A product card may show "in stock" in Moscow and "out of stock" in Novosibirsk — and if you do not check this by regions, you lose orders and do not see where competitors are winning in logistics. However, as soon as you start systematically checking stock from different geos with the same account or IP, the platform quickly notices the abnormal activity and restricts access. In this article, we present a working scheme for monitoring product availability by delivery regions without bans and captchas.
Why Product Availability Differs by Regions
Wildberries, Ozon, and most large marketplaces operate through a network of distributed warehouses. When a customer accesses the site from Krasnodar, the system shows stock from the nearest regional warehouses, not the seller's overall stock. This means that the same item may be "in stock" for Moscow and St. Petersburg but unavailable for the Urals or the Far East — simply because the product has not been physically delivered to the regional warehouse or has already been sold out locally.
For sellers, this is critical for several reasons. First, regional unavailability directly affects the position of the product card in search results — the algorithm lowers the ranking of products that cannot quickly reach the customer. Second, competitors who effectively distribute stock across warehouses capture orders in those regions where you have none. Third, without regular monitoring, you learn about the problem only after the fact — when sales in the region have already dropped, not when it is just beginning to happen.
The same applies to Avito: an ad may appear differently depending on the search city, and delivery and offer relevance may vary. Marketers and agencies managing multiple brands face this daily: it is essential to know what a real buyer sees in Yekaterinburg, not just what is visible from the office in Moscow.
Why Marketplaces Ban for Stock Monitoring
The logic of protection at Wildberries and Ozon is straightforward: an ordinary buyer accesses from one IP, views several product cards in their region, and leaves. If dozens of requests for different items with region switching come from one IP — this is a parser pattern, not a live person. Anti-bot systems respond in one of three ways: they show a captcha, temporarily block the IP due to a 429 error, or provide deliberately incorrect data (fake stock) to spoil the parser's statistics without revealing the fact of blocking.
The third scenario is the most dangerous for business — you continue to receive data, but it is incorrect, and decisions based on it will be erroneous. Marketplaces will not ban the seller's account itself if monitoring is done anonymously through a public catalog, but the IP address from which the load comes is quickly blocked — usually after 50–100 requests in a short period from one address without rotation.
An additional complication is geo-binding. To know the actual availability for a specific region, the request must come from an IP that the platform recognizes as being located in that region (or you explicitly switch the delivery region in the interface/API, but this is also tied to cookies and sessions, which can also be blocked with suspicious activity).
Monitoring Scheme Without Bans
A working scheme is built on three principles: load distribution, imitation of a live user, and regional accuracy of IP. In practice, it looks like this:
- List of Regions and Items. Create a table: needed regions (usually 15–30 key cities to start) × items that need to be tracked.
- Region-Bound Proxies. A separate IP address (or pool of addresses) is used for each region, which physically or logically corresponds to the desired city/area.
- Rotation and Delays. A pause of 3–10 seconds is maintained between requests, and the IP changes no more frequently than once every few requests to avoid appearing like a scanner.
- Browser Imitation. Requests are sent with a realistic User-Agent, headers, and, if necessary, through a headless browser with JS rendering — many stock cards load dynamically.
- Logging and Alerts. Results are compiled into a table or database, and when the status changes from "out of stock" to "in stock" and vice versa, a notification is sent to Telegram or email.
This approach is equally applicable to manual monitoring through ready-made parsing services and to custom scripts — only the tool differs; the principle of protection against bans remains unchanged.
What Proxies Are Needed for Regional Checks
The choice of proxy type is the main factor that determines whether monitoring will be stable. For the task of checking product availability by regions, ordinary cheap data center IPs without geo-binding are not suitable — the platform will either not provide an accurate regional response or quickly block the address if it has already been flagged in blacklists.
| Proxy Type | Geo Accuracy | Risk of Blocking | When to Use |
|---|---|---|---|
| Residential Proxies | High, down to the city level | Low | Primary tool for regular monitoring across 15+ regions |
| Mobile Proxies | Medium-High (by operator network) | Very Low | If the platform aggressively bans residential ranges |
| Data Center Proxies | Low, without precise city binding | High with frequent requests | Fast one-time checks without regional accuracy |
For systematic monitoring of product availability across dozens of regions, the optimal choice is residential proxies with the ability to select a city or region: they appear as IPs of ordinary internet users, so the marketplace provides accurate local availability data and does not consider requests suspicious. If the budget is limited and there are few regions, a combination can be used: residential IPs for key cities and data center ones for less critical checks where accuracy is less important.
Setting Up Geo-Targeting and IP Rotation
After selecting a proxy provider, you need to set up two parameters: geo-targeting and rotation. Most residential proxy services allow you to specify the country, region, and even city upon connection — for example, through a login format with a region parameter or through a separate port/endpoint for each city. This eliminates the need to manually search for "clean" IPs for each region.
