Demand forecasting (Beta)

Demand forecasting

Beta

Demand forecasting takes the messy guesswork out of “how much stock do I need?”. It builds a forecast from your historical sales, factors in your vendor lead times and current stock levels, and tells you what to reorder, in what quantity, by what date.

This is a Beta feature. It’s available to all subscription tiers as part of an active rollout. The frontend is being built in waves — some controls (e.g., velocity-settings overrides) may move or change shape between releases.

In this guide:

  • Open the forecast view
  • Read a forecast for one SKU
  • Override velocity settings
  • Trigger a reorder

Prerequisites

  • At least 30 days of sales history (60+ recommended).
  • Vendor lead times configured for the SKUs you want to forecast. See Lead times.
  • Beta access — visible to everyone; flagged Beta.

Step 1: Open the forecast view

Go to Inventory → Forecasting in the sidebar.

The list shows every active SKU with these columns:

ColumnMeaning
Daily run-rateSmoothed average daily units sold.
Days-of-coverAvailable stock divided by run-rate.
Reorder byDate by which a PO must be raised to avoid stockout, given the lead time.
Suggested qtyThe recommended PO quantity.
ConfidenceHilal’s confidence in the forecast (low / medium / high), driven by data volume and variance.

Sort by Reorder by ascending to see what’s most urgent.

Step 2: Read a forecast for one SKU

Click any row to open the per-SKU forecast detail.

The detail panel shows:

  • History chart — actual daily sales, smoothed.
  • Projection chart — the forecast over the next 90 days.
  • Stock-out date — when, given current stock and run-rate, you’ll run out.
  • Reorder window — the window during which a PO needs to be raised.
  • Component view (for bundles) — per-component forecasts, since the bundle’s bottleneck is the slowest-moving component.

Step 3: Override velocity settings

If you know something the forecast doesn’t (a Prime Day promo coming up, a supplier issue dropping a competitor’s stock), override:

  • Multiplier — bump expected demand by a percentage (e.g., +50% for a Prime Day month).
  • Floor — minimum daily run-rate.
  • Ceiling — maximum daily run-rate.
  • Seasonality profile — pick a preset (none / mild / strong) or define custom monthly weights.

Save the override on the SKU; it persists until you remove it.

Step 4: Trigger a reorder

When a SKU’s forecast says Reorder by is within 7 days, the row gets a Reorder now button. Click it: Hilal pre-fills a PO with the suggested vendor, quantity, and date. You can edit before sending. → Purchase orders

Limitations during Beta

  • Forecasts get noticeably more accurate after 60+ days of history.
  • The seasonality presets are a starting point — for highly seasonal SKUs, plan to tune the custom profile.
  • The frontend is in active development; expect minor UI changes between releases.

Troubleshooting

  • “Confidence: low” everywhere. Either you don’t have enough history yet or your sales pattern is very volatile. Tighten the velocity multiplier manually until enough data accumulates.
  • Reorder date is in the past. You’re already late. Raise the PO immediately — Hilal still suggests a quantity.
  • Forecast disagrees with my gut. Check the history chart for outliers (a one-off bulk sale skewing the run-rate). The detail view lets you exclude specific dates.

Related articles