Stock optimisation: indicators, methods and software

Optimising stock means having enough of every product to miss no sale, without tying up cash in what does not sell. In practice: track five or six indicators, set a reorder point per product, and decide every week what to reorder, move or leave on the shelf. Software does this work across all your products and all your shops at once, where a spreadsheet stops at a few dozen rows.

That is exactly what Sway One does for shops that sell in store and online. Here are the indicators, the methods, then the steps to optimise your stock with software.

Sway One and Sweet 1.4 in two sentences

Sway One is the app that plugs into your tools: tills like Square, SumUp, Lightspeed or Hiboutik, e-shops like Shopify, PrestaShop or WooCommerce, ERPs like Odoo or Dolibarr, logistics providers like Bigblue. More than twenty tools, which you connect yourself from the Connections page, in a few minutes.

Sweet 1.4 is its decision AI: it forecasts demand, suggests restocks, transfers and discounts, and no longer discounts by reflex. Every recommendation arrives in English, with its reason, its estimated impact and a confidence level. You approve it in one click, adjust it or discard it.

Why optimise your stock

Badly tuned stock costs you in three ways, which barely show on a balance sheet but are paid every month.

Stock-outs: the sale you do not see

A missing product leaves no trace: the customer buys elsewhere, and your history says "zero sales". If you restock on that history, you order less next time. The stock-out sustains itself.

Overstock: cash that sits idle

Every item in reserve has been paid for, and that money does nothing until it is sold. On top of that comes the holding cost: space, insurance, handling, financing, breakage, obsolescence. As an order of magnitude, it is often counted at between 20 and 25% a year of the stock value. A €10,000 stock that sits for a year therefore costs €2,000 to €2,500, before any discount.

Markdown: selling off to escape a buying mistake

A well-placed markdown saves the end of a season. A markdown out of habit turns a slow item into an item sold at a loss.

So the goal is not to have the least stock possible. It is to have the right stock, in the right place, at the right time.

The indicators to steer your stock

Calculate these six indicators per product, per family and per shop: a global average almost always hides the problems.

Turnover and cover

Example: €120,000 of cost of sales over the year for an average stock of €30,000 at cost is a turnover of 4, or about 91 days of stock on average (365 ÷ 4).

Cover is read product by product: 18 units in stock at 1.5 sales per day is 12 days of cover. If your supplier delivers in 15 days, you are already late.

A "good" turnover depends on your trade: compare yourself with yourself, family by family, from one season to the next. For the stock-out rate, which nothing records, a simple approximation: days out of stock × sales of a normal day.

Dead stock

Choose a threshold suited to your pace: 60 days without a sale for fashion, more for furniture. But an item with no sale in 60 days is not necessarily dead. While preparing Sweet 1.4, we saw on our simulated shops "dormant" items that were still selling one to five pieces a month. They were slow, not lost.

Margin after carrying cost

Example: €10,000 of gross margin over the year, an average stock of €8,000 at cost and a carrying cost of 25% a year give 10,000 − 2,000 = €8,000 of real margin.

With Sway One: the dashboard gives revenue, gross margin, sell-through rate, cover and stock-outs, channel by channel and shop by shop, tills and e-shop reconciled, net of returns.

Methods that work

Classify your products: ABC and XYZ

The ABC method sorts your references by weight in revenue or margin. Often, a small share of references makes up most of the sales: these are your As, never to be left out of stock.

The XYZ method adds regularity: X sells every day, Z from time to time. An AX product is restocked almost automatically; a CZ product is ordered on demand, or leaves the range.

With Sway One: each product receives a demand profile (smooth, erratic, intermittent or lumpy), and Sweet 1.4 chooses how to forecast it accordingly.

Safety stock

It protects you from a week stronger than expected or a supplier running late. Classic formula: service coefficient × standard deviation of daily sales × square root of the lead time in days.

With Sway One: safety stock is calculated for each product, on its own demand distribution and at the service level set for your shop.

Reorder point and min/max

The reorder point is the demand during the delivery lead time, plus the safety stock. Example: 4 sales per day, a 10-day lead time and 12 units of safety stock give a reorder point of 52.

The min/max method is the simple version: when stock falls below the min, you order up to the max. Effective on steady products, too rigid for a seasonal product.

With Sway One: the reorder point is calculated for each product on its real supplier lead time, and your own thresholds per family or per reference (minimum cover, minimum stock) are set in your deployment's configuration. The statistical calculation remains a floor: your threshold can only raise it.

The critical fractile, or how much to order for a season

For a seasonal purchase, the question is: what costs more, missing one piece or having one too many? Example: an item bought at €20 and sold at €50. Missing a sale costs €30 of margin. Keeping an unsold piece costs, say, €10 in carrying cost and write-down. The critical fractile is 30 ÷ (30 + 10) = 75%: order enough to cover demand in three seasons out of four.

With Sway One: Sweet 1.4 does this calculation for every restock, with your margin, your holding cost and the salvage value of unsold items.

FIFO and FEFO

First in, first out (FIFO), or first expired, first out (FEFO) for dated products: it is a storage and picking rule, essential as soon as your products age.

How do I optimise my stock with software

1. Connect your tills, your e-shop and your logistics

Software is only as good as the sales you give it: those of the tills and those of the site.

