Documentation

Forthcast is an AI-powered inventory forecasting app for Shopify. It connects to your store, analyzes your sales history, and helps you make smarter decisions about when to reorder, how much safety stock to keep, and which products need attention.

Core capabilities include:

• Demand forecasting at the product variant level
• Automatic reorder point and safety stock calculation
• Purchase order creation and lifecycle tracking (Draft → Ordered → Received, with Cancelled)
• Stock Health report with lost-sales and slow-moving stock breakdown
• Forecast accuracy and bias scoring per product
• Bundles, multi-packs, variety packs, and packaging-material forecasting
• Backorder tracking and PO Late / PO Low warnings
• Automatic subscription & wholesale order detection
• Supplier reliability grading
• Multi-location inventory support
• AI Chat for natural language inventory questions

New to Forthcast? Follow the Getting Started guide to get set up in under 10 minutes.

Forthcast analyzes your historical orders to calculate a demand forecast for each product variant. The forecast is expressed as a projected daily sales rate, which is then used to estimate how many units you'll sell over the next 30, 60, or 90 days.

How forecasts are calculated

Forthcast uses a time-series model trained on your store's order history. It detects trends (is demand growing or shrinking?) and seasonality (does demand spike at the same time each year?). Forecasts are recalculated automatically after each new sync with Shopify.

Data requirements

Forecasts can be produced from your very first orders, but get more precise as history accumulates. SKUs with fewer than 6 months of history are tagged Limited Data; SKUs with fewer than 2 weeks of history are tagged Low Data and the forecast is an early estimate only. Treat the recommendation as a starting point on those SKUs and lean on your own knowledge until more history builds up.

Anomalies & spike detection

Unusually large orders are flagged as Spike anomalies and excluded from the seasonal pattern by default, so a single B2B or bulk order doesn't inflate your daily reorder point. You can review each spike on the SKU detail panel and choose to include or exclude it.

Viewing forecasts

Navigate to Forecasts in the main menu. Use the time horizon selector to switch between 30, 60, and 90-day views. Click any product to open its SKU detail panel with the full forecast chart, anomalies, and history.

Found under Inventory → Replenishment. The day-to-day report for managing reordering decisions across your entire catalogue.

Reorder point

The inventory level at which you should place a new order to avoid a stockout before the next shipment arrives. Forthcast calculates it as: average daily demand × supplier lead time + safety stock. Lead time has the biggest influence on this number, so make sure it's set correctly in Settings.

Ordering on a fixed cadence instead? Switch the supplier (or the whole shop) to Scheduled Replenishment and Forthcast replaces this column with a Cycle Reorder Pt sized for the full upcoming cycle.

Safety stock

A buffer above expected demand that protects against demand spikes and supplier delays. Forthcast calculates it using demand variability and lead-time variability at your chosen service level (90%, 95%, or 99%). A higher service level means fewer stockouts but more inventory on hand.

Order quantity, days of stock & revenue to protect

• Order Quantity — the suggested quantity to bring stock back to a safe level. Override it freely based on MOQs, promotions, or seasonality.
• Days of Stock — how long current stock will last at the current sales rate. The coloured bar shows coverage relative to your lead time; if it's short or red, stock may run out before a new order can arrive.
• Revenue to Protect — the sales revenue at risk if you stock out during the lead-time window. Use it to prioritise which reorders to act on first when cash or supplier capacity is limited.

In Transit, PO Late & PO Low

Units on open purchase orders show as In Transit and are factored into your stock position so you don't double-order. A PO Late badge appears when an open PO has passed its expected arrival date; PO Low appears when the incoming quantity is below what was originally planned. Both are prompts to follow up with your supplier.

ABC classification

Every SKU is automatically classified A, B, or C based on its contribution to revenue. A items are your top revenue drivers and deserve the closest attention; B are mid-tier; C are slow movers or new products with limited history. Classifications recalculate as your sales patterns change.

Bulk settings

Apply reorder point and safety stock settings to multiple products at once via the bulk edit tool. Filter by supplier, collection, or product tag before bulk editing to target the right products.