Rotation should be configured according to two scenarios:
- Sticky session for the duration of one region check — the IP does not change while you gather data for all items for one city, to avoid disrupting the session and triggering a repeat captcha.
- Complete IP change between regions — when moving to the next city, a new address tied to the desired geo is taken; otherwise, the platform may not provide accurate local stock.
If you are using an anti-detect browser (Dolphin Anty, AdsPower, Multilogin, Octo Browser), each profile is assigned a separate proxy with the required geo, and the browser fingerprint remains unique to the profile. This provides double protection: the platform sees both a "live" IP of the region and a unique digital fingerprint of the device, which excludes linking multiple profiles into one cluster for banning.
Tools for Monitoring
For sellers who do not engage in programming, there are three ways to set up monitoring:
1. Ready-made monitoring services for prices and stock. Many such services for Wildberries and Ozon already support regional checks "out of the box" and allow you to connect your own proxies in the settings — this is the fastest option to start without technical skills.
2. Anti-detect browser + manual or semi-automated checks. For 10–20 key regions, you can create profiles in Dolphin Anty or AdsPower, link each profile to the proxy of the desired city, and methodically browse product cards, recording the availability status. This is slower but requires no coding at all and is suitable if checks are needed not every hour but once a day.
3. Automated parser (for those with a developer or freelancer). A Python script using the requests or Playwright library navigates product cards according to a specified list of regions and items, using proxies from the pool, and compiles the results into a table or database. This option provides maximum flexibility and speed of data updates but requires technical support.
Regardless of the chosen tool, the bottleneck is the same — the quality and geography of proxies. Even the most well-thought-out script will collect incorrect data or receive bans if the IP addresses are not tied to the necessary regions or are repeated too frequently.
Request Limits and Protection Against Bans
To ensure monitoring works stably for months, not just a couple of days before mass blocking, adhere to practical limits:
| Parameter | Safe Range |
|---|---|
| Requests from one IP per minute | 3–8 |
| Pause between requests | 3–10 seconds with randomization |
| IP change | after 20–30 requests or with each new region |
| Frequency of full checks across all regions | 1–4 times a day for most product categories |
If the platform shows a captcha even when limits are followed, it is a signal that the specific IP has already been blacklisted — immediate rotation to a new address and a temporary reduction in request frequency for that region for 1–2 hours helps. For highly competitive categories (clothing, electronics), marketplaces monitor bot traffic more strictly, so limits should be made even more conservative there.
Step-by-Step Setup on Wildberries, Ozon, and Avito
Step 1. Define the list of regions. Take the top 20 cities by sales volume from your seller account analytics — this is where product availability is most critical.
Step 2. Connect proxies with geo-targeting. In the proxy management panel, select the country "Russia" and the necessary cities/regions, and obtain connection data (host, port, username, password).
Step 3. Create profiles in an anti-detect browser. For each region — a separate profile in Dolphin Anty, AdsPower, or GoLogin linked to the proxy of that city. Name the profiles by region to avoid confusion.
Step 4. Set the delivery region on the marketplace site. On Wildberries and Ozon, the region is automatically determined by IP, but it is worth checking manually by selecting the city in the header of the site — this will confirm that the platform correctly recognized the proxy location.
Step 5. Record the availability status for each item. Open the product cards of interest and note the status ("in stock", "out of stock", delivery time) in a table with the date and time of the check.
Step 6. Repeat on a schedule. For Avito, the logic is similar — change the search city through the proxy of the corresponding region, after which it is recorded which ads and with what delivery status are visible to the local buyer.
Step 7. Set up alerts. As soon as the product status in an important region changes to "out of stock", send a notification to the responsible manager so they can promptly redistribute stock between warehouses.
Monitoring Launch Checklist
- A list of priority regions and items for tracking has been compiled
- Proxies with precise binding to the necessary cities have been connected
- IP rotation and pauses between requests (3–10 seconds) have been set up
- A separate profile has been created in the anti-detect browser for each region
- It has been verified that the platform correctly determines the region by the proxy IP
- Logging of results with the date and time of the check has been set up
- Alerts have been enabled for changes in product availability status
- A request limit has been established to prevent IP blocking
Conclusion
Monitoring product availability by delivery regions is not a one-time check but a continuous process that directly affects product card positions and sales on Wildberries, Ozon, and Avito. To ensure it works stably, three things are needed: precise binding of IP to the region, reasonable request limits, and load distribution among proxies to avoid raising suspicions with the platforms' anti-bot systems.
If you plan to monitor product availability across dozens of regions and do so regularly, it is advisable to start with residential proxies with geo-targeting by cities — they provide accurate data on local stock and minimize the risk of bans. For one-time quick checks without strict regional binding, data center proxies can suffice, and if the platform proves particularly strict with residential ranges, mobile proxies come to the rescue with minimal ban risk.