ML-powered · New-product ready · No implementation

Know exactly
what to order,

Demandatory gives retailers the most accurate demand forecasts in retail — 2.5× more accurate than legacy vendors on new products. No mandatory six-month rollout. Start plug-and-play, or integrate deeply — your call.

Try the demo

No signup required

WMAPE 42
vs competitors' 105
60 SEC
upload to first forecast
52 WK
forecast horizon

Categories we forecast

Built for every shelf.

Products we forecast for, across our customers' warehouses.

Live forecast
ALL-SKUS
Network aggregate
All clusters · Live
1B+
SKUs forecasted
$2.4B
inventory value
2.5X

More accurate on new products

WMAPE 42 vs competitor's 105

60S

Upload to forecast

No implementation project required

52 WK

Forecast horizon

Weekly predictions with confidence intervals
Capabilities

Demand forecasting powered by AI

Accurate forecasts without the enterprise price tag or timeline.

Cold Start

Launch-day accuracy,
zero history.

Launch-day forecast · 3 new SKUs
DAY 1 · LAUNCH
  • Oat Milk 1L
    $4.29
    1,842/wk
  • Protein Bar
    $2.49
    864/wk
  • Shampoo 250ml
    $7.90
    412/wk
Explainable

Every number is defensible.

Forecast breakdown
Vanilla Ice Cream 1L
Total predicted demand
next week
640 units
Baseline 450 units
Uplift + 190 units
  • Baseline demand 450 units
    Local trend
  • Weather penalty − 180 units
    Spring freeze
  • Easter holiday + 120 units
    Calendar
  • 20% discount lift + 210 units
    Promo prior
  • Weekend mix + 32 units
    Day of week
  • Lagged demand + 58 units
    Last 7 days
  • Competitor promo − 28 units
    Price gap
  • Stock-on-hand + 24 units
    Below threshold
  • OOS neighbour − 18 units
    Substitute bias
  • New-customer lift + 22 units
    Segment mix
  • End-cap placement + 16 units
    Display type
  • Private-label gap − 14 units
    Price ladder
  • Regional trend + 20 units
    West coast uptick
  • Day-of-month + 12 units
    Payday cycle
  • Markdown recency − 10 units
    Last 14 days
  • Basket affinity + 14 units
    Co-buy w/ cones
  • Cluster average + 11 units
    Stable repeaters
  • Competitor overlap − 8 units
    Promo clash
  • Shelf facings + 18 units
    4 → 6 facings
  • Category growth + 15 units
    YoY +8%
  • Fresh inventory age − 6 units
    Days since ship
  • Loyalty tier mix + 12 units
    High-tier %
  • Local event calendar + 14 units
    School holiday
  • Brand affinity + 9 units
    Household repeat
  • Delivery frequency − 7 units
    Gap in restock
  • Shelf height + 6 units
    Eye-level shift
Instant Setup

CSV in. Forecast out.
47 seconds.

demandatory.com/forecast/sales_2026_q1
SKUs parsed
4,218
Stores
312
WMAPE
21
00:47
SECONDS
TO FIRST FORECAST
OR Connect live — forecasts refresh as your data does
BigQuery Snowflake Postgres Databricks Redshift + more
Your Cloud, Your Data

Runs where you run.
Data never leaves.

Managed SaaS
Our cloud, SOC 2
Fastest
Private VPC
Your AWS / GCP / Azure
Most chosen
On-Premise
Your datacenter
Air-gapped
✓ Your data never leaves your environment.
Retail-Native

Every signal that moves demand.

% Promotions ★ Holidays $ Price elasticity ⇌ Cannibalisation ☁ Weather ◐ Seasonality ✦ New launches ◈ Store clusters ◇ Lifecycle ↕ Competitor price
Week of Dec 18
PROMO HOLIDAY WEATHER LAUNCH
Scale

One store, ten thousand,
every warehouse.

1
store
→
50
stores
→
10k+
network
Warehouse
Store A
Store B
Store C
Forecast warehouse demand from every store.
SAME MODEL SAME ACCURACY SAME API
Triple-layer architecture

Smart from day one.
Sharper every week.

Pre-trained on retail physics before you sign up. Fine-tuned to your SKUs the moment data starts flowing.

Prior layer · physics
WMAPE
78 ↓
Wide supermarket aisle with diverse products
Step 01
Pre-trained before you sign up

Universal retail physics

+28%
Pre-holiday uplift
−40%
Promo decay by 5th run
−1.8
Price elasticity

Laws the model knows before it sees a row of your data.

Produce display with grouped categories
Step 02
Four behaviours, every retailer

Products cluster by behaviour

38%
Stable
27%
Lifecycle
18%
Intermittent
17%
Promo

Four archetypes explain 92% of SKUs — grocery, apparel, DIY, pharma.

Close-up of bananas on a shop shelf
Step 03
Refines every week you use it

Fine-tuned to your shelf

42
WMAPE (legacy: 105)
± 4%
Forecast precision
100%
SKU coverage, day 1

Priors refine to your SKUs, stores, and customers — sharper every week.

Legacy vendors
105
WMAPE on new products
vs
2.5x
sharper
Demandatory
42
WMAPE on new products

See it on your own shelves.

Drop your email — we'll set up a private demo on your sales data within 24 hours.

Or try the sample demo · full contact form