What if you could test a product on thousands of simulated customers before spending money launching it?

AI has dramatically shortened the time it takes to build a product. Understanding how consumers might react can still take weeks. Vectorial is trying to close that gap by creating simulated audiences that product and marketing teams can test before putting something in front of real customers.

The San Francisco-based startup has raised $3 million in seed funding, with backing from Smilegate and SenseAI Ventures alongside Berkeley SkyDeck, Berkeley AI Research Lab, Stanford-linked funds and product leaders from companies including Meta, Shopify and Mastercard. The capital will be used to expand Vectorial’s behavioural datasets and develop the next version of its simulation model, SAPIENS.

Founded by Taranveer Singh, Jasmine Kaur and Rahul Garg, Vectorial combines behavioural science, AI and market research. Its platform lets companies simulate how different customer groups might respond to product prototypes, features, marketing campaigns, user journeys and even AI agents before those experiences are launched publicly.

SAPIENS builds these simulated audiences using socio-demographic, psychographic and behavioural information drawn from multiple data sources. Vectorial says it has modelled more than 2.5 million people globally, including around 500,000 in India, and has already run more than 30,000 simulations. Early customers include boAt, GoodRx, Suki AI, Everkind and Concrete Software.

India is becoming a meaningful market for the company. Vectorial says businesses can run hundreds of simulations each month rather than commissioning a fresh research study for every decision. It claims previous simulations have reproduced findings from live research with accuracy in the 80%-90% range, although those figures are company-reported benchmarks and do not guarantee that simulated consumers will predict every real-world outcome.

Vectorial’s bigger bet is not that synthetic customers will eliminate real research. It is that companies can use simulation much earlier and more often — testing dozens of ideas before deciding which ones are worth putting in front of actual people.