In a digital environment where attention spans are measured in seconds, traditional metrics like clicks and views offer an incomplete picture. They tell us what users did but not what they saw, felt, or understood. Which parts of the creative actually registered with them? Did the ad spark interest, confusion, or indifference?

That’s where online neuroresearch comes in. By analyzing visual attention, emotional response, and cognitive load, marketers can move beyond surface-level performance to discover how audiences truly engage with content. The result is a sharper creative, stronger messaging, and a deeper understanding of what drives real impact.

Technology Reveals What Traditional Metrics Miss

With today's digital tools, it’s possible to conduct neuroscience-based research entirely online — no lab or specialized equipment required. Technologies like remote eye tracking, facial coding, and EEG sensors can now be deployed directly from a user’s computer or mobile device in a natural and familiar environment.

This shift enables marketers to evaluate a wide range of assets, from video ads and social media content to display banners, landing pages, and full user journeys. By going digital, researchers get instant insight into what captures attention, what emotions are triggered, and how cognitively taxing the experience is, all without leaving the screen.

Actionable Insights Before Launch

Unlike traditional metrics that show what happened, digital neuroresearch helps us understand why it happened. We can identify which visual elements attract attention, when users disconnect or feel confused, and which moments spark positive or negative emotional responses.

These insights empower teams to make data-driven adjustments to campaigns before they go live, such as reworking layouts, refining messaging, tweaking video pacing, or emphasizing emotionally resonant elements. It’s especially effective for early-stage concept testing, A/B comparisons, message validation, and UX optimization.

Seeing Beyond the Obvious

At ThinkNow, we integrate these tools into digital studies to capture real-time, real-human responses. That is, what users see, feel, and process as they interact with advertising campaigns. It’s a smarter way to move beyond vanity metrics like clicks and views, and uncover insights that drive ad effectiveness.

If you’re interested in applying this technology to your next study, reach out. We’d love to help you discover what your campaigns are truly communicating.

Contact: ventasfullservice@thinknow.com

At ThinkNow, we’ve spent over a decade building DigayGane^TM, our proprietary online consumer panel designed to connect brands with diverse, hard-to-reach audiences in the U.S. and Latin America. Today, DigayGane^TM powers thousands of interviews monthly and fuels insights for Fortune 500 companies, government agencies, and media partners alike. But behind the clean dashboards and seamless user experience lies a constant battle: maintaining data integrity in an ecosystem vulnerable to bots, bad actors, and evolving fraud tactics.

This blog shares what we’ve learned over seven years of defending the integrity of the Hispanic online panel. These lessons have shaped DigayGane^TM into one of the most trusted multicultural research panels in the industry.

The Paradox of Accelerated Growth

When I joined the company in late 2017, DigayGane^TM was in the midst of scaling its reach and capabilities. We celebrated every new signup because expansion meant greater statistical representativeness, better segmentation, and, of course, higher revenue. However, the joy brought by volume came with a silent threat: fraud. The success of our panelist recruitment campaigns—referral programs, programmatic ads, social media promotions—turned into open hallways for bots, cloned accounts, and users who discovered how to “farm” rewards without adding any value.

As fraud threats evolved, so did the need for a more strategic approach to panel protection. Being appointed Panel Security Director wasn’t just about guarding credentials. It meant designing a socio-technical system that rewarded genuine panelists while deterring, detecting, and eliminating opportunists. This blog chronicles that journey, from early failures and hard lessons to the solutions that today form a highly effective three-phase model.

I’ll also outline the challenges on the horizon, like the rise of generative AI to global data regulation, and the strategic ideals that, in my opinion, will shape the next decade of digital market research.

Understanding the Anatomy of Fraud

In 2018, our dashboards showed seemingly healthy indicators: completion rates above 80% and average survey times below estimates. It was tempting to call it a “success.” The wake-up call came when blatant inconsistencies, contradictory and nonsensical answers, began to appear. This forced us to acknowledge that speed is not always synonymous with efficiency. Sometimes it’s pure fabrication.

Identified Fraud Typologies

1. Mass registration bots: scripts that auto-fill forms and confirm disposable emails.
2. Human “survey farms”: groups exchanging credentials and rapidly responding with copy-pasted generic text.
3. Incentive hunters: real users creating dozens of identities to multiply redeemable points.

