Crystal Balls or just Bollocks?
A Market Researcher's 1997 Internet Predictions Reviewed
November 2006
Writing this article represents some risk to the writer – after all, it’s a brave person who dares review the predictions they made many years prior, much less in such a public manner.
These predictions were made in 1997, when as a 5-year net veteran already, I wrote a paper for the 1997 Market Research Society Conference. The paper was entitled “The Internet and Market Research – Threat or Opportunity?”, and although the subject matter may now seem old hat, it was deemed worthy of publication by ESOMAR in Europe and re-presented by request on two subsequent occasions. Those facts themselves reveal just how new and controversial the internet was back then, especially for those latecomers who only stumbled across it in the late 1990s. In fact, that research paper had to open with a re-cap of what the internet actually is – there were still non-users in the audience in 1997!
Nearly ten years on, reviewing the threats and opportunities that were highlighted reveals the unfortunate fact that, whilst the internet has undoubtedly been a boon for market research, one fearfully anticipated problem has indeed occurred – and this is a problem of potentially far greater import to marketers, rather than market researchers.
But before preaching to marketers about their problematic ways, let’s first review what has and what hasn’t eventuated from some of those 1997 predictions.
Firstly, the predictions that came true. That the anticipated growth of internet usage occurred is indisputable, along with its uses for data transfer, client communications, desk research and promotional websites. But that’s just workaday business process and no great insights were needed to predict such changes – all businesses have adopted the internet for these purposes.
But it is where the internet and the unique activities of market researchers converge that the big changes have occurred – most being beneficial to researchers and their clients alike, others less so. Most of these changes have concerned the fieldwork process – the interviewing itself.
The predicted growth in online surveys did eventuate, but not in the way envisaged in 1997. Bearing in mind that most websites at this time had little true interactivity and were very expensive, the assumed methodology was that of the emailed self-executable survey programme, as already being used by Synovate for the BankDirect pre-launch research. More often than not, it was thought that email questionnaires would usually be simple Word documents emailed to participants. Thankfully that has now been replaced by the questionnaire-cum-website that is accessed via a simple emailed hyperlink.
The nature of the information gained via online surveying was also a big issue in 1997. Researchers agonized over how responses would change, who could be surveyed differently, and one what subjects. Some predicted advantages have indeed been delivered, such as the improved response rates gained when busy or hard to contact people can answer online questionnaires at a time that suits them best. For example, Synovate routinely surveys General Practitioners by email, with a large proportion of responses coming in the wee small hours.
Another issue was the quality of response – could respondents be trusted to complete questionnaires accurately, and could they be bothered to type more than a few words into those text boxes? Technology has taken care of the first issue, and experience has shown that if anything, respondents write considerably more online than when answering verbally on the telephone.
Online product testing was also predicted in 1997, but only in reflection of how it had been used to date – typically for software testing, a perfect application given the IT bias of most internet users at this time. Incredibly, no commentators I found at the time raised the potential for showing graphics within online questionnaires. The delivery of visuals is one application where the internet has significantly aided researchers, allowing for better pack-testing, prompted recall questions, concept descriptions and more. Add to that the ability to conduct complex interactive questioning formats such as conjoint trade-off research, and the internet has clearly offered significant improvements to the mix of fieldwork methods available. Predictions being made in 1997 were clearly fixated on the online delivery mechanism, rather than the nature of what was actually being delivered.
As well as the interviewing methodology, the other big question in 1997 was who to survey, and how to access your respondents. The ability to email customers was highlighted, and proven since to be one of the most useful applications of email surveying in the industry, especially for tracking purposes and when real-time instant results are required. However, when general population surveys were considered, the assumption was that once usage levels hit critical mass, one would simply buy email lists for surveying purposes – fortunately this transition from researchers to spammers did not eventuate (although some may argue otherwise!). Instead, the rise of ‘panels’, or lists, of opt-in respondents has burgeoned. Overseas, marketers can select from all manner of specialist panels, recruited specifically for online surveying. Complementing simple lists of general consumers (such as the massive international panels run by GMI) are a myriad of specialist panels, for everyone from specialist occupations to new mothers. The smaller population in New Zealand has restricted similar growth locally, but one successful example is the panel of general practitioners now being regularly surveyed through Healthtracker.co.nz.
One aspect of internet enabled market research not foreseen in 1997 was the complete reversal of the traditional surveying approach exemplified by outbound invitations to potential respondents, whether they be by phone, mail or personal approach. The youth research consultancy 18 Ltd realised some years back that youth research had to offer something of value to teens in order to facilitate informative and cost effective research – after all, what teens are at home on a Friday night to do telephone surveys? 18 Ltd’s development of www.18tracker.com shows how online communities can now be developed specifically to draw a target market to the researcher, an opposite approach to the norm.
All these developments have been beneficial to the market research industry, its clients included. But unfortunately the main negative development that I warned of in 1997 has in fact come to fruition. Back then, I predicted the development of ‘do it yourself’ programmes that would enable marketers to programme and host their own online questionnaires, and even download and analyse the results themselves. Much of this is now occurring, and local businesses such as Usuite.com report as many marketing staff using their admittedly excellent DIY services as researchers.
What these marketers have failed to realise is that the online environment through which these DIY surveys are conducted is simply a communication channel – a means of contact. Whether it is a telephone, postal, face to face or online interview, all research surveys deal with human psychology, arguably one of the most complex and challenging things on the planet to understand. The contradiction is clear - on one hand, marketers are quick to point out just how difficult their jobs are, given how they have to understand and manipulate consumers’ minds every day – and then on the other hand many assume that surveying the same consumers is a simple exercise in ‘question and answer’. Market research is so much more than providing interviewing services, after all!
If this seems like a self-interested defence of market research, consider the following examples of research peculiarities:
- when Old Spice aftershave was being pre-tested, research showed that men wanted it to sting more;
- baking pre-mix can be produced with the egg already in it – but breaking the egg into the bowl is necessary for users to feel like they’re ‘really baking’;
- few McDonald’s users actually believe McDonalds tastes better than its main competitors – but they still prefer it;
- taste-testing showed New Coke was better than the original formula;
- concentrated laundry powders need additives to make them weigh more – so consumers feel there’s more ‘packed into’ the powder;
- many consumers defect from highly satisfactory brands – and many dissatisfied consumers remain with the ones they dislike.
These examples clearly demonstrate the peculiarities of the human mind that require specialist skills to fathom – skills that are purchased with the use of specialist market researchers. When marketers decide that DIY online surveying can let them bypass the trained researcher, they’re undermining their own credibility as marketers. After all, if it’s that easy to get into people’s minds, why aren’t they setting sales records?
Jonathan Dodd