OPINION17 September 2020

Trust principles show promise in cleaning up data

North America Opinion Privacy UK

Establishing survey processes based on principles including privacy by design can increase people’s trust in research and improve the quality of data, says Paul Neto.

Data trust secure_crop

Improving data quality has long been on the minds of market researchers. We’ve already proven that instituting fraud mitigation techniques, shortening surveys and getting better at profiling are important, but those can only go so far toward better data. The truth is that we continue to be plagued by data quality issues and declining participation.

This is partially driven by poorly implemented automation in the industry, dismal user experience, and low compensation rates. Are these pitfalls causing overall distrust in the industry? It seems like this might be the case. In fact, the GRBN Consumer Trust Survey found that the same amount of people who trust the government trust market research companies. It found that Covid-19 has had a significant impact on trust overall in the latest report.

No matter how many plasters we continue to apply, such as enforcing interview length limits and minimum respondent payments, there is a central issue being missed: trust.  

In a recent study conducted by Measure Protocol, we found that 32% of respondents would refuse to participate when unsure about their level of trust with a survey provider, and 35% indicated they would provide limited information as a precaution. Similarly, when asked what contributes to their distrust in market research survey firms, 83% indicated it was their level of compensation. Other issues were overall experience ( 88%) and repeated disqualification from surveys ( 85%), while 90% cited the phrase “not trustworthy”. These initial findings appear to support a hypothesis that trust is a primary factor in participation levels and data quality.

Building trust for better data
We wanted to find out if establishing trust principles could help build the trust needed to improve data quality. These principles include:

  • Giving consumers custody of their data
  • Implementing privacy-by-design systems and processes
  • Focusing on user experience
  • Providing transparency and accountability for both sides of the industry
  • Striving for fair compensation for the participant.

To gauge the impact of these principles on respondents, we undertook a research initiative around quality across our own network and three other networks that provide access to survey participants.

Are consumers who participate in an environment built on trust principles very different from those that participate in surveys elsewhere? Are they behaving differently? If they are, how so, and what does this mean for the industry? To answer this question, we ran a purpose-built eight minute survey among a sample of 1,917 18-34 year-olds in the US and UK.

Basic findings that make a case for trust
Using a number of heuristics, we built a framework to measure a ‘quality score’ for respondents, geared to ensure respondents were real, and their answers were considered. These metrics included things like time spent on the survey, absence of straight-lining (varied responses), sensible answers to open-ended questions, avoidance of trap questions, and other industry standard data quality measures.

The differences between the two environments – trust-based app and traditional data collection – became very clear. Respondents in an environment built on trust principles were more engaged, more likely to give thoughtful responses and less likely to fail trap questions – to name a few. These initial findings point toward how data cleanliness can be achieved through networks built on trust principles.

It appears that trust is broken between data subjects and those using their data in business. These data ‘subjects’ are real people with real lives who have come to mistrust where their information goes, how it is used and who uses it. On the flip side, analysts, researchers, marketers and product developers have also come to mistrust the validity and reliability of the data they use to drive business decisions.

By building user-friendly ecosystems grounded in the principles of data sovereignty, privacy by design, fair rewards and transparency, perhaps we can begin to move the needle toward better quality data.

Paul Neto is chief marketing officer at Measure Protocol