The first lesson was paradigm-shifting as it revealed that fraud is not just a technical issue, but also an economic and sociological one. The attacker values the incentive and assesses the cost of outsmarting us. That’s why our countermeasures had to raise the “cost” of attacking while minimizing friction for legitimate users.

Phase 1 – Reinforcing the Entry Gate

Two-Factor Authentication (2FA) at Signup

We adopted a simple approach. The more dynamic the verification, the lower the chance of automated replication. Implementing SMS-based 2FA resulted in a 38% reduction in the creation of profiles using temporary emails within the first three months. The key metric wasn’t just the drop in fraudulent volume, but also the conversion rate and retention. After some UX improvements, we lost just three percentage points of legitimate signups, an acceptable cost for the benefit gained.

Real-Time Geolocation

We inserted a microservice that, before confirming registration, validates the IP and compares the declared country with physical location. Of 10,000 daily attempts, 7% came from VPNs jumping regions to access higher-paying studies. Automatic blocking, combined with manual review, created a visible deterrent: after six weeks, proxy attempts dropped to 2%.

Welcome Call

Many in the industry see welcome calls as costly. For us, it was qualitative gold. Over 90% of respondents demonstrated genuine intent. Evasiveness was highly correlated with low-quality scores in later rounds. The call also revealed linguistic nuances that enriched our demographic questionnaire. For example, we learned that certain anglicisms confused lexical verification algorithms in the southern border area near Mexico. We retrained our models and improved dialect matching accuracy by five percentage points.

Phase 2 – Ongoing Panel Monitoring

Geo-Verification via Newsletters

We turned the monthly mailing into a digital “proof of life.” We added tracking pixels to log opens and IPs, without compromising privacy (only storing partial hashes). If a panelist fails to open three newsletters or opens them from inconsistent locations, they are quarantined. This practice reduced zombie email inventory by 26% and cleaned our base for targeted campaigns.

Device Fingerprinting

We implemented device fingerprinting, which combines user-agent, screen resolution, time zone, and 40 other attributes. When the same device tries to register multiple identities, the system applies elasticity: two profiles are allowed (e.g., siblings), but three or more trigger an audit flag. Thanks to this policy, we detected a cluster in China operating 120 profiles pretending to be in California. We dismantled it in under 48 hours.

Adaptive Captchas and S2S Redirects

We migrated from reCAPTCHA v2 to an adaptive system triggered only on suspicious events, preserving ease of access for legitimate users. In parallel, we consolidated all survey redirects into an encrypted SHA-1 S2S bus. Traffic no longer exposed tokens in the frontend, and we blocked malicious URL injections that were stealing incentives. Today, every survey completion generates a checksum validated in real-time, saving 60% compared to the previous architecture.

Phase 3 – Post-Survey Safeguards and Community Reinforcement

Data Reconciliation

We trained a gradient boosting model to evaluate semantic coherence, item-level response time, and cross-survey similarity. Responses in the bottom 10th quality percentile are queued for human review. We discard an average of 3.2% of interviews per study, resulting in a 12-point increase in our Net Promoter Score and higher client satisfaction in 2024.

Social Media Groups as a Strategic Trust Layer

We created closed communities on Facebook and TikTok where top panelists receive previews of new projects and participate in AMA sessions with researchers. This “trust ring” serves two functions:

• Social identity: peer pressure discourages fraud; it’s hard to lie when your reputation is on the line.
• Early warning: members themselves report fake links or impersonations. This is how we detected an “incentive phishing” scam on WhatsApp and deployed countermeasures within two hours.

