Inventory discrepancies between website and supplier: how to avoid order problems

Why do balance discrepancies occur: 6 main reasons

Before solving a problem, you need to understand its root cause. Balance discrepancies rarely arise from a single cause—usually it's a combination of several factors.

1. Data synchronization delay

The most common reason is a time lag between a change in the supplier's inventory and the update on your website.

Typical scenarios:

• Scheduled synchronization: The supplier updates the inventory every hour, and during this time the goods have already been sold through other channels
• Nightly updates: The data is updated once a day at night - a lot changes in 24 hours
• Manual download of price lists: The manager uploads files with a delay of several hours or days
• Slow processing: Importing a large price list takes 30-60 minutes, during which time the data becomes outdated

2. Omnichannel supplier sales

If your supplier works with multiple clients at the same time, their products are sold through multiple channels in parallel.

Omnichannel Challenges:

• Competition for leftovers: The same product is sold by 10 other dropshipping stores
• Retail outlets: The supplier may have offline stores that sell from the same warehouse
• Your own online store: the supplier competes with you for its own goods
• Marketplaces: Amazon and eBay are taking a significant portion of the remaining inventory.
• No reservations: the supplier does not reserve goods for specific partners

Result: by the time you receive the updated price list, the product has already been sold in another channel, but your website still shows it as "in stock."

3. Human errors during manual updates

When managers manually update inventory, errors are inevitable.

Error types:

• Typos: instead of "15 pieces" it says "150"
• Forgotten goods: We updated 500 positions out of 520, but forgot about 20.
• Incorrect identification: Updated the balance for article number ABC-123 instead of ABC-132
• Incomplete update: They updated the price but forgot about the remaining stock.
• Copying old data: accidentally downloaded yesterday's file instead of today's

4. Accounting differences between systems

You and the supplier may have different interpretations of the term "remainder".

Discrepancies in accounting:

• Reserved items: The supplier has 10 pieces, but 7 are already reserved for other orders - only 3 are available
• Goods in transit: There are 0 units physically in stock, but 50 more are on the way from the manufacturer—the supplier indicates 50 are "in stock"
• Marriage and returns: There are 15 pieces in stock, but 5 are defective - only 10 are actually available.
• Product under inspection: returns from customers who have not yet been verified
• Minimum balance: The supplier does not sell the last 2-3 units for insurance

5. Time zones and working hours

If you and the supplier are in different time zones, additional complications arise.

Time zone issues:

• Night sales: Your site operates 24/7, and the provider updates data only during business hours (9:00-18:00 in their time zone)
• Processing delay: An order from your website arrives at night according to the supplier's time and is processed only in the morning—the item has already been sold.
• Days off: The supplier does not work on Saturday and Sunday, the stock is not updated for 2 days.
• Holidays: differences in the holiday calendars of the USA, EU, and Ukraine

6. Technical failures and integration errors

Even automated systems fail.

Types of technical problems:

• Lost connection: FTP server is unavailable, API is not responding
• Changing data format: The supplier changed the price list structure, your parser broke.
• Queue overflow: Data processing is stuck, new updates are not applied
• Expired tokens: API keys are out of date, syncing has silently stopped
• Update conflicts: simultaneous change of one product from different sources

The Cost of the Problem: How Much Does a Business Lose Due to Discrepancies?

Before investing in a solution, it's important to understand the true cost of the problem. Residual discrepancies impact several areas simultaneously.

Direct financial losses

1. Cancelled orders

• Loss of revenue: Average order value: $120, 15 canceled orders per month = $1,800 in lost revenue
• Lost profit: at a 25% margin, that's $450/month or $5,400/year
• Time spent: The manager spends 20 minutes processing each canceled order (searching for an alternative, contacting the client)

2. The cost of attracting lost customers

• CAC (Customer Acquisition Cost): The average cost of customer acquisition in e-commerce is $45-80.
• 34% will not return: If 15 orders are cancelled, you lose approximately 5 customers forever.
• Loss of investment in marketing: 5 clients x $60 CAC = $300 wasted every month

Reputational damage

Negative reviews and ratings:

• Rating decrease on platforms: Trustpilot, Google Reviews, Facebook
• Negative word of mouth: One dissatisfied customer tells 9-15 friends
• Conversion rate drop: Potential customers read reviews and go to competitors
• CPO (Cost Per Order) growth: have to spend more on advertising to compensate for the decline in trust

Operating expenses

• Managers' time: processing cancelled orders, explaining to customers, searching for alternatives
• Support load: increase in the number of support requests
• Employee Stress and Burnout: Constantly apologizing to clients demotivates the team

Strategies to prevent discrepancies

Now on to the solution. An effective strategy includes multiple layers of protection, from technical automation to business processes.

