What is Foundry Simulate?

Foundry Simulate spins up thousands of AI agent personas based on your exact target audience. Each agent has real psychology, backstory, and opinions. You test a campaign or message and get back a full intelligence report — what landed, what backfired, why, and who said what.

How much does Foundry Simulate cost?

Foundry Simulate starts at £500 for a Starter simulation (50 agents, results within 36 hours). Standard is £1,500 (200 agents), Deep is £3,500 (500 agents), Enterprise from £6,000. Compare to traditional focus groups at £5,000-£15,000.

How is Foundry Simulate different from traditional market research?

Traditional research takes weeks and costs thousands. Foundry Simulate delivers within 36-48 hours — days faster than traditional research, with expert interpretation included. The agents talk to each other — opinions form and evolve like real social environments. You can interrogate individual agents and ask why they responded the way they did.

What can I test with Foundry Simulate?

Marketing campaigns, ad creative, product messaging, pitch decks, pricing pages, feature announcements, and social content. Any message you plan to send to an audience can be tested in advance.

Foundry Simulate

Know how they'll respond.
Before you ask them.

Spin up thousands of AI agent personas — your exact target audience — and test your campaigns, messages, and ideas before you spend a penny finding out the hard way.

Book a Simulation Request Sample Report

Watch the intro

See it in 90 seconds

Every marketing decision starts with a guess. Foundry Simulate is how you stop guessing.

Every marketing decision starts with a guess. Which hook will perform? Will this positioning land? Will the feature announcement excite or confuse? Will this campaign justify the budget? You've got three options. Run it live, expensive, slow, and public. Commission research, more expense, even slower. Ask your team and get confirmation bias dressed up as feedback. None of these work at the speed modern marketing demands. So we build something different. It's called Foundry Simulate. You give it a campaign, a pitch deck, a product launch, and it spins up thousands of AI agent personas based on your exact target audience. Any persona, any language, any market, not survey respondents, agents. With real psychology, real backstories, real opinions, they talk to each other, consensus builds, you jump into any one of them and ask, why did you respond that way? And you can listen to every single conversation. The market leader charges twenty thousand for this. The number two charges eight. We're starting at seven fifty. We've already scoped our first three clients. Better briefs, fewer wasted budgets, faster decisions. Before you spend a penny finding out the hard way, this is Foundryworks, and we're just getting started. Foundryworks dot ai.

Real simulation output

This is what you get.
Not a report. A room.

Each node is an agent — a fully-formed persona with their own psychology, background, and opinions. The lines show how opinions spread between them. This is a live social simulation, not a survey.

Opinions form. Consensus builds. Polarisation happens in real time. You can see exactly who changed their mind, who influenced them, and what tipped the conversation.

Agents talk to each other — no competitor at this price point does this
Jump into any individual and ask: why did you respond that way?
See what landed, what backfired, which segment you're losing
Output: a ready-to-brief intelligence report your creative team can act on

Foundry Simulate agent network graph — 500 AI personas and their social interactions during a live simulation

Simulation network — 500 AI agents, 72 rounds, real social dynamics

Process

How it works

Three steps. Brief us on your audience and what you want to test. We run the simulation. You get an intelligence report — not a data dump, a brief your team can act on.

Brief us

Tell us your target audience, what you're testing, and what decisions you need to make. We build the persona library — real psychology, real backstories, any demographic, any market.

We run the simulation

Your content goes into the room. Agents react, discuss, argue, and form opinions — talking to each other the way real people do in social environments. 50 to 1,000+ agents, run and interpreted by our team.

You get intelligence

Not raw data. A proper intelligence report: what resonated, what backfired, which segments responded differently, verbatim quotes, and strategic recommendations. Then ask us anything.

What's included

Every simulation delivers

📊

Audience segmentation

How different personas within your target audience responded differently — and why.

💬

Verbatim agent quotes

What your AI audience actually said. Real language, real objections, real enthusiasm.

🎯

What landed / what backfired

Ranked message performance with explanation. Know your best angle before you spend.

🔍

Individual interrogation

Ask any agent directly: why did you respond that way? What would have changed your mind?

📈

Network graph

Visual map of how opinions spread through your audience. See consensus build in real time.

📋

Strategic recommendations

A ready-to-brief intelligence report. Not data — direction. Exactly what your team needs to act.

Pricing

Straightforward pricing. Extraordinary margins.

Traditional research charges £15,000 and takes four weeks. A human focus group with 50 people costs £5,000-£15,000 for one round of questions. Foundry Simulate starts at £500 and delivers a full intelligence report within 36-48 hours — including expert interpretation, not just raw data. Re-run it with a different angle for the same price.

Starter

£500 one-off

50 agents · 20 rounds

Intelligence brief
Audience segmentation
What landed / what backfired
Verbatim agent quotes
Email delivery

📋 Results within 36 hours

Foundry Simulate vs the alternatives

Method	Cost	Speed	Agents talk to each other	Interrogate individuals	Re-runnable same day
Human focus group	£5,000–£15,000	2–4 weeks	Partial	No	No
Kantar / traditional research	£15,000–£40,000	4–8 weeks	No	No	No
System1	£3,000–£8,000	1–2 weeks	No	No	No
Generic AI survey tools	£30–£100/mo	Minutes	No	No	Yes
Foundry Simulate	from £500	15–45 min	Yes ✓	Yes ✓	Yes ✓

Use cases

Who uses Foundry Simulate

Marketing agencies

Pre-flight creative testing before client campaigns go live. Show clients which hook wins before committing to production. Eliminate your weakest concepts for £500 before spending £3,000 on System1.

"Before you spend £10k testing on real audiences, spend £500 testing on synthetic ones."

In-house marketing teams

Justify campaign spend to the board with pre-flight data. Show evidence that your creative wins with segment X before you launch. Reduce the risk of campaigns that don't convert.

"Here's pre-flight data showing this creative outperforms the alternative with our target segment."

Product teams

Test feature messaging, pricing page copy, and onboarding language before launch. Which product description converts? How do different segments respond to your positioning?

"We tested three positioning angles on 200 agents before writing a single line of copy."

Founders and startups

Test your pitch before investor meetings. Test your landing page message before paying for traffic. Know which segment responds best before you double down on one audience.

"£500 to test my Series A pitch narrative before the meeting. Obvious decision in hindsight."

Get started

Stop guessing.
Start simulating.

Your next campaign, product launch, or pitch deserves to be tested before it's public. A Starter simulation is £500. You'll have a full intelligence report in your inbox within 36 hours — run by our team, interpreted by humans, ready to brief straight to creative.