Before it hits the shelf,
see who wins on a virtual one

Market Simulation Agent — using real shelves and competitors, it simulates shopper choices so you know before launch which display and which new product gets picked

Stop trial-and-error in stores — run it in the model first

Turns shelf display and competitive dynamics into quantifiable, comparable, reproducible choice simulation

Traditional display / launch testing
Market Simulation Agent
Display decisions
Gut feel and repeated in-store trials
Conjoint analysis simulates shopper choice, quantifying each display's win rate
Testing cost
Physical stocking, long cycle, high cost
One-click virtual-shelf simulation, results in minutes
Competitive dynamics
Competitive pressure hard to quantify
Competitor placement and price built into the model, quantifying share shifts
New-product forecast
Only learn after launch whether it sells
Simulate a new product's choice share on the shelf before launch
Audience differences
Only see the aggregate, not the segments
Scores by shopper segment, showing who's drawn to what
Communicating results
Piles of table numbers the business can't read
A visual shelf + a phone-tap short link anyone can understand

More than fast — you commit when the decision is fully computed

From shelf modeling to audience choice, every step rests on a choice model, quantifiable and comparable

Display all by experience
Models the shelf, price and competitors, computing a win rate for every layout
Don't know which one shoppers pick
Based on a choice model that simulates real shopping, revealing where share flows
Uncertain about a new-product launch
Before launch, simulate the new product's shelf performance — validate, then stock
Results don't reach the front line
Generates a phone-friendly visual short link the business can check anytime

You're always the decision-maker

1

Describe the scene · or upload a shelf photo

Describe in plain words the display options to compare, or upload a real shelf photo and product list

2

You review · you decide

Once the shelf, competitors, prices and constraints are set, it pauses for your confirmation; you own the parameters

3

Take comparable conclusions

Choice share and win/loss across display options at a glance, with a visual shelf and phone short link

Follow a real simulation through the full workflow

From a real shelf to visual results, see how the agent computes the display decision

→ →
Real shelf Click to enlarge
STEP 1 · Research scene · a real shelf photo as the simulation starting point

Start from a real shelf

Using the store's actual display as the baseline, it reconstructs products, facings and prices as the simulation's starting point.

Smart recognition Click to enlarge
STEP 2 · Smart recognition · structuring shelf, competitors and prices while you chat

Upload a photo and the shelf is read automatically

Against the on-site photo, the agent recognizes shelf rows and columns, competitors and prices, organizing them one by one into a structured list with confidence scores for you to review and confirm.

Optimal display Click to enlarge
STEP 3 · Optimal display · optimal position, predicted choice rate and drivers

Compute which position gets bought most

Outputs the optimal placement and predicted choice rate, marking the plan on a visual shelf and explaining the impact of shelf level, position and price.

The above are real project outputs, anonymized · swipe or click a tab to switch, click the image to enlarge

Your market simulation team

From modeling to charting, a full set of display and new-product simulation capabilities

Shelf modeling
Reconstructs shelf dimensions, facings and product positions close to the real display
Details

Natural-language shelf configuration · product-position and facing reconstruction

Competitors & pricing
Competitor placement, price, locks and no-place zones all built into the model
Details

Competitor placement / price / lock parameters / no-place cells

Shopper choice simulation
Uses conjoint analysis to simulate purchase decisions across segments
Details

Per-segment scoring · choice-model-based share estimation

Optimal placement search
Searches for better display combinations under constraints
Details

Multi-option comparison · constrained placement optimization

New-product win-rate forecast
Simulates a new product's choice share on the shelf before launch
Details

New-product scenario simulation · share and sensitivity

Visual short link
Results produce a visual shelf and a phone-friendly short link
Details

Visual-shelf output · mobile short link, one tap away

Market Simulation Agent — Market Simulation Agent  •  Powered by MatrixGorilla