4 min read
Have you ever made decisions based on data that turned out to be less reliable than expected? In market research, this can have significant and potentially damaging consequences, at best. Data quality is the pillar on which strategic decisions are made, and in an increasingly digitised environment, ensuring that samples are representative and accurate is more critical than ever.
But how can you ensure that the quality of your online samples is up to the decisions that depend on them? Here are four key points to help you do just that.
Choosing well: the key to a reliable database
How much confidence do you have in the quality of the samples you use in your research? Selecting market research providers that are transparent about the origin of the samples and guarantee their quality is essential to ensure reliable data. It is easy to be tempted to go for the cheapest option, but the consequences of not prioritising quality can be costly in the long run.
A reliable supplier not only ensures on-time delivery, but also aligns itself with high standards of quality and ethics. This includes complying with regulations such as GDPR or CCPA, being backed by certifications such as ISO or ESOMAR, and carrying out continuous quality checks to ensure that samples are representative and free of duplicates or fraud. In addition, transparency is key: a good provider will give you clear details about the origin of the samples and the incentives they use to motivate respondents, ensuring that the data is robust and reliable.
On the other hand, when a provider is unreliable, you will notice a lack of transparency, inordinate incentives that attract disengaged respondents, and a lack of rigorous controls to prevent fraud. If they do not provide clear information, if prices are too low or if they refuse to do pre-audits, it is better to think twice. While it may seem tempting to go for the cheapest, the consequences of working with an unreliable sample can be much more costly in the long run.
Anti-fraud technology: Protect your surveys from fake responses
According to a recent study, up to 18% of respondents in online surveys may be duplicated or fraudulent. Can you imagine making a marketing decision based on fraudulent data? To avoid such situations, Zinklar uses Opinion Route technology. Using the CleanID service, we detect possible fraud before respondents access surveys, ensuring that the data collected is reliable and representative.
When you use advanced fraud detection technology in your online surveys, you ensure that the responses you receive come from real, engaged people. These tools not only detect duplicates or bots trying to infiltrate surveys, but can also identify suspicious patterns of behaviour, such as respondents who respond unrealistically quickly or who show inconsistencies in their answers. By having these filters in place, you ensure that the data you get is more representative and reliable, allowing you to make decisions based on quality insights.
On the other hand, not using this type of technology means taking a considerable risk. Without these controls, the sample is much more likely to be contaminated with fraudulent or disinterested respondents, which distorts the final results. Incorrect data can lead you to make wrong decisions based on information that does not reflect market reality.
Short, mobile-optimised surveys to avoid fatigue
Who hasn’t started an online survey and halfway through you can’t wait for it to end? Studies show that respondent fatigue is one of the main causes of low quality responses. If they are bored or overwhelmed, do you think their answers will really reflect their opinion? Probably not.
At Zinklar, surveys last no longer than 9 minutes, which allows us to keep respondents interested and get more reliable responses. This is especially useful in branding studies, where it is critical to capture consumer impressions quickly and accurately, and the immediacy of responses can make a big difference in brand perception studies.
Are respondents being honest?
Another question we often ask when conducting market research is whether all respondents will be truthful in their responses. Sometimes the incentive may cause them to answer anything to qualify. For example, they may select answers at random or, in some cases, give answers that they believe qualify them for further study or better incentives, without really engaging with the content of the survey.
If respondents are not being honest or are participating only for the incentives, the results are distorted, leading to decisions based on unrepresentative data. Imagine you are conducting a study on luxury products and a significant part of the participants respond without actually belonging to the target group; the results will not only be irrelevant, but could lead to bad decisions.
At Zinklar, to avoid these problems, we implement controls such as pattern detection (known as ‘straightlining’), which seeks to identify those who repeatedly select the same options. We also use questions designed to detect unusual behaviour or inconsistent responses. Such questions, for example, can be ‘trick’ questions to assess whether the respondent is paying attention or simply marking options at random.
The quality of the sample is the cornerstone on which any valid and representative market research is built. Through these four steps, you will not only be improving the reliability of the data you collect, but you will also ensure that the strategic decisions you make are based on solid and accurate information.