Emerging Challenges: Facing the Next Tech Curve

1. AI-Powered Synthetic Identities: Generative models can now produce hyper-realistic faces and voices. I foresee attackers generating “digital twins” with seemingly valid documents. In response, we will invest in passive biometric verification—specifically, micro-expression analysis in video selfies—and decentralized proof-of-personhood using blockchain technology.
2. Contextual Content Automation: Tools like GPT can generate coherent text to fill open-ended questions, bypassing traditional lexical filters. It will be essential to combine stylometric analysis with AI signature detection (perplexity, repetition, burstiness) and, most importantly, change the type of question: incorporate multimedia, local-context microtasks, or verifiable personal references.
3. Privacy Regulation and Multi-Jurisdictional Compliance: With the EU AI Act and national frameworks, such as the revised U.S. Federal Data Protection Law, coming into force, we’ll need to constantly audit the data custody chain. One strategy will be building “clean rooms” where analysis occurs without exposing PII.
4. Panelist Fatigue and the Attention Economy: More filters mean more friction. The challenge will be choosing meaningful friction. We’re piloting variable incentives based on “historical quality” and transparent gamification: the more reliable your track record, the fewer authentication steps you see. This turns security into a positive experience.

Recommendations for Industry Peers

• Security equals UX: if your defense layer causes 20% abandonment, you haven’t won—you’ve just traded fraud for attrition.
• Iterate fast on “micro-fraud”: resist the urge to build the “perfect wall.” Target the most profitable weak spot this week, measure, learn, and move on.
• Culture of shared data: involve Support, the Product PM, and the Sales team in metric design. Each sees a different angle of the same diamond.
• Transparency with clients: sharing your anti-fraud policy, including numbers and mistakes, builds trust and reduces future external audits.

From Fortress to Resilient Ecosystem

Now active in 17 countries, DigayGane^TM handles thousands of interviews daily, of which less than 3% are flagged for review and only 0.6% are discarded due to confirmed fraud. But the most significant achievement isn’t statistical—it’s cultural. The company now understands that security and quality are inseparable from the value proposition, not just technical appendices.

My journey as Panel Security Director has taught me that the fight against fraud is a relay race, not a sprint. To professionals just starting on this path, I’d say, invest in people before tools, document every hypothesis, celebrate small wins, and embrace constructive paranoia. Because, in the end, security is the science of anticipating others’ creativity, and creativity, both good and malicious, never rests.

Want to know more about our advanced anti-fraud solution? ThinkNow Shield combines cutting-edge AI and proprietary tools to safeguard your data. Learn more.  

How government-funded university research has shaped the insights industry

The market research industry is built on innovation. Whether it’s new ways of collecting data, developing predictive models or finding more accurate ways to understand human behavior, much of what we use daily as market researchers can be traced back to one powerful engine of innovation: government-funded university research.

As an industry, we often champion advancements like synthetic data, online panels, machine learning and behavioral science. Yet we rarely acknowledge that these innovations stem from university research, which relies heavily on public and institutional funding. Advocating for its continued support is essential because many of the tools we now take for granted wouldn’t exist without it.

Here are three examples of government-funded university research that have shaped the modern market research industry.

1. Synthetic data and machine learning models

The emergence of synthetic data solutions over the past few years has completely transformed market research. Synthetic data allows researchers to simulate realistic consumer behaviors while protecting privacy, expanding datasets and reducing costs.

Much of this would not have been possible without early academic research into generative models like GANs (generative adversarial networks). GANs were first proposed by Ian Goodfellow and his colleagues in a 2014 paper while at the University of Montreal, an institution that received significant government research funding through Canada’s National Sciences and Engineering Research Council. Similarly, the foundations of generative AI that fuel synthetic data creation today, such as transformer models, were first explored in depth at publicly funded institutions like Stanford and MIT.

These academic breakthroughs, originally meant to solve theoretical problems, have been adapted and commercialized by synthetic data providers across the market research space. If public funding for these early experiments had not existed, the synthetic data solutions we now rely on might still be a decade away, or not exist at all.

2. Online sampling and survey methodology

The shift from telephone-based to online-based sampling was one of the most significant disruptions in market research over the past 30 years. But building reliable online samples required understanding how people behave differently on the internet compared to traditional media.

This understanding was made possible by research from universities like the University of Michigan and Stanford, which conducted early experiments in online behavior and social science, often with grants from the National Science Foundation and other government agencies.

Foundational concepts like nonresponse bias in online surveys, sampling representativeness in digital populations and even online panel recruitment strategies emerged from peer-reviewed academic work. Today’s leading panel companies and sample providers owe much of their methodology to these early university studies, which were often theoretical when they were first published.

3. Behavioral science and choice modeling

Behavioral economics is now a mainstay in market research, shaping everything from questionnaire design to advertising testing. However, behavioral economics itself and key methods like conjoint analysis and discrete choice modeling originated from university researchers, funded by government grants.

