Blocks when scraping marketplaces, automating social media, or monitoring competitors are a common problem faced by sellers, SMM specialists, and arbitrageurs. In 90% of cases, the reason is too high a request frequency from a single IP address. Websites and platforms track suspicious activity and block accounts or IPs if they detect inhuman behavior.
In this article, we will discuss how to properly configure request frequency through proxies so that your tools operate stably: marketplace scrapers collect data without blocks, anti-detect browsers manage dozens of accounts without bans, and SMM automation does not lead to the loss of client profiles.
Why Platforms Block for High Request Frequency
Modern platforms — Wildberries, Ozon, Instagram, Facebook, Avito — use complex bot protection systems. One of the main signals of suspicious activity is the request frequency from a single IP address.
An ordinary user cannot physically view 100 products per minute or log into 50 Instagram accounts in an hour. If the system detects such activity, it concludes: this is a bot or automation. The result is an IP block, a temporary account ban, or a captcha at every step.
Main Reasons for Blocks:
- Too many requests in a short period — for example, 1000 requests to the marketplace API per minute
- Lack of natural pauses — a robot sends requests with perfect periodicity (every 2 seconds), which a human cannot do
- Identical action patterns — the same sequence of clicks, transitions, requests
- Using one IP for multiple accounts — especially critical for social networks and advertising accounts
- Exceeding official API limits — many platforms publish rate limits in their documentation
Important: Even if you use proxies, too high a request frequency can lead to the blocking of the proxy server itself. Proxy providers may also limit the number of requests to protect their infrastructure.
Safe Request Limits for Popular Platforms
Each platform has its limits on the number of requests. Some publish them officially in API documentation, while others keep them secret, and safe values are determined through experience. Here are the verified limits for popular services used by our clients.
| Platform | Safe Limit | Recommended Delay | Task Type |
|---|---|---|---|
| Wildberries | 60-120 requests/minute | 3-5 seconds | Scraping products, prices |
| Ozon | 30-60 requests/minute | 5-8 seconds | Stock monitoring, analytics |
| Avito | 20-40 requests/minute | 4-7 seconds | Posting ads |
| 200 actions/hour | 15-30 seconds | Likes, follows, comments | |
| Facebook Ads | 200 requests/hour (API) | 20-40 seconds | Campaign management |
| TikTok | 150-200 actions/hour | 20-35 seconds | Views, likes, comments |
| Google (search) | 10-15 requests/minute | 5-10 seconds | Scraping search results, positions |
| Yandex.Market | 40-80 requests/minute | 3-6 seconds | Data collection on products |
These values are the result of practical experience from hundreds of users. They provide a balance between speed and safety. If you are working through residential proxies, you can slightly increase the frequency, as such IPs appear more natural to platforms.
Proxy Rotation: How to Distribute Load Between IPs
Proxy rotation is the automatic change of IP addresses during operation. Instead of sending all requests from one IP (and quickly getting blocked), you distribute the load among dozens or hundreds of addresses.
Two main types of rotation:
1. Time-based Rotation
The IP address changes after a set time interval — for example, every 5 minutes, 10 minutes, or an hour. Suitable for tasks where you need to maintain a session for some time.
When to use:
- Scraping marketplaces (Wildberries, Ozon) — one IP collects data by category, then changes
- Competitor monitoring — tracking prices and stock
- Working in anti-detect browsers — each profile uses the IP for several minutes
2. Request-based Rotation
The IP address changes with each new request or after a certain number of requests (for example, every 10 requests). Maximum anonymity, but not suitable for tasks requiring authorization.
When to use:
- Scraping Google, Yandex search results
- Collecting public data without authorization
- Checking website availability from different regions
Practical Load Distribution Formula:
Number of IPs = (Total number of requests per hour) / (Safe limit per IP per hour) Example for Wildberries: - Need to collect 10,000 products per hour - Safe limit: 100 requests/minute = 6000 requests/hour per IP - Needed IPs: 10,000 / 6000 = 2 IPs (round up to 3 for safety) Example for Instagram (mass following): - Need to follow 1000 accounts per hour - Safe limit: 200 actions/hour per IP - Needed IPs: 1000 / 200 = 5 IPs
If you are working with social networks or advertising accounts through anti-detect browsers (Dolphin Anty, AdsPower, Multilogin), each profile needs a separate static IP. In this case, rotation is not used — each account is permanently tied to its proxy.
Setting Delays Between Requests: Imitating Human Behavior
Even if you adhere to the limits on the number of requests, robot-like behavior can reveal automation. A human cannot click with perfect periodicity — sometimes they think, get distracted, or read product descriptions. Delays should be random and natural.