Strategy 1: Increase the sync frequency

The more frequently data is updated, the less likely it is to inconsistencies. However, a balance must be struck between relevance and the load on systems.

Frequency recommendations:

Learn more about setup automatic synchronization of prices and balances with the website for different CMS platforms.

Strategy 2: Implementing Safety Stock

Safety stock is a reserve you subtract from the supplier's actual inventory before displaying it on the website. This protects against situations where the data isn't updated in a timely manner.

Formula for calculating safety stock:

Dynamic calculation of safety stock:

For more precise control, use a formula that takes into account sales velocity and update frequency:

Setting by category:

• High demand + rare synchronization: Safety Stock = 20-30% of the balance
• Average demand + frequent synchronization: Safety Stock = 10-15%
• Low demand + frequent sync: Safety Stock = 5-10%
• Exclusive products: Safety Stock = 1-2 pieces (fixed value)

Strategy 3: Product Reservation System

A reservation is a temporary blocking of a product while an order is being processed and confirmed.

How does the reservation work:

1. The customer adds an item to the cart: the system checks the available balance
2. Proceed to checkout: the item is reserved for 15-30 minutes
3. Payment: If payment is successful, the reserve is confirmed and the balance is reduced.
4. Time expires: If the client does not pay, the reservation is automatically cancelled.
5. Sending an order to the supplier: final confirmation of availability before shipping

Backup settings:

• Basket Reservation Time: 15-20 minutes (average order processing time)
• Reservation time after payment: 24-48 hours (before dispatch to supplier)
• Priority: Paid orders have priority over unpaid reservations
• Notifications: Automatic notification to the client 5 minutes before the reservation expires

Strategy 4: Rules of Priority and Conflict

When conflicting balance data comes from different sources, the system needs to know which source to trust.

Example of priority rules:

1. Priority 1 (highest): Manual update by administrator always takes priority
2. Priority 2: Real-time provider API - if available
3. Priority 3: Automatic download of price lists
4. Priority 4 (lowest): Cached data is older than 24 hours

Conflict resolution rules:

• If the data contradicts each other: choose a lower balance value (it's safer to show "out of stock" than to sell a non-existent item)
• If the source has not been updated for more than 48 hours: Automatically hide items marked "Check availability"
• If the balance drops to zero suddenly: Do not update automatically, send an alert to the manager for verification

Strategy 5: Monitoring and Automatic Alerts

Proactively detecting problems allows you to respond before customers are impacted.

Critical alerts:

• Synchronization has not occurred for more than 4 hours - immediate notification
• Bulk zeroing of balances (>50 items at a time) — there is likely an error in the data
• The order has been placed, but the supplier does not have the product. - requires manual processing
• The percentage of cancelled orders exceeded 5%. - a systemic problem

For more information on setting up monitoring, read the article. automatic synchronization of prices and balances.

Setting up real-time synchronization with the provider

Real-time synchronization is the gold standard for preventing discrepancies. Instead of periodically importing price lists, changes are transferred instantly.

Real-time synchronization methods

1. Webhooks (callbacks)

The supplier automatically sends a notification to your server every time the balance changes.

Advantages:

• Delay 1-5 seconds
• Minimal server load
• Only modified data is transferred

Requirements:

• The provider must support webhooks.
• You need an endpoint to receive data.
• Handler for parsing and applying changes

 https://your-shop.com/api/supplier-webhook

 { "event": "stock_updated", "timestamp": "2025-01-15T14:23:17Z", "product":{ "sku": "LAPTOP-XPS-15", "stock_quantity": 8, "status": "in_stock" } }

2. REST API with frequent requests

If webhooks are not available, use polling—regular requests to the provider's API.

