CPG / Market Research
Built for Our Market Research Platform
Our market research platform had a working AI engine. It could test packaging, positioning, and campaign concepts against AI-generated audiences. But every study still needed a researcher to configure it, interpret the results, and explain the output. The AI was fast, but the workflow around it wasn't.
The Gap
$50K–$500K per study is what traditional market research costs
Even with AI models in place, the product still required manual study configuration, lacked persistent brand context, and couldn't guide non-researchers through a research workflow autonomously. The "last mile" between raw AI capability and usable intelligence was still human-dependent.
What We Built
A Value Layer that encodes 20 years of market research expertise into an agentic AI interface: context, prompting logic, and orchestration that turns a multi-model AI engine into an autonomous research partner.
Brand context that persists
The system ingests brand guidelines, historical data, and competitive positioning so every study speaks the brand's language from day one.
Research methodology as prompting logic
How to deconstruct a research task, which model handles which layer, how to weight and cross-validate results across AI-generated audience segments.
Onboarding agent
Builds a brand's first AI-lookalike audiences and configures their first study in seconds. No researcher required.
Guided workflow orchestration
Walks any team member through positioning tests, packaging validation, concept evaluation, and portfolio analysis without needing to understand the underlying methodology.
The result is Gutsy, a platform where brands run the same caliber of consumer research that used to require a $200K agency engagement, through a conversational AI agent, in seconds.
Layer 3
Applications
Conversational AI research agent (gutsy.so)
Layer 2
The Value Layer
Context
Brand guidelines, competitive positioning, 20 years of research methodology
Prompting
Study configuration logic, audience segmentation, cross-validation rules
Orchestration
Onboarding agent, guided workflow pipelines, study lifecycle management
Layer 1
Systems of Record
Brand databases, AI audience models, study history, campaign data
The Result
| Metric | Before | With Value Layer |
|---|---|---|
| Cost per study | $50K–$500K | ~$300/month subscription |
| Time to insight | 4–12 weeks | Seconds |
| Studies per year | 1–2 (budget-constrained) | Unlimited |
| Researcher required | Senior analyst + agency | Any team member |
| Brand context | Re-briefed every engagement | Persists and compounds |
Real-world validation
A craft brewery used the platform to validate packaging concepts for a new product launch. It became one of the top innovation launches in its category across Ontario retail.
A functional food brand ran Gutsy alongside traditional research ($200K+ study) and paid media testing. All three methods converged on the same winner, proving the Value Layer's accuracy at a fraction of the cost.
A former CEO of Campbell's and Kellogg's (30+ years in CPG) called the platform "unprecedented" and "radically transformational", and said he'd replace his entire marketing research function with it.
The system gets smarter with every study. Each brand interaction feeds back into the Value Layer, compounding institutional intelligence that no single research engagement could build.
What This Proves
- The Value Layer can be the core of an entire product, not just an add-on.
- Domain methodology encoded once → unlimited leverage at near-zero marginal cost.
- Validated against traditional methods. Same accuracy, fraction of the cost and time.
See what a Value Layer looks like for your business.
Book a Free Value Layer Audit