Survey Respondents Lack Identity

The recent exposure of fraudulent survey practice at an agency in the USA has once again put the spotlight on the reliability of survey data. Some industry commentators take the view that agencies are actually very good at anti fraud actions and that the real issue affecting data quality is long and boring surveys.

It is true that excessively long and boring surveys significantly contribute to respondent fatigue and poor-quality data. However, the industry's current focus on survey length, question engagement and commoditisation of sample overlooks a more fundamental issue: respondent authenticity. While improving survey design is undoubtedly crucial, placing primary emphasis on outputs without first addressing the quality of inputs, specifically, respondent credibility, risks missing the core challenge.

Surveys that extend beyond reasonable lengths, filled with tedious questions presented in repetitive grids, undeniably encourage disengagement. Respondents confronted with such tasks naturally tend towards shortcuts, guesswork or complete abandonment. It is important that the industry moves away from lengthy, uninspiring questionnaires towards shorter, more engaging formats. However, to assume that halving survey duration and offering higher incentives will automatically resolve all data quality issues is overly simplistic.

The root cause of poor survey data often lies deeper, within the respondents themselves. A significant portion of respondents engage in practices that compromise data integrity either by gaming the system to maximise incentives, conforming their answers to perceived expectations of researchers or projecting an idealised persona rather than their true selves. These practices are not necessarily driven by survey length alone but reflect broader behavioural patterns and motivations that persist irrespective of questionnaire design.

Online survey panels often inadvertently incentivise dishonest or inaccurate participation. Low incentives might discourage authentic respondents, while simultaneously attracting those with minimal regard for data authenticity. Yet increasing payments alone will not necessarily filter out individuals intent on misrepresentation; indeed, it might attract more sophisticated forms of gaming. This suggests a critical oversight: sample quality, rather than simply sample cost or survey length, demands greater scrutiny.

Efforts to combat this issue have typically focused on sample providers’ existing fraud detection mechanisms and engagement algorithms. While these are essential, there is a need for more nuanced strategies to validate respondent authenticity proactively.  Advanced behavioural analytics, sophisticated identity verification techniques and enhanced profiling methods could help distinguish genuine respondents from those presenting contrived or fantastical personas.

Furthermore, the industry must confront its reluctance to recognise this uncomfortable truth: some respondents simply invent data. They may construct imaginary lifestyles, opinions or purchasing habits, driven by internal narratives or a desire to meet perceived researcher expectations. Merely shortening the length of the survey or raising incentives will not eliminate this phenomenon. It requires concerted efforts at an industry-wide level to rigorously vet respondent credibility and continuously refine methodologies that flag and exclude unreliable data sources.

In conclusion, while reducing survey length and rethinking incentive structures are important steps towards improving data quality, the industry must first prioritise the quality and authenticity of respondents themselves.

Without addressing this foundational issue, attempts to improve outputs - however well-intentioned - will inevitably fall short. Ultimately, the path to better quality data involves deeper investment in understanding and verifying the legitimacy of those providing it.

Only then can survey data consistently deliver accurate, reliable and actionable insights.