Five ways to pursue better data quality in market research

The quest for data quality continues, and we must be relentless in our efforts to achieve it. Tired of mere lip service, the industry recently came together to take real action, as associations from around the world formed the Global Data Quality Project to “combat ongoing and emerging risks to data quality in the market and social research, consumer insights and analytics”.

Data quality is an enormous issue, and achieving it can seem overwhelming and unattainable – especially as the goalposts seem to keep shifting. Thankfully, the framework the new project provides helps to provide a path toward success, addressing each piece of the challenge comprehensively.

As an industry, we must keep this dialogue front and centre, honing in on solutions and pursuing data quality at every stage of the research process. At the potential risk oversimplifying the issue, here are five ways to boost quality in market research.

Use the right terminology: Words are important. The language the industry uses to discuss data quality is the foundation of the other work being done in the arena. Right now, the Insights Association is expanding its work on universal data quality vocabulary that focuses on accuracy and transparency. By using the proper terms to discuss the issue, further disconnects can be minimised.

If everyone uses the same words when discussing things like fraud, duplicates and survey cleaning, it can help ensure everyone involved is on the same page, and faster progress can be made. Keep a close eye on the work being done by leadership bodies in this arena.

Battle survey fraud from multiple angles: Fraud is complex and dynamic, and the tactics leveraged by cyber fraudsters continue to morph and change creating a very reactionary position for the insights industry. Any tools we have at our disposal to fight against their behaviours are matched on the fraudsters’ end with increasingly sophisticated ways to trick the system.

Ensure suppliers are employing consistent methods to combat fraud that include all the basics, such as digital fingerprinting, tracking blacklists and similar techniques, as well as more advanced practices. Making an investment in technology and human expertise to battle fraud is essential for long-term success.

Reduce bias and increase representation: As artificial intelligence (AI) becomes more widely adopted, bias has become a pivotal concern as many believe that it could become baked into AI systems without the right training data. In research, bias can easily sneak into everything from survey design to sample selection. Creating questionnaires that intentionally try to avoid collecting biased data and selecting participants that are representative of the population can help.

This boils down to how sample is sourced, and the profiling points used to segment populations. The Market Research Society’s Representation in Research group is addressing this, looking to improve representation of groups that are often underrepresented. They recommend expanding profiling points for participants to go beyond things like gender and age, and include other points like identity, region, social grade, ethnicity, sexual orientation and physical disability and/or mental health.

Improving respondent experience and engagement: This is a big one, with many facets. There are a wide number of practices, extensively covered by experts in the field, that can help to create positive and meaningful interactions between researchers and survey participants.

Survey design is mentioned above in relation to bias and for engagement, things like length, question type, user friendliness and mobile-first design are important. Providing proper rewards for an individual’s time, treating each person like a human being and protecting privacy are also vital. The bottom line: bake respect and consideration into the process, so participants remain willing, engaged and provide the best possible feedback – building data accuracy and quality from the ground up.

Data quality checks throughout the process: Employing real-time monitoring and applying data validation checks as survey responses are collected can help to identify and address data errors or inconsistencies before it is too late. Being proactive during this stage of the process, and monitoring things like response rates, completion times and overall data validity, greatly impacts the quality of the data coming out of any survey.

Likewise, employing the right techniques and technology to analyse and report on the data after it is collected can impact quality. But the quality data coming into this stage is the most important. The insights we deliver in the analysis and reporting stages are only as good as the data they represent, so we must ensure it is reputable and trustworthy by employing techniques along the entire process that are hyper focused on quality.

The only other piece that should be considered with data quality is maintaining a human element across your processes. The human feedback loop can help organisations pick up on certain elements that just feel wrong. Our intuition, while it’s not always foolproof, is really difficult for AI and technology to replicate, and is something that should not be overlooked.

Horst Feldhaeuser is group services director at Infotools