For example, Nobel Prize winner Daniel Kahneman, who developed many of the concepts behind behavioral economics, conducted his research at Princeton with funding from U.S. government agencies. Similarly, conjoint analysis, now a standard method for testing product features and pricing, was developed at the University of Pennsylvania and the University of Michigan, supported in part by government-funded research grants.

Today, almost every brand study, advertising test and product optimization project in market research benefits from behavioral science. Without these original academic innovations, many of the most effective techniques we rely on would not exist.

Championing university research

As an industry, we must recognize the importance of government funding for university research. The innovations that drive market research forward, like synthetic data, new sampling techniques and behavioral science breakthroughs, all depend on a pipeline of academic discovery.

When public investment in university research is threatened, it isn’t just theoretical fields that are at risk. It’s the foundation of our industry’s future. The ideas, breakthroughs and jobs that have long driven innovation depend on sustained support. Without it, we risk slowing the next wave of innovation before it even starts.

Our role as industry leaders, vendors and practitioners should be to champion policies and initiatives that protect and expand investment in academic research. We must recognize that today’s research paper is tomorrow’s industry disruption. 

Market research does not exist in a vacuum. It exists because of decades of research that government funding made possible. If we want to continue delivering insights that matter, we must collectively advocate for funding the university research that makes it all possible.

This blog post was originally published on Quirk's Media.

In the digital age, consumers no longer interact with brands through a single channel. Today, a single customer might discover a product on Instagram, research it on a website, receive a promotion via email, and finally make the purchase in a physical store or on an app. This fragmented and dynamic behavior is what we know as omnichannel.

But what does this mean for those of us in market research?

From Single Channel to Omnichannel Consumer

Traditionally, market research focused on more linear touchpoints. Today, the challenge is to map a user experience that unfolds across multiple platforms, devices, and moments. Omnichannel has transformed not only the way consumers shop but also the way researchers study them.

It is no longer enough to ask what they buy or where they buy it. We now need to understand how consumers move between channels, when they prefer one over another, and why they make certain purchase decisions in specific contexts.

Market Research for the Omnichannel Era

Let’s look at what market research offers in this new landscape.

• Consumer Journey Mapping: Qualitative methodologies (in-depth interviews, ethnographies, focus groups) and quantitative approaches (tracking studies, mystery shoppers) allow researchers to build a 360° view of omnichannel behavior, including everything from surveys to analysis of digital behavior data. We can also integrate sources such as web analytics, geolocation data, and sentiment analysis on social media to complete the consumer story.
• Smarter Segmentation: The omnichannel consumer is not homogeneous. Research helps identify user profiles: those who compare online and buy in-store, those who only shop via apps, or those who combine channels depending on the type of product. It also allows us to classify consumers by their level of digital engagement, price sensitivity across channels, or loyalty to certain platforms.
• Brand Experience Optimization: Understanding which channels consumers prefer and how they interact with each one allows for more personalized and consistent strategies, which translates into greater satisfaction and loyalty. This includes identifying friction points in the purchase process, inconsistencies in brand messaging, or a lack of integration between digital and physical channels.
• Real-Time Measurement: The omnichannel environment demands agility. Tools such as interactive dashboards, trackers, and online surveys make it possible to monitor consumer behavior almost in real time. In-app surveys, post-purchase experience assessments, and transactional data analysis also provide insights that can be quickly activated.

Omnichannel and Local Insights

Understanding omnichannel behavior requires localized approaches in markets like Latin America, where digital adoption is growing but diverse. For example, in some countries, WhatsApp is key, while in others, e-commerce apps or marketplaces dominate the scene.

This is where culturally contextualized market research becomes essential. It’s not just about knowing what consumers do, but understanding why they do it based on their social, economic, and digital context. A middle-upper socioeconomic consumer in Mexico City may trust delivery apps more, while someone in rural Peru might prefer informal commerce or local fairs, even if they saw the promotion on social media. Without understanding these nuances, any omnichannel strategy remains incomplete.

Research to Integrate

The key takeaway is this: omnichannel is here to stay, and with it comes new opportunities to gain deeper insights into consumer behavior. Brands that align their marketing strategies with actionable insights from solid market research adapted to the omnichannel environment will be the ones that stand out.