Types of Delays for Different Scenarios:
| Scenario | Minimum Delay | Maximum Delay | Recommendation |
|---|---|---|---|
| Scraping marketplaces | 3 seconds | 8 seconds | Random delay 3-8 sec |
| Actions in Instagram | 15 seconds | 45 seconds | Different delay for likes/follows |
| Posting ads on Avito | 30 seconds | 90 seconds | Imitating form filling |
| Scraping search results | 5 seconds | 12 seconds | Google is more sensitive |
| Working with Facebook Ads | 20 seconds | 60 seconds | Creating/editing campaigns |
How to Properly Set Random Delays:
- Use a range instead of a fixed value — instead of "delay 5 seconds," set "random delay from 3 to 8 seconds." Most scraping and automation tools support this.
- Add long pauses — every 20-30 requests, take a break of 2-5 minutes, as if the user got distracted by a call or coffee.
- Vary delays depending on the action — viewing a product may take 3 seconds, while filling out an order form may take 30-60 seconds.
- Consider the time of day — at night, you can speed up the work a bit (fewer real users), during the day — slow down.
Advice for SMM: If you manage client accounts through Dolphin Anty or AdsPower, set different delays and work patterns in each profile. Even if all accounts belong to one agency, they should behave like different people.
Which Types of Proxies to Use for Different Tasks
The choice of proxy type directly affects the allowable request frequency and the risk of blocks. Different platforms treat data center IPs, residential, and mobile addresses differently.
| Proxy Type | Best for | Allowable Frequency | Risk of Blocking |
|---|---|---|---|
| Data Centers | Scraping marketplaces, price monitoring, SEO tools | High (up to 200 requests/min) | Medium (rotation needed) |
| Residential | Social networks, advertising accounts, anti-detect browsers | Medium (30-100 requests/min) | Low (appear as real users) |
| Mobile | Instagram, TikTok, Facebook Ads, multi-accounting | Low (20-50 requests/min) | Minimal (mobile IPs are trusted more) |
Recommendations for Selection:
- For scraping Wildberries, Ozon, Yandex.Market — data center proxies with rotation every 5-10 minutes will do. They are fast and cheap, and marketplaces do not strictly check the type of IP when publicly accessible.
- For working with Instagram, TikTok through anti-detect browsers — only mobile proxies. These platforms see that the request comes from a mobile device (by User-Agent) and expect a mobile IP.
- For Facebook Ads, Google Ads — residential proxies tied to a specific city. Advertising accounts track geolocation and may block frequent region changes.
- For mass tasks without authorization (scraping search results, availability checks) — data centers with aggressive rotation (every 1-5 requests).
Tools for Automatic Request Optimization
Manually controlling request frequency and IP rotation is challenging, especially if you are working with multiple platforms simultaneously. Modern tools automate this process and adapt behavior to each platform.
Anti-Detect Browsers with Built-in Proxy Management
If you are working with social networks or advertising accounts, anti-detect browsers are your primary tool. They not only mask your digital footprint but also help manage proxies for each profile.
- Dolphin Anty — allows you to tie a separate proxy to each profile, automatically checks the IP before launching, and shows usage statistics. Convenient for arbitrageurs working with Facebook Ads and TikTok Ads.
- AdsPower — supports bulk proxy import, automatic functionality checks, and synchronization of settings between profiles. Popular among SMM agencies for managing dozens of Instagram accounts.
- Multilogin — a professional solution with advanced proxy management: automatic rotation by schedule, checking IP for DNS/WebRTC leaks, detailed statistics. Used by large e-commerce teams.
- GoLogin — a budget option with basic proxy management functionality. Suitable for beginner arbitrageurs and small SMM projects.
Ready-made Scrapers with Request Frequency Control
For scraping marketplaces and monitoring prices, there are specialized services that have already set safe limits.
- Scrapers for Wildberries/Ozon — most services (e.g., Mpstats, Moneyplace) use their own proxy pools and automatically distribute the load. You do not need to manually set the request frequency.
- SEO Tools (Screaming Frog, Netpeak Spider) — allow you to set delays between requests, the number of threads, and use proxies. It is important not to set too high a speed when scraping competitors.
- Tools for Avito — when posting ads through automation, always use delays of 30-90 seconds between publications and different proxies for each account.
Setting Limits in Popular Tools
Example of Setting in Dolphin Anty:
- Open profile settings → "Proxy" tab
- Insert proxy data in the format:
IP:PORT:LOGIN:PASSWORD - Select type: HTTP/SOCKS5 (SOCKS5 is better for social networks)
- Click "Check Proxy" — ensure the IP corresponds to the required region
- In the "Automation" section, set delays between actions: 15-30 seconds for Instagram, 20-40 for Facebook
Example of Setting a Scraper for Wildberries:
- Upload a list of proxies (at least 3-5 IPs for stable operation)
- Set rotation: every 5 minutes or every 100 requests
- Set delay between requests: 3-5 seconds (random)
- Limit the number of threads: no more than 3-5 simultaneous requests
- Enable browser emulation (User-Agent, headers)
Monitoring and Tracking Blocks
Even with the correct setup of request frequency and proxy rotation, it is important to monitor whether blocks have started. Early detection of the problem allows you to adjust settings before mass bans occur.
