In a world where a viral TikTok video can cause a product to go global in a matter of hours, the traditional market research cycle – which often takes 12 weeks – becomes a liability.
The lag between the survey question and the answers from the majority (or target) of respondents has become a major obstacle for Fortune 500 decision makers who are forced to navigate the changing political and economic landscape with data that is often out of date by the time it reaches the slide deck, as industry experts have observed.
Brox, a predictive human intelligence startup, recently announced a strategic funding round following a year in which they reported 10X revenue growth. Their proposal is as ambitious as it is technical: the creation of a “same universe” filled with digital twins of 60,000 real, living people with all their demographic profiles and consumer preferences, allowing businesses to run unlimited tests in hours rather than months.
“These digital twins are replicas of real, real people,” said Brox CEO Hamish Brocklebank in a recent video call interview with VentureBeat. “We recruit real people like a traditional panel company does, pay them to interview them, and capture all the information around them – fully consent-driven.”
The company, which currently employs 14 people, positions itself against the “crazy” research industry. By replacing statistical models with behavioral models, Brox aims to change the way the world’s biggest banks and pharmaceutical giants anticipate people’s reactions to global and dynamic events, or target product releases and employee issues, and everything in between.
The survey types and specific questions that Brox asks its digital twins are completely open and can be customized to fit any use cases the business customer has in mind.
According to Brocklebank, examples of research questions include: “What happens if America attacks Iran or Greenland? Will depositors at Bank of America put more money into their accounts or take more money out? Or, in medicine, if RFK Jr. says something next week, will that make people more likely to take vaccines or less likely to take them?”
Not artificial people – copies of real AI
The main difference of Brox technology is in the reliability of its input data.
While many competitors in the “digital audience” space rely only on artificial intelligence – ordinary people produced by Large Language Models (LLMs) – Brocklebank says that these methods are inevitably producing “AI slop”.
Artificial audiences tend to converge on a rigid distribution of responses, over-identifying “correct” or “healthy” behaviors (such as eating broccoli) due to inherent biases in the underlying models.
Brox’s “Digital Twins” are instead one-on-one behavioral simulations of real people recruited and interviewed in full depth. The process is intensive:
-
Deep Conversations: The company conducts hours of real-time and AI-driven conversations with each participant.
-
Depth of Mind: The dataset seeks to understand “decision drivers,” including upbringing, relationships, and even marital stability.
-
Data Density: For some twins, Brox stores up to 300 pages of textual data, representing what Brocklebank calls “the deepest individual data set in existence”.
To solve the “black box” problem common in AI, Brox uses “chain of reasoning” in its predictive results. When a digital twin predicts a reaction — such as how a $2 billion net worth individual might respond to a certain interest rate increase — the model imports representations and provides a step-by-step explanation of that decision.
This allows clients to understand and not just what will happen, but the basic psychology of why it happens.
Scoring the “immeasurable” interview.
Product offerings are currently live in the US, UK, Japan, and Turkey. Brox has successfully installed certain collections, high values that are difficult for researchers to reach.
This includes a panel of “high net worth” people (those worth more than $5 million) and specialized medical professionals like dermatologists – including billionaires.
However, the largest number of customers may be in the totality of all people who can be cast in large numbers and/or divided into demographics, especially those of middle and low income levels, whose purchasing power and decision-making are very delayed and who do not market-
One of the most unique features of the Brox platform is its incentive structure. To ensure that the twins stay up-to-date, their real-world counterparts are also contacted regularly.
For high-net-worth individuals not motivated by small cash payments, Brox issued Stock Appreciation Rights (SARs), making these participants “investors” in the company’s success to ensure they continue to provide high-fidelity personal reviews. The platform’s use cases currently focus on two main areas:
-
Medicines: Predicting vaccine reluctance or how doctors might react to new biologics based on a changing climate.
-
Finance: Simulating how depositors in central banks can move money in response to world events, such as conflicts in the Middle East.
As for why he goes to the trouble of interviewing and digitally integrating real people instead of just creating completely fictional, artificial characters for audiences and people using LLMs and other AI models, Brocklebank offered his opinion.
“You could create 10,000 virtual digital twins, but the answers would still be normalized into a robust distribution, which is not true if you ask real people,” Brocklebank said.
By maintaining a pre-built audience of 60,000 twins, the company enables clients to get past the research stage. A major US bank or a global pharma giant can now “interrogate” a digital population and receive a verified analysis within hours.
Price and accessibility
Unlike traditional survey companies that charge on a per-project or per-respondent basis, Brox operates as a state-of-the-art Software-as-a-Service (SaaS) platform with an enterprise-grade commercial license. The company avoids the “seat” or “use” limitations that often prevent rapid testing within large organizations.
-
Price Categories: Subscriptions are sold as a low cost, starting at a minimum $100,000 a year.
-
High Level Contracts: For larger applications involving multiple parties and global data access, contracts scale up $1.5 million a year.
-
Use Rights: Clients are provided unlimited use during the contract period. This allows them to run thousands of simulations without worrying about rising costs, promoting a culture of “testing everything” before shipping.
From a legal and privacy perspective, digital twins are built on a “fully consent-driven” framework. While twins can be traced back to real human data for internal validation, the platform is designed to provide aggregated behavioral data that protects participants’ anonymity while maintaining the ability to predict their digital counterparts.
Reject the rise of Kalshi, Polymarket and ‘prediction markets’
The tech industry has recently seen an increase in valuation and interest in “prediction markets” such as PolyMarket and Kalshi, which allow users to bet on the outcome of various world events.
However, the leadership at Brox maintains a distinct distance from these platforms, citing “personal disdain” for betting markets from an ethical and intellectual perspective.
Brocklebank says that although the betting markets are predictable results (eg, who wins an election), they provide zero help to business decision makers because they fail to provide the “why”.
Knowing that there is a 60% chance that someone will win does not help the company to adjust its customer strategy; to know why A certain group of investors is feeling worried.
Investors including Scribble Ventures, Wonder Ventures, and Vela Partners have backed this “human first” approach to AI, betting that the moat created by deep human data will prove stronger than commercial models of artificial data providers.
As Brox prepares to launch in the Middle East and APAC, the company is moving toward its ultimate goal: to emulate the rest of the world as a “same world” for risk-free decision-making.
