QUICK OVERVIEW
What to take away from the article
- Physical status, bookings and expected income are different data.
- The order point depends on the consumption during delivery and on the selected reserve.
- Approve purchase proposals first; expand automatic ordering after data verification.
In stock does not mean available for sale
There can be 80 pieces on the shelf, of which 50 already belong to confirmed orders. If the e-shop shows all 80 as available, it can sell the same item twice. Conversely, the expected income for tomorrow is not a physically received good. The status for the customer must correspond to the rules of sales and delivery reliability.
Differentiate physical condition, reserved quantity, expected income and planned expenses. Each value should have a source and update time. Separately deal with damaged goods, returns awaiting inspection and another company's stock. The right model of ownership and updates is explained by e-shop and ERP integration without duplicates.
A simple order point model
The fictitious store sells an average of 12 pieces of packaging material per business day. Delivery takes five working days. Expected consumption during delivery is 12 × 5 = 60 pieces. The company chooses a reserve of 24 pieces for this simple model, so the order point is 84 pieces.
A reserve of 24 pieces is a chosen assumption, not a general recommendation. In practice, it will be affected by sales fluctuations, supplier delays, required availability and inventory value. Seasonal product or irregular large B2B orders may not be well described by the average. For the pilot, therefore, compare the proposal with specific known orders.
| Area | Postup | What to check |
|---|---|---|
| Average consumption | 12 pieces / working day | Assumption of stable sales. |
| Delivery time | 5 working days | From ordering to usable receipt. |
| Consumption during delivery | 60 pieces | 12 × 5. |
| Selected reserve | 24 pieces | Illustrative safety cushion. |
| Order point | 84 pieces | 60 + 24. |
How much to order: the replenishment point is not the target state
An order point tells when to act. The target state determines where you want to replenish the stock. In the continuation of the model, let's choose a target of 180 pieces and a stock position of 72 pieces, while in this position we already include known income and binding expenses according to the defined rule. The design is 180 − 72 = 108 pieces. In a pack of 24, rounding up would give 120 pieces.
Before sending the purchase, the minimum quantity, price, space in the warehouse and open orders must be verified. Otherwise, each new calculation may establish another demand for the same missing goods. Therefore, the quantity already ordered and its expected date must be taken into account.
How Odoo fits into it
Odoo describes replenishment rules with a minimum and maximum quantity and with a link to a suitable supply route. Depending on the configuration, they can create purchase or production requisitions. The choice of automatic or manual mode must correspond to how reliable your data and approvals are.
Do not take the setting from the template without checking. For the product, verify the unit, packaging, supplier, delivery time and rounding method. In particular, try partial deliveries and moving the date. Automation can be technically successful and yet commercially inappropriate if it systematically buys too much unnecessarily.
Exceptions to be handled by a human
Approvals include an unusual price jump, a large order, an uncertain deadline, and a product with declining sales. AI can point out a deviation and suggest an explanation, but a binding purchase must respect the company's limits. Seasonal stocking is a business decision, not just a mathematical continuation of last week.
The work report should show the reason for the proposal, open orders, available budget and the consequence of the postponement. In this way, the worker approves a specific decision instead of manual data collection. Morning exceptions can be listed in Owner and Manager Dashboard.
Evaluate the pilot according to the availability of committed money
Select a limited group of stable products and first compare the automatic suggestions with the buyer's decision. Track days out of stock, value of slow-moving goods, urgent backorders and processing time. The number of auto-generated requests alone is not an indicator of success.
After verification, introduce limits and expand the assortment by groups. In case of major changes in prices or delivery times, review the rules again. High-quality warehouse automation should reduce manual tracking and improve availability without uncontrolled inventory growth.