Signs that the request frequency is too high:
- Captcha appearance — the first signal that the platform suspects automation
- Errors 429 (Too Many Requests) — a direct indication of exceeding the API limit
- Errors 403 (Forbidden) — IP or account temporarily blocked
- Empty responses or redirects — instead of data, the server returns placeholders
- Increased response time — the server artificially slows down responses from suspicious IPs
- Request for re-authorization — in social networks and advertising accounts
How to Organize Monitoring:
- Log all requests — save the server response code (200, 403, 429), response time, and used IP. This will help identify the blocking pattern.
- Automatic notifications — set up alerts in Telegram or email when errors 429/403 or captchas appear. Many scrapers and anti-detect browsers support this.
- Check proxies before use — ensure that the IP is not on blacklists (can be checked through services like IPQualityScore, Scamalytics).
- Rotate at the first signs of problems — if one IP starts getting captchas, replace it immediately without waiting for a complete block.
- Testing on a small volume — before launching large-scale scraping, test the settings on 100-200 requests, assess the platform's reaction.
Important for Arbitrageurs: If you are farming Facebook Ads or TikTok Ads accounts, monitor the "health" of proxies through the built-in tools of Dolphin Anty or AdsPower. They show blocking statistics for each IP and warn of problems.
Practical Cases: Scraping, SMM, Arbitrage
Theory without practice provides little benefit. Let's examine real scenarios of using proxies with optimized request frequency for various business tasks.
Case 1: Scraping 50,000 Products from Wildberries in One Day
Task: A seller wants to collect prices, stock, and ratings of competitors for 50,000 products in the "Electronics" category.
Solution:
- Used 10 data center proxies with rotation every 10 minutes
- Set delay between requests: 3-6 seconds (random)
- Limited the number of threads: 5 simultaneous requests
- Added a long pause of 3 minutes every 30 minutes of operation
Result: Collected all 50,000 products in 24 hours without a single block. Average speed: ~35 products per minute (2100 per hour). No IP received a captcha or ban.
Case 2: Managing 30 Instagram Accounts for an SMM Agency
Task: An SMM agency manages 30 client accounts on Instagram, needing to post, respond to comments, and like without blocks.
Solution:
- Assigned a separate mobile proxy to each account (30 IPs)
- Configured profiles in AdsPower with different digital fingerprints
- Set limits: no more than 150 actions per hour per account
- Delays between actions: 20-40 seconds (likes), 40-80 seconds (comments)
- Worked during "human" hours: 9:00 AM - 10:00 PM, accounts inactive at night
Result: In 6 months of operation, no account received a ban. Average subscriber growth: 500-800 per month per account. Instagram did not suspect automation due to mobile IPs and natural delays.
Case 3: Farming 20 Facebook Ads Accounts
Task: An arbitrageur is farming 20 Facebook Ads accounts for launching advertising campaigns. They need to warm up the accounts without chain bans.
Solution:
- Assigned a residential proxy with an IP from the USA to each account (city matches BM)
- Used Dolphin Anty to create isolated profiles
- Warmed up accounts: 2-3 logins per day for 10-15 minutes
- Actions: scrolling the feed, liking friends' posts, occasional comments
- Delays between actions: 30-60 seconds, imitating reading posts
- Launched ads: no earlier than 14 days after warming up
Result: Out of 20 accounts, 18 successfully warmed up and launched ads. 2 accounts were subject to document verification (not related to proxies). Thanks to the low request frequency and residential IPs, Facebook did not suspect multi-accounting.
Case 4: 24/7 Competitor Price Monitoring on Ozon
Task: An e-commerce business tracks prices of 5000 competitor products on Ozon every hour to promptly adjust their prices.
Solution:
- Used 15 residential proxies with rotation every 15 minutes
- Divided 5000 products into batches of 300-400 products per IP
- Delay between requests: 5-8 seconds
- Full cycle for checking all products: 45-50 minutes
- Automatic notification in Telegram when a competitor's price changes by more than 5%
Result: The system has been operating stably for 4 months without blocks. Thanks to prompt monitoring, the company manages to react to changes in competitor prices within an hour, increasing conversion by 12%.
Conclusion
Optimizing request frequency through proxies is not just a technical setup, but a key factor for success in scraping, SMM automation, and traffic arbitrage. The right balance between speed and safety helps avoid blocks, maintain accounts, and consistently obtain the necessary data.
The main principles to remember: adhere to safe limits for each platform, use random delays between requests, choose the right type of proxy for the task, and always monitor the first signs of blocks. Modern tools — anti-detect browsers, ready-made scrapers, rotation systems — automate much of the work, but understanding the basic principles remains critically important.
If you plan to work with marketplaces, social networks, or advertising platforms, choosing the right type of proxy will be the foundation of your strategy. For tasks with high anonymity requirements and minimal risk of blocks, we recommend trying residential proxies — they provide natural behavior and a high level of trust from platforms. For working with Instagram and TikTok, the optimal solution will be mobile proxies, which fully imitate real users of mobile networks.