With Sway One: more than twenty connections, including Shopify, PrestaShop, WooCommerce, Square, SumUp, Lightspeed, Zettle, Hiboutik, Odoo, Dolibarr, Bigblue and Amazon FBA. You connect them yourself, the connection is tested straight away, and Sway One only reads: it writes nothing in your tools.

2. Make the data reliable

Duplicate references, missing barcodes, zero sales during stock-outs: clean up before calculating.

With Sway One: products are matched across channels by reference or barcode, and stock-out days are corrected: a product that is out of stock is not a product that does not sell.

3. Set thresholds per family and per product

Your best-sellers deserve more cover than your end-of-line items.

With Sway One: thresholds per family, then per reference, and every recommendation cites the rule applied.

4. Let the software calculate the restock

The calculation must start from forecast demand, stock on hand, what is already on order and the supplier lead time.

With Sweet 1.4: demand is forecast day by day, with an uncertainty band visible on the product page, then translated into a quantity to order, shop by shop. What AI changes here is detailed in Stock optimisation with AI, including on Shopify.

5. Place consolidated purchase orders

Restocking line by line misses free-shipping thresholds and order minimums.

With Sway One: restocks are grouped by supplier and by shop, with minimums, free-shipping thresholds and lead times, then turned into editable purchase orders, exported as CSV or PDF.

6. Receive stock-out alerts per shop

The useful alert is not "you are out of stock". It is "you will be out of stock before the restock has time to arrive".

With Sway One: that is exactly this alert, calculated shop by shop, in the app and on Discord.

7. Transfer between shops before reordering

When one shop is short of a product that is sitting in another, the transfer costs less than an order.

With Sweet 1.4: transfers between shops are suggested with their quantities, next to the restocks.

8. Measure the outcome

With Sway One: once the horizon has passed, the estimated impact of each decision is compared with real sales, and the following estimates are recalibrated.

Spreadsheet, ERP, WMS, e-commerce app or optimisation software

Each family of tools meets a different need, and many shops combine several.

Sway One belongs to the last family. It replaces neither your till nor your ERP: it builds on them.

The criteria for choosing your software

Before signing, check:

• All your sales: shop tills and e-shop, reconciled.
• Multi-shop for real: every screen filterable by shop, and transfers suggested.
• A forecast per product, which corrects stock-outs and takes seasons into account.
• Adjustable thresholds per family and per reference.
• Purchase orders that respect minimums, free-shipping thresholds and lead times.
• Stock-out alerts before it is too late to react.
• Discounts justified by what they earn, not by the age of the stock.
• Explained recommendations, with a quantified impact and a confidence level.
• A measure of the outcome, decision by decision.
• Your data stays yours: dedicated deployment, and nothing used to train a model.

This list is Sway One's specification: Sway One ticks every box.

Common mistakes

• Systematically discounting what does not sell. A slow item is not a dead item. Sweet 1.4 only suggests a discount if it earns more than waiting at full price: that is the principle of Sweet 1.4.
• Restocking on last month. Last month perhaps contained a stock-out, a public holiday or a heatwave.
• Applying the same rule to the whole shop. A best-seller and an end-of-line item do not have the same stock-out cost.
• Steering by margin rate alone. You can gain margin points by selling less: the result is also read in euros, after carrying cost.
• Never measuring. Without comparing the estimate to the outcome, you repeat the same mistakes.

Why Sway One and Sweet 1.4

• Everything plugs in: more than twenty tills, e-shops, ERPs and logistics providers, connected by you in a few minutes.
• All your shops and your e-shop on one screen, filterable shop by shop.
• Recommendations in English, with their reason, their estimated impact and a confidence level, approved in one click.
• From restocking to the purchase order, with minimums, free-shipping thresholds and lead times, exported as CSV or PDF.
• Sweet 1.4 no longer discounts: a discount is only suggested if it earns more than waiting.
• Every decision measured, and your AI assistant connected via MCP.

The measured result: on four simulated shops (sport, outdoor, furniture, fashion), replayed over 26 weeks, Sweet 1.4 adds +4.7% to +11.6% of margin after carrying cost to the retailer's routine, with fewer stock-outs in all four cases. These are simulated shops: the real gain is measured on your data.

Sway One costs €289 excl. VAT per month, with a deployment dedicated to your shop.

How much would optimised stock earn you?

The gains calculator starts from your revenue, your margin and the time spent on stock, and puts a figure on three scenarios over twelve months, subscription deducted.

To see your own opportunities, connected to your data, book a demo: 30 minutes, no commitment.

Frequently asked questions

How do I optimise my stock?

Track turnover, cover, stock-outs and dead stock per product and per shop, set a reorder point per reference, then decide every week what to restock, transfer or keep on the shelf. Then measure what each decision earned.

Which software should I choose to optimise my stock?

Software that reads all your sales, shops and e-shop, forecasts per product, respects your supplier constraints and measures the result. Sway One, with its decision AI Sweet 1.4, is built for this.

How do I optimise my stock with software when I have several shops?

Connect the till of each shop and your e-shop, track stock by location, and let the software suggest transfers before reordering. Sway One reasons shop by shop and sends stock-out alerts per shop.

How do I calculate stock turnover?

Divide the cost of sales for the period by the average stock at cost. A turnover of 4 over a year means the stock renews four times, or about 91 days of stock on average.

Is a spreadsheet enough to manage stock?

For a few dozen references in a single shop, often yes. Beyond that, with several shops or an e-shop, manual entry and the lack of forecasting cost more than software.