Company profile · 3 min read
Foresight uses consumer simulations to narrow CPG research decisions
Foresight simulates consumer behavior for pricing, positioning, packaging, campaign and product-concept decisions before fieldwork.
Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).
Published · Updated
Foresight is a market-research product for CPG and consumer teams that want to test pricing, positioning, packaging or campaigns before spending on a real launch.
What it does
The company describes AI-powered simulations of human behavior. A team can test possible moves, see how simulated consumers are likely to react and choose a path before going into fieldwork. Foresight YC profile Foresight homepage
The public YC launch reports a blind benchmark against a Fortune 500 company with 100+ paired estimates, 95% accuracy against fieldwork and a 0.88 Lin’s concordance correlation coefficient. Those are company-reported results from a launch claim, not an independent replication. YC launch
That distinction is the whole buyer decision. A simulation can help narrow questions and compare directions. It should not automatically replace real customer research when the decision is expensive, novel or hard to reverse.
Why I’d look closer
The product fits a team that needs more iterations before committing to a survey, focus group or launch. The company says it can test pricing, positioning, packaging, campaigns and product concepts against target audiences. That is a useful pre-fieldwork layer if the output helps a researcher decide what to test next.
The founder context supports both sides of the product. The YC profile describes Antoine Bertrand as a former Bloomberg engineer and second-time founder who previously grew a gaming platform to 100,000 monthly active users. Eytan Rozenblum is described as CEO and cofounder of the simulation product.
What could make it the wrong choice
The central risk is false confidence. A simulated consumer is not a real respondent, and a benchmark against fieldwork does not tell a new buyer how the model performs for a different category, audience or price decision.
Pricing was not published in the sources checked. A buyer should ask what consumer data shapes the simulation, how target audiences are defined, how calibration works, what the blind benchmark included, and where the model is expected to abstain.
My editorial take
I would shortlist Foresight for a CPG or consumer-insights team that already knows the question it wants to explore and needs a faster way to narrow options. I would not use it as the only evidence before a major launch. The fit is strongest as a pre-fieldwork filter: run more ideas, kill weaker ones, then validate the finalists with real consumers.
Quick facts
| Field | Sourced detail |
|---|---|
| Product | AI-powered consumer-behavior simulations |
| Buyer | CPG, marketing, advertising and consumer-insights teams |
| Public validation | Company-reported blind benchmark against fieldwork |
| Pricing | Not published in the checked sources |
| Main fit question | Is the simulation narrowing a real research decision or replacing one? |
Sources checked
Foresight YC profile, homepage, YC launch post and linked case-study path were checked on 2026-09-19.
Cohort context
Foresight is listed in Spring 2026. In our 2026-09-18 directory snapshot, 112 of 193 listed companies in that cohort have YC’s primary industry label B2B (58.0%). This is a current-directory comparison, not an original intake count or a performance ranking. Nine-cohort dataset.
Public website snapshot
Observed 2026-09-19T16:16:14.380Z in raw homepage HTML. This records visible metadata and advertised links, not agent execution or product quality.
| Signal | Homepage observation |
|---|---|
| Product description metadata | Observed |
| Canonical link | Not observed in this response |
| H1 or H2 heading | Observed |
| Typed structured data | Observed |
| Docs/developer link | Not observed in this response |
| Pricing link | Not observed in this response |
| llms.txt link | Not observed in this response |
| Markdown alternate | Not observed in this response |
Public observations · Collection method. Missing links here do not establish that a capability or file is absent elsewhere.
About the author
Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).
I cofound Lazyweb and publish Mudpie. This is an owner-written publication, not an independent testing organization. Research notes distinguish observations, sourced reporting and editorial judgment.
First1000 ↗ · X ↗