Editor’s Intro: It seems that worrying over the state of online sample quality has been an ongoing and growing concern for market research. Innovative solutions are needed. Hugh Carling of Liveminds provides a provocative viewpoint on how participant recruitment using social networks can help alleviate some of the challenges the industry is facing regarding sample quality.
Having read the GRIT Report for many years I didn’t think that I could still be shocked by the continued pessimism about the quality of participants in the industry, but I was. The statistic that blew me away wasn’t that ‘39% of researchers surveyed expect sample quality to worsen over the next three years’ – we carried out our own Research on Recruitment earlier this year and discovered much the same. The kicker for me this time was that ‘fewer than a third of sample providers expect panel quality to get better’!
It’s well known that researchers have long had limited options at their disposal, with most suppliers providing the same thing (and likely the same people), packaged slightly differently. What is a revelation is that there is such limited belief from suppliers that they can address this situation. The report’s conclusion that ‘the traditional model is no longer sustainable’ is clear. Yet most researchers are still relying on trying to find the right consumers on small, static databases. Even for surveys, these database are limited to 20 million repeat respondents globally. It’s time to break out of that rotting walled garden.
One of the major themes raised by those GRIT respondents who viewed sample quality as getting worse was that ‘the proliferation of surveys is stretching a limited resource (the respondent pool) too far’. So why go back to the same people again and again when it’s affecting the quality of research, the reputation of the industry and it’s fundamentally not necessary?
Even if we aren’t completely comfortable with it, we now realise that social networks and search engines know a great deal about us – more than our best friends and family in fact (Wu Youyou, Michal Kosinski, and David Stillwell). Social ad networks allow us to hyper-target over 2 billion participants in 190 countries based on their demonstrated behaviour, demographics and interests. Marketing and advertising teams are routinely using these ad networks to personalise ads and reach their preferred consumers – why isn’t the research industry?
By finding fresh, representative participants for each project we will avoid the problem of repeat respondents getting fed up with re-entering their personal data or getting ‘survey fatigue’.
To look at the raw numbers – the biggest online panels have around 20 million online users worldwide and we’re told often struggle to find young and old recruits. Social networks have access to more than 2 billion online users worldwide, with hundreds of millions of Gen Zs and over 100 million aged over 60. Social ad targeting based on demographics, interests and behaviour can be used to find anyone; from the general population to fanatics of niche products and everything in between.
Using social networks for participant recruitment can’t be the answer for all projects. If you need to speak to people who don’t use the internet then you’d likely have more luck using a phone or going out into the street. Or if you’re doing a ‘gen pop’ quant study, you’d probably be able to find that sample cheaper through a panel. However, panel recruits wouldn’t be fresh and would still have the associated negative issues that many researchers are well aware of. For typical qual and niche quant projects the cost of social network recruits tend to be the same or slightly lower than through existing databases. Recruiting through social networks is not perfect, and perhaps no method ever will be, but it is a huge leap forward from any approach that uses databases of repeat respondents as it’s starting point.
We recently did a study of our research platform, looking at more than 35 million English words submitted in online qual projects. We compared the responses of participants recruited using social networks vs. traditional methods, which included recruiters employed by a client base of nearly 200 companies. The participants found using social networks are genuinely keen to talk about subjects that are of interest to them and submitted 47% more data than than those found through traditional methods.
Now of course, no modern research article would be complete without at least a brief reference to the General Data Protection Regulation (GDPR). The good news is that by finding fresh participants for each project there is no requirement to maintain a database of repeat respondents’ personal data, so the GDPR is far easier to deal with than traditional methods. Once a given project is completed there’s no need to contact participants’ again, so their data can simply be deleted.
The opportunity already exists for the industry to stop relying, more out of hope rather than expectation, on damaged, depleting resources. If we don’t embrace that opportunity, clients will increasingly go elsewhere for the insights they need to make critical business decisions.