Polling optimization:

• Differential frequency: Check popular items every 5 minutes, and check others every hour.
• Use the "since" parameter: request only the changes since the last request, not the entire directory
• Batch requests: Group the check of 50-100 products into one request
• Caching: keep the latest values, update only when there is a real change

3. Hybrid approach

Combination of methods for maximum reliability:

• Webhooks as the main channel - instant updates
• Polling as a backup — We check every 30 minutes to make sure no webhooks have been missed.
• Night full synchronization - once a day a full comparison of all products to eliminate discrepancies

Checking availability before finalizing an order

Even with real-time synchronization, add a final check before sending the order to the supplier.

Two-step verification:

1. Time of order placement: check against local database (instant)
2. Before sending to the supplier: Real-time API request for final confirmation of availability

This approach protects against the situation where several customers simultaneously order the last unit of an item.

Troubleshooting: Solving Common Problems

Even with a perfect setup, unexpected situations can arise. Here's a guide to diagnosing and resolving the most common issues.

Problem 1: Synchronization works, but there are still discrepancies

Possible reasons:

• Insufficient safety stock: the reserve is too small for the speed of sales
• Omnichannel sales: the supplier sells through other channels between your syncs
• Different understandings of "remainder": the supplier includes reserved goods in the general balance

Solution:

1. Increase safety stock by 5-10 units for problematic products
2. Ask your supplier to provide "available balance" instead of "total"
3. Increase the synchronization frequency for popular products to every 5-10 minutes
4. Implement a final availability check before order dispatch

Problem 2: Mass zeroing of balances

Signs:

• Suddenly, 100+ items show a remaining stock of "0"
• Happened in one go during synchronization
• The supplier actually has the goods

Possible reasons:

• Error in the supplier's price list: technical failure, empty column with residues
• Changing data format: The "remainder" column has been renamed or moved
• Parsing error: Your script interpreted the data incorrectly.

Solution:

1. Data rollback: Restore the previous version of the remains from the backup
2. Check the source file: Download the latest price list from the supplier manually and check the format.
3. Set up protection: If >10% of products are reset at a time, block the update and send an alert
4. Logging: keep source files for analysis

 // Pseudocode of the protection rule if (more than 10% of products are reset) { cancel_update(); restore_previous_values(); send_alert_to_administrator(); block_automatic_updates(); wait_for_manual_check(); }

Problem 3: Balances are not updated for some products

Signs:

• Some products update correctly, others do not.
• Problematic products are always the same

Possible reasons:

• Article numbers mismatch: your SKU and the supplier's SKU are different
• Duplicates in the database: one product was created twice with different IDs
• Incorrect mapping: The connection between your catalog and the supplier's price list is broken.

Solution:

1. Export a list of problematic products
2. Check the part numbers against the supplier's price list (the supplier may have changed the SKU)
3. Check the mapping table - update the mappings
4. Find and merge duplicate products in the database

Learn more about resolving duplicate issues in this article. How to eliminate duplicate products in the catalog.

Problem 4: Synchronization delays during peak hours

Signs:

• At certain hours, synchronization slows down
• The update queue is piling up
• The processing time for one item increases by 5-10 times

Possible reasons:

• High server load: the site's peak traffic coincides with synchronization
• API rate limits: The provider limits the number of requests per hour.
• Insufficient resources: the server can't handle the load

Solution:

1. Shift your schedule: run heavy synchronizations at night
2. Use queues: asynchronous processing via RabbitMQ, Redis Queue
3. Optimize queries: batch updates instead of single ones
4. Scale your infrastructure: add workers for parallel processing

Problem 5: Conflicts between multiple suppliers

Situation:

The same item is available from two suppliers with different stock levels. How do I determine which stock level to display?

Solution - aggregation rules:

• Summation: total balance = sum of balances from all suppliers (apply safety stock to each)
• Preferred Supplier: choose data from your preferred provider
• Best conditions: choose the supplier with the highest balance or the lowest price
• Source marking: show the client "Available from 2 suppliers"

Read more about aggregating data from multiple providers in this article. How to combine price lists from multiple suppliers.

FAQ: Frequently Asked Questions

1. What if the provider does not provide a synchronization API?

Answer: Есть несколько альтернатив:

• Автоматическая загрузка прайсов: настройте автоматическое скачивание прайс-листов с FTP, email или по прямой ссылке. Периодичность — каждые 1-4 часа.
• Парсинг сайта поставщика: если поставщик публикует остатки на своем сайте, можно настроить автоматический парсер. Легально, но требует согласования.
• Google Sheets + API: попросите поставщика вести остатки в Google Sheets, подключитесь через Google Sheets API.
• Консервативный safety stock: при редких обновлениях увеличьте страховой запас до 30-50% от остатка.