Scheduled Replenishment (Fixed Cycle)

If you order on a regular cadence — a quarterly container, a monthly review with a domestic supplier, or anything driven by MOQs — switch from On Demand to Scheduled (Fixed Cycle) mode under Settings → Replenishment model. Forthcast will then size every PO around the upcoming cycle instead of the next reorder threshold.

When to use Scheduled mode

Pick Scheduled mode when one or more of these are true: you receive containers on a fixed shipping cadence; your supplier has a high MOQ that forces you to batch orders; you have a recurring supplier review (e.g. every Tuesday or the 1st of each month); or you're paid-on-order and want to lock spend to a predictable schedule. Stay on On Demand if you can place a PO any day of the week.

Setting a default cycle & next order date

From the Replenishment model panel, set a default order cycle in weeks (4 = monthly, 8 = bi-monthly, 13 = quarterly) and a default next order date. These apply to every supplier that doesn't have its own override.

Per-supplier overrides

On the Suppliers settings page each row has its own Order cycle (weeks) and Next order date. Supplier-level values always win over the shop default; if you leave a supplier blank, it falls back to the shop default. If both are blank, the cycle defaults to 4 weeks.

Cycle Reorder Pt — what it actually means

In Scheduled mode the Reorder Pt column on the Replenishment table is replaced by Cycle Reorder Pt. The math is plain English: cycle weeks of forecasted demand + (optional) lead-time demand + your existing safety buffer. When available stock drops below the Cycle Reorder Pt, the row turns into Order Now.

The "include lead time in cycle horizon" toggle controls whether lead time is added on top. Turn it on for long lead times or transit-heavy supply chains where stock you order today won't arrive until well into the next cycle; leave it off for short, predictable lead times where the cycle alone covers you.

Manual vs. auto Next Order Date

Each next order date carries one of two sources: manual (you set it) or auto (Forthcast manages it). Manual dates are never overwritten — when a PO ships, Forthcast prompts you to review the date so you can roll it forward yourself. Auto dates advance by exactly one cycle each time a PO is created or received with that supplier, anchored to the PO's date and clamped to today or later. You can flip between sources at any time.

Worked example — a quarterly container

Say you import a category from one overseas supplier, cycle = 13 weeks (one container per quarter), lead time = 8 weeks, "include lead time" turned on. On the next order date, Forthcast looks at every SKU on that supplier and computes Cycle Reorder Pt = 13 weeks of demand + 8 weeks of demand + safety buffer. Anything below that line shows up as Order Now with a suggested quantity that brings stock back above the Cycle Reorder Pt for the upcoming 13-week window. You build the PO, send it, and Forthcast advances the supplier's next order date by 13 weeks automatically. Receive the container 8 weeks later and stock is replenished to cover the next quarter.

Switching modes mid-month

You can switch a supplier (or the whole shop) between On Demand and Scheduled at any time. No data is lost — the Reorder Pt and Cycle Reorder Pt columns simply swap places, and you keep all of your demand history, safety stock settings, and supplier lead times.

Every row in the Replenishment report carries a status badge so you can scan urgency at a glance:

• Order Now (red) — stock is already at or below the reorder point. Act today.
• Plan Ahead (orange) — stock will reach the reorder point within your lead-time window. Order soon.
• Pre-season — a seasonal demand peak is approaching within your lead time. Consider ordering ahead even if stock looks sufficient now.
• Limited Data — fewer than 6 months of sales history; recommendations are less precise.
• Low Data — fewer than 2 weeks of sales history; forecast is an early estimate only.
• Spike — an unusually large order was detected and excluded from the seasonal pattern by default.

Click any SKU name in the Replenishment table to open a full breakdown for that product.

Stock breakdown — the four tiers

The Inventory section splits stock into four tiers so you can see exactly where every unit is:

• Available — physically on hand and free to sell. Same number Shopify shows on the storefront.
• Committed — units already attached to unfulfilled customer orders. Sold from a forecasting standpoint, but still in your warehouse.
• On PO — units on open purchase orders not yet received. Hover for expected-arrival date.
• Backordered — units customers have ordered that you couldn't fulfil. Demand you owe and need to clear with your next inbound shipment.

