Predictive Demand Forecasting

Know what your customers will buy before they do.

Predictive demand forecasting for non-subscription commerce. We anticipate each account’s next purchase from the signals you already have.

forecast_output.tab — Weekly Predictions
CustomerProductWeek 23Week 24Week 25Action
C-04821Classic Tee — Black M0.870.120.04Email Now
C-11247Wool Coat — Navy L0.050.230.91Pre-Stock
C-08390Linen Shorts — Sage S0.740.310.08Email Now
C-15602Cashmere Scarf — Cream0.020.680.15Restock

4 of 12,847 predictions · Updated 2 hours ago · Next refresh: Monday 06:00 UTC

The Problem

Your customers don't subscribe.
They come back when they're ready.

Traditional forecasting assumes predictable patterns. Non-subscription commerce doesn't work that way.

Stockouts on your best sellers

Size 8 in black is gone again. Your top variant sells out while odd sizes collect dust.

Overstock eating your margins

You ordered based on gut feel and last year's numbers. Now you're marking down 30% of inventory.

Blind email campaigns

You're blasting the same promotion to everyone because you don't know who's ready to buy.

New launch guesswork

Every product launch is a coin flip. How many units? Which sizes? You won't know until it's too late.

Proven Results

Numbers that speak for themselves

0%
WAPE

Forecast accuracy on Olist e-commerce dataset

0.0%
Baseline WAPE

Traditional forecasting baseline

0M+
Rows Processed

Validated on real e-commerce datasets

How It Works

From data to decisions in three steps

1

Deploy In Your Environment

StratTab runs as a container in your cloud environment with read-only access to your data warehouse. Your data never leaves your infrastructure.

Compatible with Snowflake, BigQuery, Databricks, and Postgres. Setup takes under an hour.

2

We Build Your Forecast Model

Our AI analyzes purchase patterns at the customer × product × week level, generating a sparse synthetic fact table with purchase probabilities.

Hierarchical reconciliation ensures forecasts are consistent from the top line down to individual SKUs.

3

Get Predictions In Your Warehouse

Receive forecasts directly in your data warehouse. Query via SQL, power your email campaigns, and optimize inventory decisions.

Predictions write to your analytics schema. Use any BI tool or downstream system you already have.

Capabilities

Built for non-subscription commerce

Every feature designed for the unique challenge of predicting when customers will come back — without a subscription to tell you.

Warehouse-Native Deployment

Runs in your cloud environment. Your data never leaves your infrastructure. No security review nightmares.

Customer × Product × Week Forecasting

Granular predictions at the individual customer and product level. Not aggregate SKU guesses.

Synthetic Fact Table Output

Predictions in the same format as your transaction tables. Query with SQL. No black boxes.

Hierarchical Reconciliation

Bottom-up forecasts that sum correctly to top-line targets. Consistent across all aggregation levels.

Intermittent Demand Handling

Purpose-built for sporadic, non-subscription purchase patterns that break traditional models.

SQL-Native Integration

Join predictions to your existing dimension tables. Works with any BI tool or downstream system.

Use Cases

Purpose-built for your industry

Fashion & Apparel

Predict size and color demand across seasonal collections. Reduce overstock on slow variants while never missing size 8 in black.

30% reduction in variant-level stockouts

Problems We Solve

  • Size/color variant stockouts
  • Seasonal collection planning
  • New launch quantities
  • Markdown optimization
InventoryMarketingMerchandising

Stop guessing.
Start forecasting.

Join data-driven brands that predict demand instead of reacting to it. See what StratTab can do with your data.

No credit card required · Setup in under an hour · 12+ months of order history recommended