A Real-Time and Historical Threat Assessment, Profiling and Prevention System.This is our answer to, how we can counter fraud within market research panels.
Online Surveys have greatly reduced the time, effort and budgets that was once involved in market research projects to a large extent but online research comes with its own set of challenges that, researcher must overcome in order to deliver insightful market research. One of the most crucial challenge is to get high quality samples from online panels, most of which are prone to manipulation by malicious actors for increasing their rewards for completing surveys and this puts in question, the data quality of samples from those panels and also the result of your market research project which is based on samples from those panels.
To prevent bad actors from accessing surveys, we developed a machine learning and artificial intelligence based multi-tier respondent risk scoring system, which works in at every stage of panellist interaction with our panel, from sign-up to survey attempt to cash-out and maintain a risk profile of every panellist.
Get Started
Similarly, this technology keeps track of every respondent from third party sample providers, which goes through our systems and maintains a risk profile on them as well, so we are always in control of quality of respondent no matter where they come from.
Get Started
A panellist risk profile is updated with every interaction with our panel, on the basis of data points, we capture from the panellist’s device like time-zone, language, device type and which are given by panellist themselves like demographics, owned devices, preferences, habit, and many more.
Get Started
This risk profiling is then used in real-time to assess probability of panellist defrauding the system whenever any panellist is trying to attempt any survey from our panel and are blocked from accessing the survey if their risk score is higher than the threshold risk score. This technology is used in tandem with variety of other fraud prevention techniques to make sure only qualified and engaged panellists are sent to live surveys.
Get Started
We have developed a machine learning and AI based technology which can eliminate malicious respondents before entering surveys.
After extensive development and testing, this technology is first put in use in January 2022 and what we have observed is that, there exists a huge number of low-quality panellists which has really plagued the whole online market research industry as these panellists are signed up with multiple online research panels for sole purpose of maximising their rewards and give very little thought towards giving quality responses in online surveys. After running over 500 surveys with over 29 different panel companies and 4 sample marketplaces, we have found some disturbing trends.
Over 41% of all respondents from these panels rated high risk by our systems.
Only 27% of all respondents were of medium risk to low risk.
31% of all respondents were registered with more than 7 panel companies.
34% of all respondents were registered with between 7 to 3 panel companies.
Only 17% of all respondents were registered with only one panel company.
You can download our panel book to view our panel numbers and know more about our offerings, so you have all the information you need to make an informed decision to make us your preferred sample provider.