Data quality – Reliable responses
In our fourth article on Vypr’s data quality, we focus on reliable responses. Vypr ensures that the insight you use is as robust and reliable as possible by concentrating on the 5 quality pillars below.
In this article, we delve into the reliability of responses from our community, discussing a few of the tactics that we use to ensure that consumers are providing truthful and honest responses and the measures we have implemented.
The first step is recruiting new members to our community via the mobile app. We have made the sign-up process straightforward; consumers add their personal details and can answer steers right away. The first steer questions (we call demographics) are all about building up a detailed individual profile, but this raises the question of whether consumers are genuinely who they claim to be. Fortunately, we have a range of strategies in place to tackle this. Our targeted recruitment is via acquisition partners who have done lots of profiling and validation checks themselves. This ensures that our community members are authentic and reliable.
Our steer questions have a character limit to ensure they are easy to understand and answer. Most of our steer types only require the consumer to click on one option or on a line. After answering one question and collecting their points, they can exit the app if they wish. This streamlined process encourages consumers to provide accurate responses, as it is not overly complicated or time-consuming.
We know that mistakes can happen, so we don’t want to jump to conclusions and assume that everyone is being dishonest. That’s why we run a series of integrity steers every month that help us identify any bad actors and ensure that our panel is as reliable as possible. These simple steers are factual, with only one correct response, for example, “how many candles are on this cake?” Consumers who answer incorrectly are contacted, highlighting the error, and they receive an official warning about accuracy. Recurrent warnings lead to an automatic ban from the panel, removing these respondents.
With our Free Text and Sentiment steers, we allow consumers to type in a response. We allow our customers to flag a response for review if it doesn’t make sense (like “rjhghgdtfertrtr”), or if a response is unrelated to the question asked. Consumers are again informed of the error and warned that further inaccuracy could lead to being banned. This information then gets added to our steer violation tracker, which helps us monitor and maintain the quality of responses within our community.
In conclusion, we hope that this article allows you a better understanding of what it means to have reliable data. Numerous internal and external checks are put into effect to ensure that only the most reliable responses are reported back to our customers. By focusing on recruitment, making it easy for consumers to respond accurately, and combatting those who don’t, we strive to provide our customers with the most robust and reliable data possible.