Synthetic Personas Are the New Normal of User Research

Market researchers are using generative AI tools like ChatGPT to create synthetic personas that can mimic, describe, and (sometimes) successfully predict human behavior. The goal is simple: reduce marketing costs while improving user experience. Built with large language models and artificial intelligence, these synthetic participants can potentially replace real people in surveys, interviews, and other research studies.

Traditional market research studies are critical to understanding customers. However, they’re expensive and time-consuming. They need big budgets and human resources, people who are willing to participate in a survey and answer your questions. With marketing costs on the rise and people becoming sensitive about the data they share, market researchers often find it difficult to collect high-quality customer information.

Synthetic personas solve these bottlenecks, i.e., collecting, processing, and analyzing consumer demographic and behavioral data. Although they cannot fully replace the qualitative research involving human interactions, you can use them to supplement your studies and get a general idea of the reaction your products or marketing campaigns will receive. You don’t need a big budget or a dedicated researcher; you just need a clear picture of who your customer is and the right tool to get started with synthetic users.

What are Synthetic Personas?

A synthetic persona, also known as a synthetic user, is a virtual customer created using AI and machine learning algorithms based on user demographics, psychographics, and behavioral data. It provides an accurate representation of a specific audience segment, describing their likes, dislikes, thoughts, preferences, needs, and challenges. They might seem similar to customer personas, but unlike personas, you can actually interact with them (eg, talk to them about their problems).

Synthetic customers are no different from the AI humans created by companies using model generation and sophisticated AI puppeteering. The latter can track eye movements, facial expressions, and pupil dilation to understand human emotions and provide helpful instructions. In some cases, you can also use AI avatars to market your products or services.

Depending on the goal of your research study, you can create a diverse set of synthetic respondents using ChatGPT or a dedicated synthetic research software like Delve AI. They are fairly easy to create and are cheaper than recruiting and conducting research with real people (for instance, Delve AI offers research functionalities at $0.99 per synthetic user). You have complete control over the data fed into the system and can ask follow-up questions to take the interview forward.

FYI, ChatGPT takes a lot of prompting to understand the right customer segment and adopt the related user persona.

For synthetic personas to work, you need to ensure there are certain standards in place. Your data needs to be validated before it’s fed into AI systems. The personas generated also need to be cross-checked, compared, and validated with real-world respondents before you can use them.

How do synthetic users work?

To create a synthetic persona, you need to feed a selected amount of customer data into large language models like Claude and prompt it to adopt the “persona” of an actual customer based on the data you’ve provided. That mostly contains demographic information, like age, gender, and location, along with psychographic attributes, such as personality traits, attitudes, lifestyles, and interests. The persona generated can then be used in customer surveys to give human-like responses.

The accuracy of synthetic personas is directly proportional to the quality of the training data. The higher the quality, the better your personas. If your input data is generic and inaccurate, the output will also be the same.

Data concerns aside, you can use synthetic users to simulate a diverse range of customers and ask them about their needs, pain points, and challenges regarding your product or service. You can use them in qualitative research to get user insights, refine product designs, and test ad creatives across multiple target audiences. One of the best things about synthetic audiences is that you can interview them from the comfort of your home or office, 24/7.

Now, synthetic user panels take this capability a step further. You can run group research studies in regions or scenarios where collecting real respondent data is legally complex.

AI-powered virtual assistants can conduct these interviews on your behalf, sometimes in multiple languages, and analyze responses in real time to ask personalized follow-up questions. Synthetic personas also make it possible for quant researchers to study demographics that would otherwise be unreachable, i.e., niche segments too small or too dispersed to recruit through traditional means.

Synthetic Users vs Real Users

As you can tell from the name itself, synthetic personas are AI-generated, and the other is human. Both differ in the efficiency and quality they bring to the research process.

For starters, you have to recruit human participants. That’s setting up screener questions and waiting for confirmations. Most of the time, you have to chase responses. It’s not the same with synthetic users. You don’t need special moderators and calendars, or be mindful of different time zones.