Reorder calculations use all four. Your effective stock position is Available + On PO − Committed − Backordered. Looking at on-hand alone is misleading once you have unfulfilled orders or POs in flight.

Reorder Intelligence chart

Plots how demand is expected to move alongside your current stock trajectory, with a shaded uncertainty band and a horizontal reorder-point line. Where the stock line approaches zero is where you'd run out if no order is placed.

• Average cycle chip (e.g. ~14mo) — the typical gap between orders for slow-moving SKUs, so you can right-size safety stock for infrequent sellers.
• Amber ~X/yr chip — shown when a SKU sells fewer than roughly one unit per month annualised, signalling a low-frequency item where the standard daily-rate model is less precise.

Retail vs. wholesale split

The Order Breakdown section labels each historical order as retail, bulk, or wholesale and shows the percentage of recent demand from each. Bulk and wholesale orders are detected automatically (see Order Tag Detection) and excluded from the seasonal pattern by default, so a single large B2B order doesn't inflate your daily reorder point. You can override the classification on any individual order.

Demand history & anomalies

Actual sales (bars) are plotted alongside the forecast line so you can see how well the model has been tracking. Anomalies — orders significantly larger than typical — are flagged and excluded from seasonality by default; review each one and choose whether to include it.

Found under Inventory → Performance. Shows how recent forecasts compared to actual sales and lets you apply manual demand adjustments.

• Velocity — average units sold per month across months that had sales. A quick read on how fast a SKU moves.
• Demand Trend — your most recent 3 months vs. the 3 months before. Tells you whether demand is rising or slowing.
• Accuracy — how close the forecast was to what actually sold. Above 70% is healthy; lower accuracy on a high-velocity SKU is worth investigating, lower accuracy on a slow-moving SKU is normal.
• Bias — direction of forecast errors. Positive bias = you're consistently selling more than predicted; negative = forecast is over-estimating. Consistent bias is more actionable than random errors.
• Last Month comparison — predicted vs. actual for the most recent completed month, as a quick monthly sanity check.

Adjust Forecast

The Adjust Forecast panel at the top of the Performance tab lets you apply a percentage uplift or downlift to a SKU's forecast for a specific date range. Running a promotion in April and expecting +40%? Enter +40% with start and end dates and Forthcast factors it into reorder calculations for that window only. Use a negative percentage to dampen demand when you expect a softer period. Adjustments stack with the underlying model — seasonality and trend are preserved — and active adjustments are listed below the form so they can be removed at any time. Adjustments are visible in the SKU detail panel too, so anyone reviewing a recommendation can see why the number is what it is.

Found under Inventory → Stock Health. A bird's-eye view of your whole catalogue, grouping SKUs by health so you can see where inventory is solid and where it needs attention.

Lost Sales banner

If a SKU was out of stock during a period when it would normally have been selling, Forthcast estimates the revenue lost as a result. The total appears in a banner at the top of the report and is broken down per SKU below — use it to quantify the cost of stockouts and to decide which items to prioritise.

Slow-moving stock

Items selling much slower than expected relative to how much stock you're holding. They tie up capital and storage. The report highlights them so you can consider markdowns, promotions, or reducing future order quantities.

Found under Settings → Bundles. Map bundle parents to their components so demand and stock add up correctly.

Why bundles need mapping

When you sell a bundle, customers buy the parent SKU but the stock you actually consume is the components inside it. Without a mapping, Forthcast would forecast the parent on its own and miss the demand it places on each component — leading to overstock on the parent (you don't reorder a kit, you assemble it) and understock on the components.

Mapping a bundle

Choose the bundle parent SKU, then add each component SKU and the quantity per unit. A 6-pack is one component at quantity 6. A gift set with three different items is three components at quantity 1 each. A variety pack is each flavour as a component at quantity 1, so demand spreads across all of them in proportion to how often the variety pack sells.

