AI-driven demand planning for e-commerce

Know what sells.
Before it sells.

astock.ai forecasts demand, senses price elasticity and plans your inventory across every SKU, warehouse and channel — so you stop losing revenue to stockouts and dead stock.

30-minute walkthrough on your own numbers · Connects to Amazon, Shopify, eBay & Walmart

app.astock.ai / demand-forecasting
Home
Demand Analysis
Demand Forecasting
Price Sensing
Inventory Planning
Product Research
Ask AI
Revenue · 365d
£4.2M
▲ 18.3% vs prior year
Units sold
312k
▲ 11.9%
Stockout risk
7 SKUs
▼ from 19
Pricing upside
£216k
next 90 days
WAN-W1B+ — Pro Boxing Gloves 16oz
History, forecast & 80% confidence band
— History — Forecast ▒ Confidence band
0
SKU-warehouse combos planned
0
avg. revenue lift in 6 months
0
fewer stockout days
0
daily forecast horizon
RDX SPORTSNORDIC GEAR CO.PEAKFITURBANCYCLEHOMEBARNVELOX SUPPLY
Everything in one workspace

Six planning superpowers.
Zero spreadsheets.

Purpose-built for multi-channel sellers: every module reads the same live data, so your forecast, your prices and your purchase orders finally agree with each other.

Demand Forecasting

180-day daily forecasts per SKU and warehouse with uncertainty bands, trend and seasonality decomposition you can actually explain to your team.

Price Sensing

Elasticity estimated from your own sales history — see exactly how demand reacts to each price move before you make it.

Inventory Planning

Stockout and overstock exposure in revenue terms, recommended order quantities, and a what-if price slider that re-simulates your stock curve live.

Product Research

Review analytics and semantic search across your marketplace reviews — find the sizing complaint before it becomes a return rate.

Ask AI

Chat with an analyst that knows your data. "What should I reorder this month?" gets a table, a chart and a straight answer.

Channel Sync

Amazon, Shopify, eBay and Walmart feeds unified into one clean product hierarchy — warehouses, regions and channels stay comparable.

How it works

From raw sales feed to confident decisions

Forecasts you can interrogate, not just trust

Every forecast ships with its confidence band and a full decomposition — trend, yearly and weekly seasonality, out-of-stock corrections — so "why is the model saying that?" always has an answer.

  • Daily granularity per SKU × warehouse, 180 days ahead
  • Stockout-aware training: lost sales don't poison your forecast
  • Additive or multiplicative seasonality, one click apart
Model components — yearly seasonality
above yearly average below average

Price with evidence, not gut feel

astock.ai fits a demand curve to your own price history, then finds the price that maximises revenue given the stock you actually hold. Drag the slider, watch the stock curve and 90-day revenue re-simulate instantly.

  • Per-SKU elasticity with fit quality (R²) shown honestly
  • Scarce stock → hold price. Overstock → the exact markdown to clear it
  • Revenue response curve with today's and recommended price marked
Revenue response to price — HGR-F6M
£24.50optimal £38.20 · +11.4% revenue£45.50

See the cost of bad stock — in pounds, not units

Every understock and overstock is translated into revenue at risk. Replenishment quantities target your optimal cover, and in-transit shipments are already on the curve.

  • Stockout revenue loss ranked across the whole catalogue
  • Order exactly to optimal cover — 120 days by default, yours to tune
  • Slow-mover markdowns computed from elasticity, not guesswork
Top stockout items by revenue loss
Loved by operators

Planners sleep better on astock.ai

★★★★★

"We cut stockout days by 40% in one quarter. The what-if price slider paid for the subscription the first week — we cleared £60k of slow movers without torching margin."

AS
Ahtesham Sadiq
Demand Planning Lead, RDX Sports
★★★★★

"The forecast decomposition is the killer feature. When finance asks why we're ordering heavy for January, I show them the seasonality chart and the meeting ends early."

MK
Maja Kowalski
COO, Nordic Gear Co.
★★★★★

"Ask AI replaced a weekly hour of pivot tables. 'What should I reorder this month' gives me a ranked list with quantities I can paste straight into a PO."

DR
Dev Raghavan
Founder, PeakFit
Pricing

Starts free. Scales with your catalogue.

Every plan includes all six modules. You're only ever paying for scale, not features held hostage.

Starter

£0
forever · up to 100 SKUs
  • All six modules, full UI
  • 1 warehouse, 2 channels
  • 90-day forecast horizon
  • Weekly data refresh
Book a demo
MOST POPULAR

Growth

£249
per month · up to 5,000 SKUs
  • Unlimited warehouses & channels
  • 180-day horizon, daily refresh
  • Price optimizer & what-if simulation
  • Ask AI analyst, unlimited questions
  • CSV / API export
Book a demo

Enterprise

Custom
unlimited SKUs · SLA & SSO
  • Dedicated data pipelines & onboarding
  • Custom optimal-cover policies
  • Private cloud or on-prem deploy
  • Named success engineer
Talk to sales
FAQ

Questions, answered

How does astock.ai get my sales data? +
Connect Amazon, Shopify, eBay or Walmart with read-only API keys, or upload CSV exports. We normalise everything into one product hierarchy — SKU, warehouse, region, channel — and refresh on your plan's schedule.
How accurate are the forecasts? +
Accuracy depends on your history and demand volatility, which is exactly why every forecast ships with an honest confidence band instead of a single number. Most catalogues see weekly-level errors of 10–20% after two seasonal cycles of data — and the inventory math is built to plan against the band, not the midpoint.
Do I need a data team to use it? +
No. The app is built for operators: every chart answers a planning question in plain language, and Ask AI turns questions like "what do I reorder this month?" into ranked tables. If you *do* have a data team, everything exports via CSV and API.
What happens to my forecasts during stockouts? +
Out-of-stock days are detected and excluded from model training, so a stockout reads as lost availability rather than vanished demand. Those days are also flagged on the forecasting charts so you can see exactly where history is being corrected.
Can I try it without connecting my store? +
Yes — book a demo and we'll walk you through every module on a realistic sample catalogue (12 SKUs, 4 warehouses, 3 years of history), or on your own CSV export if you bring one. You connect your store only when you're ready.
Get started

Your next stockout is already
on the forecast.

See it coming — and the price, order quantity and timing that prevent it.