These days, data overload presents challenges more sinister than wasted time and resources—poor data quality can be a big obstacle to a brand or business’s trust-building capability with consumers. Blockchain-powered ethical data marketplace Measure Protocol explores this impact in an eye-opening new study on market research data quality, Is trust the solution to dirty data?
The firm’s new report explores the effect of trust principles such as data sovereignty, privacy by design, fair reward and transparency on data quality. With a side-by-side comparison between data collection using these principles and traditional data sources, the research provides a detailed examination of various quality metrics.
“Data-rich industries like market research are facing ongoing challenges that affect data quality, with poorly implemented automation in the industry, dismal user experience, and extremely low compensation rates just to name a few,” said Guy Wates, director of operations and programmatic for Measure Protocol, in a news release. “We wanted to test how these newer approaches based on building trust with the consumer could impact quality. By using the right technology can consumers and businesses come to trust each other again?”
The report sets out a guideline for defining quality, trust and future research
The paper illustrates a first step in establishing that data quality can be positively affected by a focus on consumer data privacy and control, user experience, compensation, transparency, accountability and more.
The firm recommends a “clean water” approach to consumer data usage, in which clean inputs (sample and data from real, validated consumers) through a clean instrument (short, engaging, mobile surveys) should result in clean outputs (data) to be used for decision making.
Respondents in the Measure network outperformed across a number of dimensions which led to a greater quality score. These participants:
- Were less likely to speed through the survey (took 10 percent more time than the average)
- Were 50 percent less likely to be caught by one of the trap questions
- Were less likely to over-claim products they purchased in the last three months or over-claim tech products they owned
- Were less likely to be influenced by order bias and closer to the population mean for measures such as left-handedness in both the U.K. and U.S.
- Attained a higher derived quality score in the study on average
“These principles are proving to be a form of ‘honesty priming,’ which has already been proved to positively influence data quality,” Wates added. “They can have a far-reaching impact on data cleanliness and consumers’ overall willingness to share for holistic insights.”
Measure is partnering with Women in Research (WIRe) for a webinar to discuss the findings of the study and how market researchers can start to build trust with their own audiences for better data quality. Register here for the session on September 23 at 11am EDT.
Measure conducted a research initiative around quality across its own mobile-only network and also across three other prominent data collection networks that provide access to survey participants. The questionnaire was built with a combination of standard questions used in the industry to gauge quality and a number of questions specifically adapted to our approach. The research was conducted in each network at the same time to control for any timing effects. The samples were also controlled by age and gender to ensure adequate comparison of results across sources.