Auto-detection

Forthcast scans your catalogue for likely bundles by looking at product titles and SKU patterns ("6-pack", "variety", "set of 3", and similar) and surfaces suggested bundles in Settings → Bundles with a confidence score. Review each one and confirm — nothing is applied to your forecast until you accept the mapping.

Hiding assembled SKUs from reorder

Once a bundle has a mapping, the parent SKU is automatically excluded from reorder recommendations — the report will only show its components, since that's what you actually buy. To remove the parent from Stock Health and other reports as well, add it to Settings → Exclude Products; the bundle mapping itself stays active.

SKU multipliers (case packs)

Bundles describe how you sell. SKU multipliers describe how you buy. If a supplier ships a SKU only in cases of 12, set the case size on that SKU's lead-time settings — Forthcast rounds order quantities up to the nearest case so what's recommended is also what's orderable.

Found under Inventory → Packaging. Forecast cartons, mailers, inserts, and other packaging materials from your finished-goods demand.

How packaging forecasts are built

For each packaging material you add, link it to the finished SKUs that use it and set how many units of the material each finished unit consumes (a mailer is usually 1; a shared insert may be 1 per order; tissue paper might be a fraction). Forthcast rolls up demand for the linked SKUs into a projected packaging quantity with its own reorder recommendation — calculated the same way as a regular SKU.

Why this matters

Running out of mailers or branded boxes stops you shipping just as effectively as running out of product. Treating packaging as a forecasted item — instead of guessing from past purchase invoices — means your packaging reorders track real customer demand, so you stop overbuying or running short.

Found under Orders. Manage incoming stock orders and track what's in transit.

Creating a purchase order

Click New PO on the Purchase Orders page, or start one directly from suggestions in the Replenishment table. Forthcast pre-fills suggested order quantities. Edit any quantities, set the supplier, expected arrival date, and supplier reference number, then save the PO as a Draft.

PO statuses

• Draft — saved but not yet confirmed with your supplier.
• Ordered — sent to the supplier. Units are now counted as In Transit in the Replenishment table.
• Received — stock has arrived. Units are removed from In Transit and added to your Shopify inventory at the selected location. Partial receiving is supported — the PO stays open until all items are received.
• Cancelled — the order didn't proceed. Units are removed from In Transit.

Backorder tracking

When a customer order can't be fulfilled because stock has run out, the unfulfilled units are tracked as backorders. Forthcast counts them per SKU and shows the total in the optional Backorders column on the Replenishment table — turn it on via the Customize button. Reorder calculations factor backorders into priority so SKUs with customers waiting are surfaced first; resolved backorders disappear automatically once the order ships.

Forwarding supplier emails

Each store has a private inbound email address shown on the Purchase Orders page. Forward your supplier's confirmation, invoice, or shipping document to that address — Forthcast reads the message and any PDF attachments and matches them back to the right open PO automatically. The PO is updated with the supplier reference, confirmed quantities, prices, and ETA, and the In Transit numbers in your Replenishment table follow. If a forwarded email can't be matched with confidence, it's held for review on the Purchase Orders page so you can attach it manually. Nothing is changed on a PO until the match is confirmed.

Multi-location

Assign each PO to a specific receiving location. Inventory is updated at the correct Shopify location when the PO is received.

Forthcast automatically excludes subscription and wholesale orders from your seasonal demand pattern using seven detection signals. They run on every order with no tag setup needed for most stores.

Subscription signals

• Order tags: Skio, ReCharge, Bold, Smartrr, Loop, Ordergroove, Appstle, Seal, Stay AI, Subify, Paywhirl, plus any tag containing subscription, recurring, or autoship.
• Shopify Selling Plans (native subscription billing) on any line item.
• Third-party app properties on line items (ReCharge, Bold subscription metadata).
• Recurring customer pattern: customers with 3+ orders in 12 months averaging less than 60 days apart.

Wholesale / B2B signals

• Order tags: wholesale, b2b, trade, reseller, bulk, distributor, stockist, key-account, and related patterns.
• Discount codes with wholesale-indicating text (e.g. WHOLESALE20, B2B-TRADE).
• Sales channel: Draft Orders and the Shopify B2B portal.
• Company field: orders placed via the Shopify B2B company portal.
• Statistical outlier: order quantity 3+ standard deviations above your SKU average.

