Mission Update: ReDem 3.0 Activates a New Line of Defense Against Survey Fraud
Bots, click farms, AI-generated responses, and professional survey fraudsters have leveled up—and are quietly infiltrating our data. What seems harmless on the surface is, in truth, a targeted attack on the quality of our insights.
ReDem reinforces the defense with focused AI power
With Version 3.0, we’re deploying a powerful upgrade to our data quality protection—built to uncover complex fraud strategies, hidden patterns, and intelligent manipulation.
Don’t miss it: May 28 – 02:00 PM (BST)
Sign up for the launch webinar and see the new version in action. Learn how to upgrade your quality assurance and defend yourself against modern threats.
Why Traditional Checks Can’t Keep Up Anymore
Many traditional methods like trap questions, VPN filters, and CAPTCHAs—once reliable protective measures—have become little more than toothless checkpoints. At the same time, AI-generated responses have become so linguistically convincing that it's increasingly difficult to tell them apart from those written by real people.
Survey fraud today is no longer “obvious and loud”—it’s subtle, cleverly hidden behind deceptively human answers. To counter this development, ReDem has developed a new version of its quality assurance system, which will be officially launched at the end of May.
What's New in ReDem 3.0
1. Behavior-Based Review of Open-Ended Responses: The Behavioral Analytics Score (BAS)
ReDem 3.0 introduces the Behavioral Analytics Score (BAS), which analyzes typing behavior in open-ended survey responses. By analyzing factors such as keystroke rhythm, pauses, and input speed, the BAS helps distinguish between natural human typing and responses that may have been artificially generated or inserted via bots.
A score is assigned on a 0 to 100 scale, with lower scores reflecting artificial, non-human input or copy-pasted answers. The BAS is intended to support researchers in identifying cases where AI-generated or scripted responses might bypass traditional quality checks.
The BAS classifies input behavior into categories such as:
Copy-Pasting
Unnatural Typing (e.g. bots or automated scripts)
Low-Effort Human Typing (inconsistent but human)
Auto-Complete Enhanced Typing (manual input + auto-complete)
Natural Human Typing (typical manual input)
"As AI-generated responses become more sophisticated, evaluating how an answer is entered, beyond just what is written, adds another important layer to data validation.” — Florian Kögl, CEO of ReDem
2. AI-Based Consistency Check: The Coherence Score (CHS)
The Coherence Score uses artificial intelligence to evaluate the logical consistency and plausibility of responses throughout the entire interview—regardless of question type. Contradictory answers are automatically identified without the need for specially programmed control or trap questions.
For example, the CHS flagged a case where a respondent stated they prefer to listen to podcasts on Spotify but failed to recognize the brand in later questions. In another case, a participant claimed to frequently drive a specific car model — but the CHS uncovered that the vehicle had only been announced and was not yet available on the market at the time of the interview.
Each interview is assigned a score between 0 and 100that reflects the internal consistency of the responses. A lower score indicates lower coherence and multiple contradictory or implausible answers. The AI behind the CHS also generates a brief explanatory statement indicating why a particular score was assigned and which questions showed inconsistencies.
The AI behind the CHS also provides a brief explanatory statement that clarifies why a particular score was assigned and in which questions inconsistencies occurred.
“In a conversation with a potential client, we discussed their existing quality checks—and the blind spot that remains: there was no method to verify whether answers within a survey actually made sense together. Anyone who’s ever tried to do this manually knows: checking for internal consistency across responses is extremely time-consuming and mentally taxing.” — Florian Kögl, CEO of ReDem
Neues Design & leistungsstarke API
With ReDem 3.0, the platform not only introduces new quality checks but also delivers a completely redesigned user experience. The modern interface makes it easier than ever to understand and apply the results of data quality checks
At the same time, the new, more powerful API enables seamless integration with existing survey tools. Large volumes of data can be evaluated efficiently and in real time for quality.
Don't Miss the Official Launch Webinar
Wed, May 28th | Online | Register now
Agenda:
- Market Overview & Fraud Trends Current industry challenges
- NEW Quality Checks: Deep dive into the two new quality checks
- Live Demo – Interface & API: Practical use cases and workflows
- Q&A: Direct answers to your questions
