have also been falling for telephone research, as a result of the increasing use of silent numbers, the popularity of the telephone preference service and the move to mobile phones as the primary phone line. What this means, increasingly, is a move to a permission-based model."
Logically, Seebold argues, permission-based online panels, where consumers respond at a time convenient to them, simply have to yield more considered and accurate data than that collected via unsolicited telephone calls or knocks at the door.
The tried-and-tested models may indeed not have been nearly so pure and effective as people occasionally imagine, but while some suggest that qualitative research might just as well migrate entirely online, others advocate a hybrid approach, in which traditional methodologies feed into new ones.
"There is no reason we shouldn't meet people in a qualitative research environment and then continue the dialogue digitally," says Roberts. "Why would you abandon the relationship and rapport you have built up over the course of a focus group when you can keep up the conversation online, easily and cheaply?"
Lightspeed is among the research specialists that are in the process of running tests on a new mobile research offering. "The strength of it is that it allows us to provide responses to our clients very quickly," says Day. "It also allows us to target specific groups, such as younger people, who might be harder to reach through traditional online surveys."
The further possibilities of such a method are numerous, including the ability to conduct research during a particular event, or to elicit visual information from the respondents. "One of the advantages is that we can ask respondents to send us images, such as ads they have been exposed to - and that is something we are trialling," says Day.
Mobile is not the only new frontier of research. Procter & Gamble chief executive, Alan Lafley, recently stated his belief that we live in "a let-go world" and urged marketers in general to adopt new modes of communication with their customers.
The buzz phrase might be a new one, but the sentiment is already widely held, and nowhere more so than in the research community, where other areas of interest include social networking sites, virtual worlds such as Second Life and the potential for panelists to record and upload video diaries using webcams.
Youth research specialist Q Research carries out surveys by mobile phone, asking respondents to send in photos and videos of themselves, as well as replying to surveys via WAP technology. A recent project with BDH/TBWA, which was pitching for a project from hair-styling company GHD, focused on girls who use styling irons, asking them to video themselves on their phones.
GfK NOP, meanwhile, has created a Bebo page called "BrandedMinds", as a means of researching hard-to-reach target populations in younger age groups. "Our goal is to expand the number of channels available and to offer different demographics a choice of research modalities," says GfK NOP global director of online research, Mike Cooke. "The point is that they can choose to respond by the most appropriate means."
Second life
GfK NOP has also been conducting interviews in the Second Life virtual world using an avatar by the name of Gabrielle, although the research at this stage is strictly qualitative.
The possibilities of online do indeed mark the platform as something very different from every method the research industry has had before. But does offline offer a wholesale replacement for online methods? Probably not, and certainly not without a great deal of care.
"In terms of migrating traditional research methodologies, we would certainly argue that there is a place for the traditional face-to-face groups, one-to-one interviews and telephone research as well," says Russell. "Certainly, the growth of online research is phenomenal, and it is so much less expensive, but it needs to come with a certain number of health warnings."
ONLINE MARKETING RESEARCH TIPS
- All methodologies are not necessarily equal, and they certainly don't all produce the same results.
- Traditional is not necessarily best.
- Neither is online: telephone research can match online for speed and digital cost-savings may sometimes be reflected in the quality of the data.
- Over-incentivising may lead people to complete surveys for the wrong reasons.
- Popular brands can leverage those assets to mobilise research uptake.
- Online can make the world a smaller place, but it can also be a misleading one - beware professional panelists.
- Where you have the ability to interrogate the entire world, make sure you are not always interrogating the same bit of it.
ONLINE APPROACH BREAKS DOWN INHIBITIONS FOR DUREX
Some subjects simply don't lend themselves to face-to-face research, and sex - in the abstract sense - is one of them. For around a decade, until 2005, Durex's Global Sex Survey had been carried out in the street with pen-and-paper questionnaires.
"That can be quite expensive and quite time-consuming," says Gail Dudleston, managing director of digital agency twentysix Leeds, which transplanted Durex's entire survey into an online model.
Respondents were recruited through an online campaign, which included links on Durex.com and its 60 branded country sites, as well as with an email campaign to the previous year's respondents, banners on MSN and Yahoo!, pop-ups and site links.
Real-time statistics reported on the number of surveys completed per country, enabling Durex to roll out extra online marketing activity in those countries where responses were running below the 1,000 required for eligibility.
In its first year, the campaign succeeded in increasing the number of respondents to 317,000 (up by 135 per cent), while increasing the number of participating countries to 41. In 2006, the results were more dramatic still, with 650,000 responses across 52 nations.
"Some of the material we got was quite personal, about people's attitudes to sex and sexual behaviour, so online worked really well," says Dudleston. "It significantly brought the brand's costs down but at the same time allowed scaleability, in terms of increasing responses from different countries."
.jpg)
.jpg)