Detected wholesale and bulk orders remain visible in your demand history so you can see the full picture, and any individual order can be re-classified by hand.

Add your suppliers in Settings → Suppliers. For each supplier you can set a name, contact email, and average lead time. Forthcast uses lead time to calculate reorder points and safety stock.

Supplier reliability grades

Forthcast automatically grades each supplier (A through D) based on the delivery performance recorded across your purchase orders — primarily on-time rate and order completion rate. Grades are updated each time a PO is received.

Grades give you a quick view of which vendors are most consistent. To hold extra stock for an unreliable supplier, add a buffer on the affected SKUs in Settings → Buffers — buffers are the supported way to lift safety stock above the calculated value.

The Settings section lets you configure global defaults for Forthcast:

• Lead Time — the single most important input. Set a store-wide default and override per supplier or per SKU. If lead time is wrong, every reorder point is off by the same margin.
• Service level target — choose 90%, 95%, or 99% to control how conservatively safety stock is calculated. 95% is the recommended starting point. Can be overridden per SKU.
• Buffers — add a fixed extra quantity on top of the calculated reorder point or safety stock for a SKU or group of SKUs. Use buffers when you have a known reason to hold more stock that the forecast doesn't capture (a hard-to-restock item, a planned promotion, an unreliable supplier). Buffers can be edited or removed without affecting the underlying forecast.
• Suppliers — add, edit, and remove suppliers. Set default lead time, MOQ, and purchase price per supplier.
• Bundles — map bundle parents to component SKUs so demand and reorder suggestions land on the components you actually buy.
• Exclude Products — remove specific SKUs from your reports entirely. Useful for discontinued items still in your Shopify catalogue, samples, gift cards, or anything you manage stock for outside Forthcast. Re-include any time by removing the SKU from the exclusion list.
• SKU multipliers — case-pack sizes for products that ship from suppliers in fixed lots (e.g. 12 per case). Set on the SKU's lead-time settings; Forthcast rounds order quantities up to the nearest case.
• Notifications — Configure low-stock email alerts. Choose a daily or weekly digest of products below their reorder point, or instant alerts when a specific product drops below its threshold.
• Sync settings — control how frequently Forthcast syncs inventory and orders from Shopify.

The Customize button at the top of the Replenishment table opens the column chooser — toggle any column on or off, including optional ones like Backorders. Your preferences are saved in your browser so the same layout appears next time you open the app.

The search bar filters by SKU name or product title. You can also filter by status badge (Order Now, Plan Ahead, etc.) and by supplier so you can quickly pull up only the SKUs relevant to an upcoming order.

Forthcast connects to additional platforms beyond your main Shopify store. Set up everything from the Integrations page.

• Shopify — your primary connection. Order history, inventory levels, and product data sync automatically. No additional setup needed.
• TikTok Shop — connect your TikTok Shop account to include TikTok sales in your demand data.
• Notion (Purchase Order Sync) — export your purchase orders to a Notion database for tracking and team visibility.
• Amazon FBA (Beta) — available to connect, but contact us before relying on it for live inventory data.
• QuickBooks (Beta) — available to connect, but contact us before relying on it for live data.
• Xero (Beta) — available to connect, but contact us before relying on it for live data.

The Forthcast AI Chat is available from the Reports page — open it from the bottom-right corner of that page. Ask any inventory question in plain English and get a specific answer based on your store's data.

Example questions

• "Which products are at risk of stocking out in the next 30 days?"
• "What should I reorder from Supplier A?"
• "How accurate were my forecasts last quarter?"
• "Which products have been out of stock the longest?"
• "Show me my most overstocked items."
• "What's my total value of inventory on order?"

The AI Chat understands Forthcast's data model — forecasts, safety stock, purchase orders, supplier grades, and more — and can surface insights you might not immediately find by browsing the app manually.

Have a question not covered in these docs? Check the Frequently Asked Questions page for answers to common questions about setup, billing, data privacy, and more.

Still stuck? Contact our support team — we respond within one business day.