Because in a world of multiple channels, the true differentiator remains customer knowledge. And today, that knowledge requires listening and connecting the dots between every click, conversation, and step in the consumer journey.

Imagine market research as navigating a vast ocean. For years, we've used simple maps – surveys and panels – to guide us. But the ocean is changing, new currents are emerging, and those old maps just aren't enough anymore. That's where the new online sampling comes in – it's like having a smart compass that shows you exactly where to go and enabling researchers to collect data from diverse, geographically dispersed audiences quickly and efficiently. Removing barriers like travel constraints and logistical delays offers a more accessible and cost-effective way to reach the right respondents. Think of it as a “smart compass” getting you precisely where you want to go.  

What's different now with online sampling?  

Think of it this way:

• More Than Just Surveys: We're not just asking people questions anymore. We're listening to what they say on social media, looking at what they buy online, and even getting clues from their smart devices. It's like having a network of spies giving us real-time updates.
• Making Sense of the Mess: All this information can be overwhelming. But new tools can sort through the chaos, finding what’s important and ignoring the rest. It's like having a super-smart filter that cleans up the data.
• Predicting the Future: These tools can even predict what might happen next. They can spot trends before they become big news, helping businesses stay ahead of the game. It's like having a crystal ball that shows you what's coming.
• Instant Answers: We don't have to wait weeks for results anymore. We can get answers almost instantly, allowing businesses to react quickly to changes. It's like having a real-time weather report for the market.
• Talking Directly to You: Instead of sending out generic surveys, businesses can now talk to specific groups of people with questions tailored just for them. It's like having a personalized conversation with each customer.
• Mobile is King: Most people use their phones for everything, so research needs to work seamlessly on mobile. It's like having a compass that works perfectly on your phone.
• Being Responsible: As we collect more data, it's crucial to be careful and respectful of people's privacy. It's like having a code of ethics for how we use our smart compass.

What's in it for Businesses?

This new approach to online sampling can help businesses:

• Understand Customers Better: They can get a much clearer picture of who their customers are, what they want, and why they do what they do.
• Make Smarter Decisions: With better information, they can make better choices about what products to create, how to market them, and where to invest their money.
• Move Faster: They can react quickly to changes in the market, staying ahead of the competition.
• Sell More: By understanding their customers better, they can create products and services that people actually want to buy.
• Stay Ahead of the Curve: Businesses that use these new tools will have a big advantage over those that don't.

In short:

The future of online sampling is all about being smarter, faster, and more personal. It's about having a smart compass that helps businesses navigate the ever-changing market and reach their destination successfully. It's not just about collecting data; it's about using that data to make better decisions and build a stronger business.

Shifting consumer dynamics are reshaping how companies and brands connect with their audiences. While market research strategies often account for basic demographics like race, ethnicity, and gender, truly inclusive panels go beyond these factors by creating spaces where individuals feel comfortable sharing their opinions and experiences.

This approach not only enhances the depth and quality of the data collected but also ensures a more accurate and authentic representation of diverse target audiences.

Inclusive Online Panels Checklist

1. Diverse and Representative Recruitment: An inclusive panel starts with recruitment that ensures broad societal representation. This involves identifying and attracting individuals from various demographic and cultural groups. Partnering with local communities, specialized networks, or diversity-focused platforms can be instrumental in achieving this goal. Additionally, it’s crucial to avoid selection bias by ensuring equal opportunities for all segments to participate.
   
2. Accessible Panel Design: Platforms must be technologically inclusive. Incorporating real-time captions, audio descriptions, and compatibility with screen readers expands accessibility and ensures participation by individuals with disabilities. It’s also important to prioritize usability, making interfaces intuitive and functional across devices with varying technical capabilities.
   
3. Inclusive and Neutral Language: The language used in online panels must be carefully considered. Phrasing questions in ways that avoid cultural or social assumptions helps prevent bias. For instance, asking about “the head of the household” might be inappropriate or irrelevant for some participants. Using a neutral and inclusive tone while avoiding gendered language helps create an environment where participants feel both comfortable and understood.
   