Second, incentives run at $50–$150 per person for an online survey, $400–$600 per person for an in-depth interview, and $15,000–$30,000 per session for a focus group. Synthetic users cost a fraction of that. Plus, you don’t need to spend hours analyzing response data; synthetic research tools, like the ones from Delve AI, do that for you automatically. You get themes, sentiments, response transcripts, and a full report analysis.

A traditional, complete research study generally takes four to six weeks; a synthetic study covering the same questions can be done in a day.

Now, the quality of research is one point that can be debated. Real people may surprise you with unprompted feedback in ways AI cannot. After all, synthetic audiences mimic patterns from existing data. Qualitative research with actual users, hence, remains the most reliable way to decode unexpected and authentic audience behaviour.

Question: If you can’t conduct research, is it okay to use synthetic personas? Maybe, maybe not. You can always learn something new, as opposed to not learning at all. Yet, basing critical product or marketing decisions on synthetic data alone is not ideal.

Suppose your new feature is a hit with your synthetic customers. You launch it in the market, but it fails with real users. What happens then? You waste resources and lose money. Bottom line: always take synthetic user responses with a grain of salt.

How to Create Synthetic Users: Delve AI + Other ChatGPT Prompts

The process of creating synthetic personas is quite simple. It involves data aggregation, generative AI modeling, and persona generation.

You start with data collection, wherein you gather all sorts of important customer or user data. This can be your web analytics data, CRM records, social audience insights, reviews, ratings, user feedback, survey transcripts, transaction histories, and more. Of course, it needs to be anonymized to maintain privacy. Next, you bring in machine learning algorithms to analyze this data and present synthetic data similar to the source data.

These models segment these synthetic records to create different synthetic personas. Or just a single one, depending on your instructions.

Each persona gives you the traits, characteristics, and behaviors of your desirable and non-desirable audience segments. With AI persona generators like Delve AI, you can find user demographics, job profiles, goals, motivations, pain points, values, interests, jobs to be done, and hobbies. You also get insights into their favorite social networks, brands, websites, tools, marketing channels, and events.

What differentiates synthetic users from traditional marketing personas is their chat functionality. Basically, you can interact with these personas and get their input on a new product feature or marketing copy. You can use AI applications like ChatGPT or tools like Delve AI’s Digital Twin of the Customer software for this purpose. The dashboard is similar to WhatsApp or any other messaging platform. Just go to the platform, select your target persona, and pop the question.

In the following subsections, we’ll discuss how you can create synthetic users via ChatGPT and Delve AI.

1. Using ChatGPT prompts to create synthetic personas

ChatGPT can process vast amounts of unstructured data – numerical data, job profiles, literature, research papers, forums, reviews, and websites – and present it logically. As such, it is a free alternative to create interactive personas that you can use in consumer research.

It’s a simple exercise, really:

• Upload your customer data into ChatGPT
• Direct it to list down the possible segments
• Select the audience segment you want to target
• Prompt ChatGPT to adopt the persona of that segment

You can then ask the persona to be a participant in a survey session. Note that ChatGPT personas are only useful when you have broad customer profiles and not a niche audience. Unless, of course, you have enough qualitative and quantitative data for ChatGPT to form assumptions.

Here’s how your prompts can play out:

Prompt 1: Analyze and structure the customer data provided, focusing on key demographic, behavioral, and psychographic attributes. Summarize and highlight any clear trends or patterns across different customer attributes.

Prompt 2: Using this structured data, list the possible customer segments. Name each segment and briefly describe its defining characteristics, including their interests, purchasing behaviors, and other notable demographic details.

Prompt 3: From these segments, focus on [Segment Name]. Give more details about their major motivations, typical pain points, and potential brand interactions.

Prompt 4: Adopt the persona of a typical customer from the [Segment Name] audience. Describe their day-to-day lifestyle, interests, key values, and specific needs. Assume this persona’s character for further interaction.

There are just a few things you need to remember.

