Trick or Treat: Meet the ‘Ghost Complete’

Published Oct 30, 2024

Marie Hense, Global Head of Quality at Toluna
Gene Saykin, VP/Head of Quality Strategy at Toluna

Share this

Something scary is haunting the research industry… Speak to any Research Project Manager about their greatest pain points, and you’re likely to encounter the ghoul of the industry: the ghost complete. 

In the vast corners of the internet, online survey fraudsters are constantly attempting to manipulate survey links. Their goal? To rush through a survey without actually answering any questions, so that they can receive the incentive upon completion. This practice results in a ‘ghost complete’ – and it’s all trick, no treat. Here’s why. 

What is a ghost complete?

In simple terms, a fraudster manipulates a survey’s URL, so the survey system believes they’ve completed the survey. However, when they reach the final survey page, they receive the incentive without having genuinely participated. In the dataset, this person appears as a complete, though their responses are often missing or sparse – like a ghost in a dataset. 

This type of fraud typically occurs when researchers provide a survey link to a third-party sample provider. For instance, if you programme your own survey and then use a panel company or sample marketplace to gather responses, you could fall victim to ghost completes. However, surveys programmed and executed within a single platform (such as Toluna’s Toluna Start) are generally unaffected. 

Why are ghost completes a problem? 

The misrepresentation created by ghost completes causes several issues. For starters, the survey system might assume all responses are collected, only for the analyst to discover missing answers during the data review process. At this point, the project must be reopened to replace the fraudulent entries, which wastes valuable time. 

In another scenario, fraudsters might detect a survey with weak protection and work their way through it, manipulating the system as they go. This results in partial answers that need to be painstakingly cleaned during post-field data processing. 

To put this into perspective: in Toluna’s data, surveys without ghost complete protection experience a clean-out rate three times higher than those with safeguards in place. But the issue extends beyond mere operational inconvenience. The easier the research industry becomes to exploit, the more fraudsters will target it, escalating the problem over time. 

How can you protect your survey against ghost completes? 

Ghost completes and link jumping are industry-wide challenges, but their impact varies depending on how well a survey link is protected. Well-protected projects will rarely see ghost completes, whereas poorly protected surveys can be heavily affected – depending on factors like the survey market or the incentive offered. 

Here are three ways to protect your study: 

    1. Encrypted links: Use encrypted redirect links to make it harder for fraudsters to manipulate the URL and reach the incentive page. The more complex the encryption, the tougher it is for fraudsters to crack. Regularly updating your encryption keys can help maintain security.
    2. Direct system-to-system API integration: A direct API link between the survey platform and the sample provider allows real-time status comparisons. If a participant is marked as a partial complete in one system but a full complete in another, the discrepancy is flagged, which helps in catching ghost completes early. This is the safest and recommended method.
    3. Script your survey on a panel provider’s platform: Surveys created directly within a panel provider’s platform, such as Toluna Start, don’t require external redirect links, eliminating the vulnerability of URLs. 

 

Working together to deter fraud 

The responsibility of protecting the research industry from fraudsters is shared by all. Collaboration and best-practice sharing through initiatives like the cross-association Global Data Quality Initiative are essential steps. Ultimately, though, it’s up to researchers to implement quality measures. 

If you suspect your survey may be vulnerable to link jumping, reach out to your sample provider to discuss how they can support you in safeguarding your study. 

Share this
This site is registered on wpml.org as a development site. Switch to a production site key to remove this banner.