Recruiting respondents for market research used to be an art, and it remains so because judgment is required from recruitment managers. There was a time when recruitment relied on informal networks and phone lists based in specific geographic areas. Close to the turn of the century, Internet research accelerated the creation of national, even global databases of people with certain profiles important to research sponsors. Where are we today in terms of the nature of respondent recruitment? For one thing, it is remarkable how the panel model has endured the test of time. After almost two decades of social media and data proliferation, consumer and other respondent panels continue to be the backbone of market research recruitment. This is not to say that there are factors affecting the evolution of respondent recruitment. As a qualitative and survey researcher, I can see these trends:
- Artificial Intelligence (AI) as an assistant companion. AI is helping moderators and other researchers improve their screeners and questionnaires. AI is helping researchers identify what pieces of information they may be missing, and how to rephrase certain things.
- Active online communities. The global COVID-19 pandemic made many people stay put in their homes, and this led to the strengthening of online communities. Name a hobby or important consumer characteristic, there is probably a group accessible somewhere. We are literally just clicks away from whomever we want to talk to for research purposes.
- Big data analytics. Consumer and patient panels are now massive, and this has led to the application of large data models. Questions like where might potential respondents come from or how likely are they to accept a research invitation can be easily predicted. To the benefit of our market research profession, while not being fully automated, what we have now is close to a self-refreshing machinery.
Some of the trends above, while positive, could create some challenges. Research sponsors and end clients should always ask their insights agents where respondents originate from and how they are being managed. This idea is based not only on customary quality and regulatory reasons. The trends above impact how relevant results are. Corporate researchers need to understand to what extent data comes from tech-literate respondents as well as from less-tech savvy respondents. Recruitment methods should attempt to create a balanced mix if such variability is expected in their target population.
There are two possible solutions to ensure respondent variability. One is to incorporate panel respondents in our research as well as non-panel respondents. The other one is to ensure that panels themselves incorporate respondents originating from sources that depend on varying degrees of technology. It behooves us, market researchers, to think of the respondent who did not make it to the research because our outreach failed to get to them. In the long run, brands could suffer because of this. While technology can help with recruitment execution in a very positive way, we need to be on the lookout for potentially exclusionary effects that might limit the applicability of research results.
