From your suppliers’ spreadsheets to sent purchase orders.

RestockIQ connects through Amazon’s Selling Partner API — the official mechanism Amazon provides for seller tools. You authorise it from your own Seller Central account, and you can revoke that authorisation from Seller Central at any time. RestockIQ requests the standard roles for product listings, inventory and order tracking, and finance data. It never sees your Amazon password, and it never stores your customers’ names, addresses or contact details.

Drop in the file your supplier actually sent — CSV, XLSX or XLS. Map its columns to RestockIQ’s fields once (product, barcode, cost, stock, case size); the mapping is remembered for next time. Currencies are handled per supplier and converted with daily historical FX rates.

Each supplier barcode is matched to its Amazon listing. Barcodes are occasionally reused for entirely different products — the classic trap that turns a pallet of the wrong item into a very expensive lesson. So before any purchase order can be sent, every line’s supplier title is checked against the Amazon listing title by an automated language-model comparison. Anything that doesn’t plausibly describe the same product is flagged, and a human must confirm or remove it before the order goes anywhere.

Discover shows one row per barcode across every price list you have uploaded, with the cheapest supplier for that product on the row and the rest a click away. Each row carries the full profit breakdown: the Amazon price including VAT, VAT extracted at 20/120, the referral fee, the FBA fulfilment fee, inbound shipping, and your supplier cost. What’s left is profit; profit over landed cost is ROI.

The market price column is deliberately three-state. A clean number means a confirmed buy box. Where there is no buy box, it says so. Where the figure is only the lowest offer, it says that instead. You are never left guessing which one you are looking at.

Prices spike, so profitability is gated on the lower of today’s price and the average over your chosen horizon. A spike is flagged in amber, and one click re-runs the row at the average price so you can see the deal you would actually be getting.

For each marketplace, Replenish assembles your true position: stock available at Amazon, shipments already in transit to Amazon, stock in your own warehouse, and any inbound you have declared. Units in transit are credited on the day they are projected to arrive, not the day the shipment was created — a container landing in three weeks does not get to cover next week’s sales.

Demand comes from your own sales history, shaped by recurring seasonal curves per category and marketplace, with the seasonal effect removed from the variance before safety stock is computed so the season is never counted twice. The reorder point covers your supplier’s lead time plus Amazon’s inbound transit time, because stock in your warehouse is not sellable stock yet.

Build a draft basket from Discover and Replenish, set a budget, and the optimiser returns several complete purchase plans at different spending levels — so “what would we do with £5,000 versus £15,000?” is a click, not an afternoon.

Every plan respects each supplier’s minimum order value, minimum order quantities, case pack sizes, your budget, and cheaper alternate suppliers for the same product. Under the hood this is genuine mixed-integer optimisation — the same class of mathematics industrial supply chains run on — not a sort-by-ROI spreadsheet. It optimises economic profit, which prices in your cost of capital and the cost of holding slow stock, so it will not tie up capital in inventory that takes months to sell just because the margin looked good.

Orders move through draft, ready, sent and received. Nothing reaches “sent” until every line has passed — or a human has explicitly confirmed — the title-match check.

Every sale is matched to a cost through a six-tier, time-aware cascade — from an exact purchase cost where one exists, down through progressively broader fallbacks — and the interface tells you which tier each figure came from, so you know how firm it is. Fees come from your actual Amazon settlement data where available: what Amazon really charged, not an estimate of what it usually charges.

Multi-marketplace sellers earn in several currencies and buy in others. Profit is computed at the daily historical rate for the sale date, then reported in your home currency.

Market prices and sales-rank history come from Amazon’s own APIs and from Keepa, the industry-standard price-history service. Your sales, stock and fee data come directly from your Amazon account. Every derived figure in the interface can be traced to its source — hover it and it will tell you what it used and when that data was last updated.