Is synthetic data going to change market research?
For decades, market research has relied on familiar methods—surveys, focus groups, and ethnographies. These tools help brands understand how consumers think, feel, and behave. But they also come with trade-offs: time, cost, and logistical complexity.
Synthetic data, or AI-generated data simulating consumer behavior, is now being promoted for faster, cost-effective insights. Companies like Evidenza are at the forefront, claiming they can simulate human responses using vast datasets, eliminating the need to conduct interviews or surveys. The founders, Brian Watroba, Jon Lombardo, and Peter Weinberg, have been popping up on pods and in several recent articles to share their ideas. The Evidenza bros refer to this as synthetic research while more traditional research firms think of it as synthetic data. I’m not sure what to call it – but it is intriguing.
What does Evidenza do?
Evidenza is the world’s first synthetic research platform. We help you make smarter, faster go-to-market decisions by interviewing and surveying AI-generated copies of your customers. Our platform delivers evidence-based sales and marketing plans swiftly, transforming the way businesses understand and interact with their markets. By leveraging advanced AI, Evidenza offers in-depth insights in hours, not months, enabling businesses to make informed decisions rapidly. (From their FAQ)
But while the promise is exciting, seasoned, experienced researchers urge caution.
“Synthetic data provides directional insights that can be useful for exploration, but it’s not a replacement for rigorous, validated research,” says Michelle Elster, CEO of Rabin Roberts Research.
What Is Synthetic Data?
Synthetic data uses AI to model “synthetic respondents” trained on publicly available and licensed datasets, from product reviews to social media behavior. These simulated personas can then be queried to predict how real people might respond to marketing concepts, ads, or products.
But it’s crucial to understand the limitations.
“Directional answers from synthetic data are more like qualitative insights,” Elster explains. “They are not derived from large, structured samples or unbiased, validated methods. They can help refine hypotheses or identify red flags—but they’re inappropriate for high-stakes decisions.”
When (and When Not) to Use Synthetic Data
Appropriate uses:
- Exploring early-stage ideas
- Refining hypotheses
- Detecting potential issues or reactions before investing in complete research
- Supplementing human insights during fast turnarounds
Risky or inappropriate uses:
- Making multimillion-dollar decisions
- Launching nationally without real consumer validation
- Skipping real-world testing in favor of AI predictions
“Would you base a major business decision on a model that’s never been validated?” Elster asks. “I wouldn’t.”
Serious Limitations
Michelle Elster outlined several core concerns with synthetic data:
- Bias Risk: If the original datasets contain bias, synthetic outputs will reflect and amplify those issues.
- Lack of Emotional Nuance: AI lacks human context and empathy, making it challenging to interpret emotional responses.
- Quality Control: Unlike traditional research, synthetic data isn’t governed by industry standards or third-party audits.
- No Validation: “You can’t verify synthetic data unless you have validated benchmarks,” says Elster. “And often, that’s missing.”
Don’t Confuse Speed with Rigor
It’s tempting to use synthetic data for everything—especially when timelines and budgets are tight. But as Elster stresses, just because it’s fast doesn’t mean it’s right.
“Synthetic data has a place in modern research—but it requires training and judgment to use it correctly. Without that, it’s easy to misuse or misinterpret what the data tells you.”
What Other Research Leaders Are Saying
These views align closely with Michelle Elster’s: synthetic data can be a useful exploration tool when used with care. However, it is not a one-size-fits-all replacement for traditional methods rooted in real consumer engagement.
Even among innovation-forward consultants, there’s a call for balance. Kristin Luck, founder of ESOMAR’s Global Data Council and a long-time advocate for research tech, has noted: “Synthetic data is promising—but only in the hands of professionals who know its limits. It’s not a DIY tool for brands to skip proper validation.”
JD Deitch, Chief Operating Officer of research tech firm Cint, echoes these concerns in an industry blog: “Synthetic data might seem like a shortcut, but the risk of compounding errors or embedded bias is significant. There’s no shortcut to understanding emotional context or spontaneous reactions, often the most valuable parts of consumer insight.”
Melanie Courtright, CEO of the Insights Association, has said publicly that while AI and synthetic data have a role to play in evolving methodologies, they should not replace rigorous sampling and validation standards:
“There is a real danger in mistaking simulated consensus for real-world complexity. Synthetic data can help form hypotheses, but accurate consumer understanding still requires engagement with real people.”
The rise of synthetic data has generated cautious interest across the market research industry—but many established firms remain skeptical of their reliability for critical decision-making.
The Bottom Line
Synthetic data is a valuable tool—but not a magic wand.
It can be a helpful starting point for brands seeking quick, low-cost, exploratory insights. However, for strategic decisions that carry real business risk, traditional, validated research remains essential.
“Think of synthetic data as a supplement, not a substitute,” Elster concludes. “Used properly, it can accelerate learning. If you use it blindly, it can lead you astray.”
Three Key Takeaways for Marketing Professionals:
- Synthetic data is best for directional, exploratory insight, not statistically validated decision-making.
- Bias, lack of transparency, and absence of validation make synthetic data risky for major strategic bets.
- Savvy marketers should blend synthetic and traditional research, using each tool for what it does best.
What’s your view on this topic?
Do you have first-hand experience using synthetic data in your marketing efforts?
What’s your perspective on this topic?
Here is an interview with the Evidenza founders about their work. The synthetic discussion starts at 31:26 into the podcast. Their silly outfits undermined the interesting work they are doing. They dress as Renaissance men to be memorable.
But the conversation is worth listening to if you are interested in this topic and want to learn about some innovations in research through disruptive technology.
This post is written by a food entrepreneur who built a successful multi-million-dollar business from scratch, It took my wife and I thirteen years to get to a successful exit, If you need help, let’s talk.
Connect with Jeff at The Marketing Sage Consultancy. Interested in setting up a call with me? Use my calendly to schedule a time to speak. The call is free, and we can discuss your brand and marketing needs.
If you want to learn more about my new offering, The Trusted Advisor Board, click here. Feel free to email me at jeffslater@themarketingsage.com or text 919 720 0995.
Thanks for your interest in working with The Marketing Sage Consultancy.
This post featured insights from Michelle Elster, CEO of Rabin Roberts Research, a traditional research firm with a long history and heritage supporting marketing and brands.
Photo by Possessed Photography on Unsplash
This post is written by a food entrepreneur who built a successful multi-million-dollar business from scratch, with a successful exit. I have done what I am recommending. If you need help, let’s talk. Connect with Jeff at The Marketing Sage Consultancy. Interested in setting up a call with me? Use my calendly to schedule a time to speak. The call is free, and we can discuss your brand and marketing needs.
If you want to learn more about my new offering, The Trusted Advisor Board, you can click here to learn the details. Feel free to email me at jeffslater@themarketing sage.com or text 919 720 0995. Thanks for your interest in working with The Marketing Sage Consultancy.
