Aggregating sample or multi-sourcing has increasingly become the norm for online data collection in quantitative market research. A confluence of factors, including overall demand for research, difficult-to-reach audiences, and recruitment budgets, necessitate drawing on multiple sources to complete nearly all projects. This sample blending practice can broaden a project’s reach and allows for faster and more diverse fielding. But as sample aggregation evolves, so has transparency into the blending practices and sources.
Why are researchers increasingly adopting marketplaces that focus on multi-sourcing transparency? There are several crucial implications driving demand:
Data Bias
Researchers have long monitored supplier blends for wave and tracking consistency. The new increase in multi-sourcing requires more transparency to understand potential shifts in these projects or any deviations from norms. It is common to have 30 different sample sources supplying a single project. While this may present challenges in keeping exact blends, this heavily fragmented sourcing increases the representation of audiences and reduces the risk of a few large sources changing allocation wave to wave. Transparency into the sources is required to see no single source spikes to an allocation that could bias data (>40%). These blending issues can concentrate by device type or demographic quota. A sampling platform like PureSpectrum’s Marketplace provides a simple dashboard and excel outputs to analyze blends to a demographic quota allocation level.
Sample Quality
If in the uncommon event of a fraudulent breach, it is most likely that a single source is the cause. Tracing bad quality responses to the source and comparing the distribution across the total distribution of interviews identifies and neutralizes the issue. When executed quickly, fielding can resume without the offending source, and insights are delivered to clients with confidence and timeliness. Moreover, identifying sources that have experienced an attack helps the resolution and security improvement, adding to the feedback loop for better quality data.
Insights Expertise
Clients select their insights partners for expertise in problem-solving and in the research process. As multi-sourcing becomes the dominant fielding approach, questions arise from clients who have previously been accustomed to responses from a single, consistently managed panel. Clients may inquire about recruitment, sample quality, or fear multi-sourcing. Researchers who require transparency will be better equipped to answer these questions and advocate for sample blending. A clear grasp of how the data was sourced, and a point of view of the reasons and advantages behind the approach, drive confidence in the insights delivered and in the researcher as an expert partner.
Sample Feedback Loop
Less discussed but equally important in programmatic marketplaces is information on the sample buying part of a transaction. A panel source can better optimize asset management if they can understand the researchers that are purchasing the sample. It creates trust in the market, better clearing prices, and reduces duplication and bad respondent experiences. Transparency must flow in both directions for programmatic sample marketplaces to function optimally. James Rogers recently highlighted how critical the first 15 minutes of fielding are to transmit accurate, transparent information to panels if researchers want to realize the potential of programmatic marketplaces.
While several factors have forced multi-sourced sampling, it has also produced several advantages in online data collection. Sample marketplaces offer a deeper pool of respondents, less inherent bias from single panels, reduced business risk, and faster fielding. Without transparency into the methodology and sourcing, a researcher cannot enjoy these benefits because of risk to data analysis, quality issues, or mismanaged client expectations.
Interested in learning more about the sample source transparency offered on the PureSpectrum platform? Reach out to our team below: