ModaViva (real case — data altered under NDA), a fashion retail chain with 5 physical stores in central Portugal and an online shop, managed a catalogue of 3,200 product references with a stock system that was updated only once a day — manually, at the end of the working day. The result was chaotic: online sales of products already sold out in stores, stock transfers between stores based on guesswork, and returns caused by items shipped in the wrong size or colour. The director, Pedro Alves, calculated that stock problems were costing the company over EUR 95,000 per year in lost sales, returns and operational inefficiency.
The Problem in Detail: Stock Nobody Could See
Invisible stock-outs. When an item sold out in a store, the information only reached the central system at the end of the day. This meant that throughout the entire day, the website showed a product as available that no longer existed. On the busiest dates (sales, Black Friday, Christmas), the lag was even worse — products sold out within hours but remained on sale online for the rest of the day. Online orders for unavailable products had to be cancelled with an apologetic email to the client, generating frustration, negative reviews and loss of trust.
The numbers were alarming: on average, 12% of online orders had to be cancelled due to stock unavailability. Each cancellation cost the company not only the lost sale but also the processing cost (team time, client communication) and the risk of losing the client permanently. Analysis showed that only 15% of clients whose orders were cancelled placed a new order — the remaining 85% never returned.
Inefficient stock transfers. When one store ran out of a popular item but another had excess, the transfer was coordinated by phone between managers. There was no centralised real-time visibility of stock by store — each manager knew only their own stock, and the information was frequently outdated. Transfers took 2 to 3 days to materialise, during which the store without stock lost sales and the store with excess occupied warehouse space with products that could have been selling at another location.
Avoidable returns. The online shop's return rate was 22%, significantly above the Portuguese fashion e-commerce sector average (15–18%). Analysis of return reasons revealed that 35% of returns were motivated by stock problems: an item shipped in a different size from the one ordered (because the correct size had sold out between order and fulfilment), a colour slightly different from the photograph (stock mixed between seasons), or a defective item that should have been removed from circulation.
The Solution: Unified Real-Time Stock
Multi-Store POS Integration
We implemented a unified point-of-sale (POS) system across all 5 stores, connected in real time to a central stock database. Every sale, every return, every transfer was recorded instantly — not at the end of the day, but at the moment it happened. When a client bought a size M t-shirt at the Coimbra store, the stock for that reference was updated in under 10 seconds across all other stores and the online shop.
The POS system was integrated with barcode readers in every store, eliminating the need for manual counting. Every item entering the store (supplier delivery), leaving (sale or transfer) or being returned was recorded with a simple scan. Counting errors — which previously caused significant discrepancies between theoretical and actual stock — were virtually eliminated.
For the online shop, the integration meant that the stock visible to the client was the actual aggregate stock across all stores and the central warehouse. If only 1 unit of an item remained across the entire chain, the website displayed "Last unit available" — creating urgency and preventing the sale of non-existent items. When stock reached zero, the item was automatically removed from the online shop or marked as "Unavailable — Notify me when back in stock".
Does your retail chain have stock problems?
We integrate POS, e-commerce and ERP systems for full real-time stock visibility. No stock-outs, no surprises.
View ERP & Integrations →Intelligent Replenishment Alerts
For each product reference, we configured two alert levels: the replenishment level (when to reorder) and the critical level (when to urgently transfer from another store). These levels were calculated automatically based on sales history, seasonality and supplier lead time.
For example, if a jeans reference sold an average of 8 units per week at the Aveiro store and the supplier took 10 days to deliver, the system set the replenishment level at 16 units (2 weeks' cover) and the critical level at 5 units. When stock reached 16, a supplier order suggestion was automatically generated. When it reached 5, an alert was triggered for the manager with a transfer suggestion from the store with the most stock.
The system also detected anomalous patterns: if an item that normally sold 3 units per week began selling 10 (for instance, due to a social media mention or a trend), the replenishment alert was automatically adjusted to reflect the new demand. This predictive capability prevented dozens of stock-outs that, under the previous system, would only have been detected once the item was already sold out.
Management Dashboard and Automated Transfers
Pedro now had a real-time dashboard showing the stock of each reference at each location, with visual alerts for items at critical level, items with excess stock and discrepancies between stores. This enabled informed decisions in minutes rather than hours — or rather than not making them at all, as frequently happened under the previous system.
Inter-store transfers went from a 2–3 day process coordinated by phone to a digital workflow of under 24 hours. When the system identified that the Viseu store had 20 units of an item that was not selling and the Leiria store had a stock-out of the same item, it automatically suggested a transfer. The Viseu manager received a notification, approved with a tap, prepared the items with an automatically generated transfer note, and the Leiria store's stock was updated at the moment of receipt.
Implementation: The Store-by-Store Migration
The implementation was phased to minimise operational disruption. The total project lasted 12 weeks, with each store migrating to the new system in a dedicated week. The first store served as the pilot — the Coimbra store, being the smallest and closest to the support team. After 2 weeks of stable operation, the remaining 4 stores migrated in sequence.
The biggest challenge was the initial stock count. The discrepancy between theoretical stock (in the records) and actual stock (on the shelves) averaged 8% — significantly more than Pedro had expected. Some references had more stock than recorded (items received but not entered); others had less (theft, unrecorded breakages, prior counting errors). This initial count, although labour-intensive, was fundamental to the integrity of the entire system — and revealed problems that had been costing the company thousands of euros without anyone knowing.
Staff training focused on simplicity: scan the product on receipt, scan at the point of sale, scan on return. Managers received additional training on the management dashboard and the transfer system. Within 2 weeks, the entire team was comfortable with the new system.
Results After 6 Months
• Online orders cancelled due to stock unavailability: from 12% to 0.3%. Virtually eliminated — the rare remaining cases were due to simultaneous sales of the last unit (online and in-store in the same minute).
• Online return rate: from 22% to 9%. The elimination of incorrect shipments and stock accuracy reduced returns by 60%.
• Sales lost to stock-outs: estimated 78% reduction. The alert and automatic replenishment system ensured the best-selling items were always available.
• Inter-store transfer time: from 2–3 days to under 24 hours. The agility in stock redistribution maximised sales across all locations.
• Inventory accuracy: from 92% to 99.2%. The discrepancy between theoretical and actual stock was virtually eliminated.
• Annual revenue: 18% increase, from EUR 1,350,000 to EUR 1,593,000 — an additional EUR 243,000 in sales, primarily from reduced lost sales and returns.
• Stock management time: from 15 hours/week (counting, phone calls and manual coordination) to 2 hours/week of dashboard monitoring.
Impact on the Client Experience
The most significant impact was on online client trust. The NPS (Net Promoter Score) of the online shop rose from 18 to 52 — a radical transformation. Negative reviews about "cancelled orders" and "item different from what was ordered" virtually disappeared. And the "Notify me when back in stock" feature generated sales that previously would have been permanently lost: 28% of clients who activated the availability alert actually purchased when the item returned to stock.
In the physical stores, the impact was equally visible. Staff began trusting the system to check availability of sizes and colours at other stores — instead of saying "we do not have it", they could say "we do not have it here, but I can reserve it from the Aveiro store and have it delivered to your home tomorrow, at no cost". This assisted selling capability between stores generated a 7% increase in total in-store sales.
Conclusion
ModaViva went from a retail chain that did not know what it had to an operation with full real-time visibility. The EUR 12,500 investment (hardware, software and implementation) generated a return of EUR 243,000 in additional revenue in the first year — an ROI of nearly 20x. But the true gain was operational peace of mind: Pedro, who previously spent half his time resolving stock problems, now dedicates that time to what truly matters — buying better, negotiating with suppliers and planning the chain's growth.