For one, ChatGPT cannot understand meaning, so it cannot be used to create patterns of meaning that synthetic users communicate to you in a study. It does not help you understand the different approaches people take to reach a goal or complete a task. Also, AI cannot use products if you think about it; it can just imagine. You can run tests to validate your product ideas, but every opinion might just be a positive one – ChatGPT personas aim to please the interviewer.

Case in point: A comparative study by Emporia Research between three groups of survey responses – LinkedIn-verified respondents, synthetic users generated by AI using LinkedIn profile data, and users based on AI-generated personas – representing IT decision-makers found that B2B synthetic users generated by AI show a strong positive bias compared to real survey respondents. They followed a herd mentality, and the quality of insights was not that great either.

2. How to create simulated users in Delve AI

Generic ChatGPT personas can only do so much. With Delve AI’s Synthetic Research software, you can generate hundreds or thousands of AI-powered synthetic personas that mimic your target audience or user base.

The Synthetic Research software enables you to:

• Create synthetic panels
• Run synthetic surveys and interviews
• Conduct an AI-assisted focus group discussion

Synthetic panels consist of simulated users based on your existing personas, created with Delve AI, and can be used to conduct user research studies.

To generate simulated users, log in and go to the synthetic market research software. Then, click Panels in the sidebar menu and buy the number of users you need (eg, 100). Once done, you’ll be directed to a screen where you can select the persona(s) or persona segments you'd like to use to generate your synthetic audience. You can also recreate niche audience panels using Filters and target users by attributes such as gender or location.

As shown below, the software has built 100 simulated users based on that persona segment. Each card comes with a “Start Chat” option, which you can use to engage with them.

Compared to traditional research methodologies, Delve AI’s simulated users are a cost and time-efficient way to understand different audience segments before you run a campaign or launch a new product in the market. You can simply upload a CSV file with different types of survey questions, like multiple-choice, rating-scale, and image-based questions, and get quick results.

Synthetic Personas in Market Research: Use Cases and Examples

Marko Sarstedt exemplifies the use of synthetic personas in market research and design in his research on silicon samples. He used them to get feedback on survey questions and refine packaging design. Besides his applications, synthetic users can also be used for other marketing and product-related use cases.

Marketing, segmentation, and branding

You can create two types of synthetic personas: those representing individual customers and those representing groups of customers. Each enables you to get information about user behavior and preferences without accessing private customer data. It’s a good starting point for professionals who are not familiar with their target audience. You can develop different customer segments to test your campaigns and identify which ones respond best.

Consequently, you can customize your marketing assets, content, and copy to suit the preferences of your target audience segments.

Simply upload your marketing plans, graphic designs, videos, and ads to get feedback in the campaign development process. After uploading your material into the model, ask the synthetic user for inputs, like:

• Does this video ad invoke happiness or sadness (replace it with your desired emotion)?
• Where would you like to see this ad? On billboards, Facebook, Instagram, or LinkedIn?
• Do you like the overall message that the ad copy tries to convey?
• How likely are you to respond to this email campaign?

You can further add performance data of past marketing campaigns to check what worked and what didn’t – it gives the AI model more context to work with. You will have to tweak your prompts to get efficient outputs, but the results are well worth it. Who knows, it might just help you create an ad that goes viral!

Brand positioning and messaging are other things you can look at. Run your value propositions through multiple synthetic segments and check whether the messaging resonates with audience values.

In product development, design, and user testing

Synthetic users help product teams empathize with their users and create new products, features, and functionalities that align with consumer needs and expectations. You can simulate usage scenarios to identify potential problems and ensure that your product functions well in different situations.

Brands can use them to validate their product designs before launching them in public. For example, they can test how well the user interface and interactions work in a relatively risk-free environment (since they won’t have to test it on actual customers). Doing so will help them refine the overall product user experience (UX) and greatly reduce the likelihood of issues popping up when the product is officially launched, bringing in new users and reducing customer churn.

