Relevance & Research Question

Social networking sites have become popular tools for recruiting survey respondents through targeted advertisements. Ad design is crucial, as it must capture users’ attention within seconds.While previous studies highlight the relevance of ad design in recruitment performance, sample composition, and data quality, they have almost exclusively focused on static images. However, static images represent only one possible design format, and the potential effects of animated visuals remain underexplored. This study extends prior research by systematically comparing the effects of static and animated ad images on two key aspects of survey recruitment: sample composition and response quality. It addresses the following research questions:

How are different visual elements (static vs. animated) related to sample composition?

How are different visual elements (static vs. animated) related to response quality?

Methods & Data

Data stem from the recruitment campaign for the new online panel GP.dbd, which combines survey data with digital behavior data (e.g., web tracking and app data). The target population comprised adults living in Germany, and recruitment was conducted in 2023 via Facebook and Instagram. Four static images and their animated counterparts were tested. Differences between the two ad formats were examined using descriptive and comparative analyses focusing on respondent demographics and data quality indicators.

Results

Findings show that the visual format of an ad image influences both who participates and how attentively respondents engage with the survey. Static images tend to attract women and highly educated individuals, whereas animated ads appeal more to men, those with lower or middle education, and older respondents. Analyses of extreme response times, item nonresponse, and break-offs yielded mixed findings, suggesting that animation influences attentiveness in complex and context-dependent ways.

Added Value

These results highlight that ad design choices can subtly shape both the composition and engagement of recruited samples. Rather than favoring one format over the other, the findings suggest that static and animated ads serve different purposes in recruitment. A practical implication is to combine both formats strategically. Together, these insights provide nuanced, evidence-based guidance for researchers and practitioners seeking to optimize recruitment on social media platforms.

Relevance & Research Question

Social media platforms, such as Facebook and Instagram, are increasingly used for survey recruitment, particularly for targeting hard-to-reach populations. Previous research has shown that the visual design of advertisements plays a key role in the effectiveness and costs of the recruitment and the data quality of the resulting samples (e.g., Donzowa et al., 2025; Höhne et al., 2025). Pictures are typically used in the advertisements to attract the social media users’ attention and motivate them to click on the survey invitation. However, there is yet a limited understanding of how other visual formats, in particular videos, influence survey recruitment and data quality. In this study, we examine the effectiveness of pictures vs. short videos for the survey recruitment of young adults with social media. We hypothesize that videos are more engaging than pictures, resulting in a larger number of completed surveys at a lower cost, but do not expect differences in sample composition and response quality.

Methods & Data

We will conduct an online survey in December 2025 among young people aged 18-25 from Germany who traveled with Interrail in the last year. The survey includes questions about their travel behavior, attitudes towards the European Union, and European identity. Survey respondents are recruited through Meta (Facebook and Instagram). Using snowball sampling, respondents are also asked to forward the survey invitation to other people who fit the target criteria. In the ad campaign, we use pictures showing different contents, such as a person within a train, trains in front of different landscapes, and railway stations. We create corresponding videos with the AI software Midjourney by setting the respective pictures as the starting frame.
Results

We compare the effects of the two advertisement formats on survey recruitment regarding their effectiveness, measured by the number of completed surveys and referrals through snowballing sampling, and their cost efficiency. Furthermore, we evaluate differences in sample balance across sociodemographics, in particular age and gender, and response quality, measured by completion time and item-nonresponse.
Added Value

The findings will be highly relevant for survey practitioners who plan to recruit respondents through social media.

Relevance & Research Question

Traditional survey methods consistently face critical challenges in achieving adequate coverage and response rates among migrant populations, leading to significant sampling bias in market and opinion research. Understanding these diverse groups is vital for both commercial and public sector decision-making.

Research Question: Can targeted, non-probability sampling methods utilizing social media platforms (SMs) effectively recruit demographically diverse and representative samples of specific migrant populations in European countries, and how do the resulting data quality and efficiency metrics compare to surveys via online access panels?

Methods & Data

We’ve run 5 online surveys between January and October 2025 focusing on first- and second-generation migrants from Turkish and Arabic origin countries residing in France, Germany and Belgium. We used stratified recruitment campaigns across Meta platforms (Facebook/Instagram), utilizing the advertising API for targeting based on age, gender and language. The surveys were run in local languages + Arabic + Turkish.

The collected data was compared with online access panel data regarding metadata on response behaviour (survey speed, devices used, drop-outs…) as well as survey results including a deeper analysis and comparison of language impacts on survey results.

Results

The Social Media Sampling (SMS) method demonstrated clear advantages in efficiency and reach. Crucially, SMs proved highly effective at accessing younger and lower-assimilation migrant cohorts who are severely underrepresented in standard frames.

The results showed, for example, that participants had significantly better knowledge of Turkish and Arabic compared to panel members. In addition, the proportion of Muslims was on average 20 percentage points higher than in the online access panels. Furthermore, we observed that many of the participants via social media were 1st generation migrants and an overrepresentation of people who recently moved to the respective country.

Added Value

This research provides an essential, validated framework for survey practitioners, demonstrating that social media can be leveraged as a rapid and reasonably cost-effective primary recruitment tool for hard-to-reach, mobile populations. It offers a robust, tested procedure for mitigating sampling biases compared to online panels. Ultimately, this study promotes greater inclusivity and accuracy in survey results by ensuring the reliable representation of migrant voices.