4. Empathetic Facilitation: Moderators play a critical role in fostering inclusion. They must be trained to handle sensitive conversations and create an environment where everyone feels safe sharing their perspectives. This responsibility includes identifying and addressing any instances of exclusion or discrimination, whether occurring between participants or involving the moderator themselves. By doing so, moderators can ensure that interactions remain respectful, equitable, and conducive to open dialogue.
   
5. Continuous Evaluation and Feedback: Achieving inclusion in online panels is an ongoing journey that demands intentional effort and consistent evaluation and refinement. Implementing mechanisms to receive feedback from participants helps identify areas for improvement and refine approaches as needed. By doing so, organizations reaffirm their commitment to the principles of diversity and inclusion, which are well-established drivers of innovation and economic growth.

Why Inclusive Online Panels Matter

Adopting inclusive practices in online panels not only strengthens market research strategies but also benefits brands. A more accurate representation of audiences helps inform the development of products and services that genuinely meet their needs. Furthermore, brands that demonstrate a genuine commitment to inclusion often earn the loyalty of consumers who value diversity and equity.

From a research perspective, prioritizing inclusion yields richer, more diverse insights by accounting for cultural norms and other factors that influence survey responses. These authentic insights enable brands to innovate and adapt in a globalized market where diversity is the norm rather than the exception.

Brands that incorporate inclusive online panels into their market research strategies are better positioned to engage audiences over the long term, driving sustained topline growth.

Online sampling has revolutionized the way businesses gather insights and feedback. However, the rise of digital platforms has also heightened the risk of fraudulent activities. In this blog post, we'll delve into the strategies and techniques available to detect and prevent fraud in online sampling.

Understanding Online Sampling Fraud

Let’s start with a definition. Online sampling fraud occurs when individuals or groups manipulate the sampling process to obtain incentives or rewards without providing genuine feedback. By “gaming the system,” these fraudsters create favorable outcomes for themselves to the determent of the research. Some common types of fraud include:

• Professional Respondents: These individuals participate in multiple surveys solely for the rewards, often providing dishonest or fabricated answers.
• Bot-Based Fraud: Automated bots can be used to complete surveys rapidly and in large numbers, skewing the results.
• Identity Theft: Fraudsters may use stolen identities to participate in surveys, compromising data integrity.

Key Strategies for Fraud Detection and Prevention

Preventing and detecting fraud requires a proactive, multifaceted approach leveraging technology and human intervention to identify and eliminate threats effectively. Here are a few strategies to do that:

1. Robust Screening and Validation:
   • IP Address Verification: Monitor IP addresses to identify multiple submissions from the same device or location.
   • Email Validation: Implement email verification processes to ensure authentic respondents.
   • Demographic Profiling: Cross-reference demographic information with real-world data to detect inconsistencies.
2. Advanced Analytics and Machine Learning:
   • Behavioral Analysis: Analyze respondent behavior patterns, such as response times, skipping rates, and consistency, to flag anomalies.
   • Statistical Modeling: Use statistical models to identify outliers and unusual patterns in the data.
   • Machine Learning Algorithms: Employ machine learning algorithms to learn from historical data and predict future fraudulent behavior to inform strategies to safeguard against it.
3. Real-Time Monitoring and Alerting:
   • Dashboard and Alerts: Create real-time dashboards to monitor key metrics and set up alerts for suspicious activity.
   • Immediate Action: Implement procedures to quickly investigate and address fraudulent behavior.
4. Incentive Structure Optimization:
   • Balanced Rewards: Design incentive programs that reward quality over quantity.
   • Tiered Rewards: Offer higher rewards for completing more complex surveys or providing detailed feedback.
5. Collaborate with Industry Partners:
   • Shared Blacklists: Collaborate with other panel providers to share information about known fraudsters and bot networks.
   • Industry Standards: Adhere to industry best practices and standards to maintain data quality.

Conclusion

Fraud is an industry-wide problem, not an isolated event. By collaborating with industry peers and adopting proactive strategies, sample companies can significantly reduce the risk of online sampling fraud and ensure the accuracy and reliability of their insights. As technology advances, so too does fraudulent tactics. To stay ahead of these evolving threats, organizations must invest in robust fraud detection and prevention measures. By doing so, they can drive successful business outcomes for their clients.