You can also use synthetic personas much earlier in the process:

• Map out the questions a user arrives with,
• Find out the specific tasks they’re trying to complete,
• Identify their emotional state at first contact, and
• See what a successful outcome looks like for them.

To learn more about the applications of synthetic users in marketing and user testing, read our post titled “13 Practical Use Cases of Synthetic Research by Delve AI.”

Are Synthetic Personas Reliable? The Problems with Synthetic User Research

No matter how fancy AI gets, there are way too many limitations of synthetic users than meets the eye. Synthetic personas are built using machine learning models, so they are very useful in scenarios where actual customer data is scarce or limited. However, you need to keep some things in mind before incorporating them into your workflow.

To begin with, synthetic personas have a tough time capturing the full range of human behavior and actions; their responses can be too simplified, generic, or logical. Human beings are not always logical – we are influenced by various internal and external factors (like our personal experiences, personalities, opinions, and values) that govern the way we function and respond to things.

Artificial intelligence and machine learning systems, the foundations of synthetic personas, are not yet able to comprehend these factors. For instance, AI respondents always want to please the person running the study, fully complete any given task, and answer questions in the affirmative (a phenomenon called sycophancy). They provide one-dimensional responses, and seem to care about everything; providing a long list of needs, challenges, and interests. People don’t have such lists – they care about specific things, and other things don’t even cross their minds.

In their paper titled, “A Manager and an AI Walk into a Bar: Does ChatGPT Make Biased Decisions Like We Do?” Chen et al. found that ChatGPT is often unable to replicate effects that characterize customer behavior, like the sunk cost fallacy. People continue bidding on an item even after the price has gone beyond what it's worth, just because they want to win the bid. There is no hierarchy in their responses – how do you understand what’s important and what’s not?

Another concern is the data used to create synthetic personas; if the training data is biased, the responses will contain and perpetuate stereotypes. Data scarcity also makes it difficult to represent underrepresented or marginalized users. Validation is tricky, too. It's not easy to test synthetic users against real-world situations and outcomes. They need to be constantly monitored, updated, and verified with real user data. We are talking about anonymized data that does not reveal sensitive customer information.

There’s a more structural issue, too. Most LLM-based persona generation is optimized for density matching. So the tools generate the most probable customer, not the full range of customers that could plausibly exist. This works fine if you just want to know your core audience, but it underrepresents minority behaviors, edge cases, and rare user types. Research published in 2026 (Paglieri et al.) confirmed that even when LLMs are explicitly asked to generate “diverse personas,” the output collapses around a narrow cluster of stereotypical responses.

Even though you have all these things crossed off your checklist, integrating synthetic users into your existing systems can prove to be difficult. Technical and data problems aside, convincing skeptical stakeholders about their usefulness might be your biggest challenge. What’s worse, your company might get accustomed to using synthetic customers for user research and not want to invest in the real thing.

The Future of Synthetic User Personas

Synthetic personas reveal unknown insights about your customers. But can we trust synthetic data? Some say synthetic users are the future of market research; others say they lack the nuance and variation generally found in human respondents. Yet, in a world where market research costs an arm and a leg, synthetic personas are a cost-effective way to understand your customers.

The research frontier is moving fast. A 2024 study by Park et al. at Stanford simulated 1,000 real individuals as LLM-based personas and found that the synthetic versions accurately reproduced attitudes, beliefs, and behavioral patterns, outperforming baselines built on demographics alone. In 2026, researchers at Google DeepMind introduced “Persona Generators,” functions that use evolutionary AI to optimize how synthetic audiences are created. These generators were tested on topics they'd never encountered before and successfully represented over 80% of the range of human opinion on those topics.

As indicated by these studies, the next generation of synthetic persona tools won’t just generate the most common user; they’ll be capable of creating the full range of user behaviors, including rare and edge-case behaviors that are hardest to anticipate. However, you still need to validate. The best approach, right now, is to quickly build synthetic personas, get directional insight, then pressure-test findings against real users before making important decisions.