Additionally: автоматическая загрузка прайс-листов по email и FTP.

2. Как объяснить клиенту, что товар закончился уже после оформления заказа?

Answer: Честность и проактивность — лучшая стратегия:

1. Немедленное уведомление: свяжитесь с клиентом в течение 2-4 часов после заказа
2. Explanation: "К сожалению, последняя единица товара была зарезервирована за несколько минут до вашего заказа."
3. Alternative: предложите аналогичный товар, возможно, с небольшой скидкой
4. Compensation: промокод на 5-10% для следующей покупки
5. Priority: "Как только товар появится, вы будете первым в очереди"

Превратите негатив в возможность показать отличный сервис. 67% клиентов прощают проблему, если она решена быстро и с заботой.

3. Стоит ли скрывать точное количество остатков от клиентов?

Answer: Зависит от стратегии бизнеса:

Показывать точное количество:

• Plus: прозрачность, клиент видит дефицитность товара
• Plus: срабатывает FOMO (fear of missing out) — "осталось 2 шт, надо брать!"
• Minus: если товаров много (>100), клиент может подождать со покупкой

Показывать статус без точного числа:

• "В наличии" — более 10 штук
• "Мало" — 3-10 штук
• "Последние единицы" — 1-2 штуки
• "Под заказ" — 0 штук, но можно заказать

Recommendation: для дропшиппинга лучше не показывать точное количество. Используйте статусы — это безопаснее и все равно создает urgency.

4. Что лучше: показывать "нет в наличии" или "под заказ"?

Answer: Зависит от вашей модели работы и возможностей поставщика:

"Нет в наличии" + скрыть товар:

• Подходит для быстро меняющегося ассортимента
• Не перегружает каталог недоступными товарами
• Меньше разочарованных клиентов

"Под заказ" + возможность заказать:

• Подходит если поставщик может доставить товар за приемлемый срок (7-14 дней)
• Не теряете потенциальные продажи
• Позволяет собирать предзаказы
• Обязательно указывайте реальные сроки: "Доставка 10-14 дней"

Recommendation: если товар вернется в наличие через 3-7 дней — используйте "под заказ". Если непонятно когда — лучше скрыть и добавить кнопку "Уведомить о поступлении".

Больше информации о предотвращении проблем с остатками: как предотвратить дефицит товаров и потерю продаж.

Conclusion

Расхождения остатков между сайтом и поставщиком — решаемая проблема при правильном подходе. Комбинация технических решений (частая синхронизация, real-time API, webhooks) и бизнес-процессов (safety stock, резервирование, правила приоритетов) снижает вероятность проблем с 8-10% до менее 2%.

Key recommendations

• Увеличьте частоту синхронизации: популярные товары — каждые 15-30 минут, остальные — каждые 2-4 часа
• Внедрите safety stock: 10-30% от остатка в зависимости от скорости продаж и частоты обновлений
• Настройте резервирование: блокируйте товары на время оформления заказа (15-30 минут)
• Используйте real-time синхронизацию: webhooks или API polling каждые 5-15 минут для хитов
• Установите правила конфликтов: при противоречивых данных выбирайте меньшее значение остатка
• Настройте мониторинг: алерты при массовых обнулениях, задержках синхронизации, высоком проценте отмен
• Финальная проверка: перед отправкой заказа поставщику делайте real-time запрос наличия

Поэтапное внедрение

1. Аудит текущей ситуации (неделя 1): анализ частоты расхождений, процента отмен, причин проблем
2. Настройка базовой автоматизации (неделя 2-3): автоматическая загрузка прайсов, увеличение частоты синхронизации
3. Внедрение safety stock (неделя 4): расчет и применение страховых запасов по категориям
4. Real-time синхронизация (неделя 5-8): настройка webhooks или API polling для критичных товаров
5. Мониторинг и оптимизация (постоянно): отслеживание метрик, корректировка правил

ROI от внедрения

Средний интернет-магазин с выручкой $50,000/месяц экономит $1,200-1,500/месяц после внедрения комплексной системы предотвращения расхождений. Инвестиции $2,000-5,000 окупаются за 2-4 месяца.