How to Scrape Websites Without Getting Blocked

Web scraping is a task that has to be performed responsibly so that it does not have a detrimental effect on the sites being scraped. Web Crawlers can retrieve data much quicker, in greater depth than humans, so bad scraping practices can have some impact on the performance of the site. While most websites may not have anti-scraping mechanisms, some sites use measures that can lead to web scraping getting blocked, because they do not believe in open data access.

If a crawler performs multiple requests per second and downloads large files, an under-powered server would have a hard time keeping up with requests from multiple crawlers. Since web crawlers, scrapers or spiders (words used interchangeably) don’t really drive human website traffic and seemingly affect the performance of the site, some site administrators do not like spiders and try to block their access.

In this article, we will talk about the best web scraping practices to follow to scrape websites without getting blocked by the anti-scraping or bot detection tools.

Web Scraping best practices to follow to scrape without getting blocked

1. Respect Robots.txt
2. Make the crawling slower, do not slam the server, treat websites nicely
3. Do not follow the same crawling pattern
4. Make requests through Proxies and rotate them as needed
5. Rotate User Agents and corresponding HTTP Request Headers between requests
6. Use a headless browser like Puppeteer, Selenium or Playwright
7. Beware of Honey Pot Traps
8. Check if Website is Changing Layouts
9. Avoid scraping data behind a login
10. Use Captcha Solving Services
11. How can websites detect web scraping?
12. How do you find out if a website has blocked or banned you ?

Respect Robots.txt

Web spiders should ideally follow the robot.txt file for a website while scraping. It has specific rules for good behavior, such as how frequently you can scrape, which pages allow scraping, and which ones you can’t. Some websites allow Google to scrape their websites, by not allowing any other websites to scrape. This goes against the open nature of the Internet and may not seem fair, but the owners of the website are within their rights to resort to such behavior. 

You can find the robot.txt file on websites. It is usually the root directory of a website – http://example.com/robots.txt.

If it contains lines like the ones shown below, it means the site doesn’t like and does not want to be scraped.

User-agent: *

Disallow:/ 

However, since most sites want to be on Google, arguably the largest scraper of websites globally, they allow access to bots and spiders. 

What if you need some data that is forbidden by Robots.txt. You could still scrape it. Most anti-scraping tools block web scraping when you are scraping pages that are not allowed by Robots.txt.

What do these tools look for? Is this client a bot or a real user? And how do they find that? By looking for a few indicators that real users do and bots don’t. Humans are random, bots are not. Humans are not predictable, bots are.

Here are a few easy giveaways that you are bot/scraper/crawler –

• Scraping too fast and too many pages, faster than a human ever can
• Following the same pattern while crawling. For example – go through all pages of search results, and go to each result only after grabbing links to them. No human ever does that.
• Too many requests from the same IP address in a very short time
• Not identifying as a popular browser. You can do this by specifying a ‘User-Agent’.
• using a user agent string of a very old browser

The points below should get you past most of the basic to intermediate anti-scraping mechanisms used by websites to block web scraping.

Make the crawling slower, do not slam the server, treat websites nicely

Web scraping bots fetch data very fast, but it is easy for a site to detect your scraper, as humans cannot browse that fast. The faster you crawl, the worse it is for everyone. If a website gets too many requests than it can handle it might become unresponsive.

Make your spider look real, by mimicking human actions. Put some random programmatic sleep calls in between requests, add some delays after crawling a small number of pages and choose the lowest number of concurrent requests possible. Ideally, put a delay of 10-20 seconds between clicks and not put much load on the website, treating the website nice.

Use auto throttling mechanisms which will automatically throttle the crawling speed based on the load on both the spider and the website that you are crawling. Adjust the spider to an optimum crawling speed after a few trial runs. Do this periodically because the environment does change over time.

Do not follow the same crawling pattern

Humans generally will not perform repetitive tasks as they browse through a site with random actions. Web scraping bots tend to have the same crawling pattern because they are programmed that way unless specified. Sites that have intelligent anti-crawling mechanisms can easily detect spiders by finding patterns in their actions and can lead to web scraping getting blocked.

Incorporate some random clicks on the page, mouse movements and random actions that will make a spider look like a human.

Make requests through Proxies and rotate them as needed

When scraping, your IP address can be seen. A site will know what you are doing and if you are collecting data. They could take data such as – user patterns or experience if they are first-time users.

Multiple requests coming from the same IP will lead you to get blocked, which is why we need to use multiple addresses. When we send requests from a proxy machine, the target website will not know where the original IP is from, making the detection harder.

Create a pool of IPs that you can use and use random ones for each request. Along with this, you have to spread a handful of requests across multiple IPs.

There are several methods that can change your outgoing IP.

• TOR
• VPNs
• Free Proxies
• Shared Proxies – the least expensive proxies shared by many users. The chances of getting blocked are high.
• Private Proxies – usually used only by you, and lower chances of getting blocked if you keep the frequency low.
• Data Center Proxies, if you need a large number of IP Addresses and faster proxies, larger pools of IPs. They are cheaper than residential proxies and could be detected easily.
• Residential Proxies, if you are making a huge number of requests to websites that block to actively. These are very expensive (and could be slower, as they are real devices). Try everything else before getting a residential proxy.

In addition, various commercial providers also provide services for automatic IP rotation. A lot of companies now provide residential IPs to make scraping even easier – but most are expensive.

Rotate User Agents and corresponding HTTP Request Headers between requests

A user agent is a tool that tells the server which web browser is being used. If the user agent is not set, websites won’t let you view content. Every request made from a web browser contains a user-agent header and using the same user-agent consistently leads to the detection of a bot. You can get your User-Agent by typing ‘what is my user agent’ in Google’s search bar. The only way to make your User-Agent appear more real and bypass detection is to fake the user agent. Most web scrapers do not have a User Agent by default, and you need to add that yourself.

You could even pretend to be the Google Bot: Googlebot/2.1 if you want to have some fun! (http://www.google.com/bot.html)

Now, just sending User-Agents alone would get you past most basic bot detection scripts and tools. If you find your bots getting blocked even after putting in a recent User-Agent string, you should add some more request headers.

Most browsers send more headers to the websites than just the User-Agent. For example, here is a set of headers a browser sent to Scrapeme.live (Our Web Scraping Test Site). It would be ideal to send these common request headers too.

The most basic ones are:

• User-Agent
• Accept
• Accept-Language
• Referer
• DNT
• Updgrade-Insecure-Requests
• Cache-Control

Do not send cookies unless your scraper depends on Cookies for functionality.

You can find the right values for these by inspecting your web traffic using Chrome Developer Tools, or a tool like MitmProxy or Wireshark. You can also copy a curl command to your request from them. For example

curl 'https://scrapeme.live/shop/Ivysaur/' \
  -H 'authority: scrapeme.live' \
  -H 'dnt: 1' \
  -H 'upgrade-insecure-requests: 1' \
  -H 'user-agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_4) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/83.0.4103.61 Safari/537.36' \
  -H 'accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9' \
  -H 'sec-fetch-site: none' \
  -H 'sec-fetch-mode: navigate' \
  -H 'sec-fetch-user: ?1' \
  -H 'sec-fetch-dest: document' \
  -H 'accept-language: en-GB,en-US;q=0.9,en;q=0.8' \
  --compressed

You can get this converted to any language using a tool like https://curl.trillworks.com

Here is how this was converted to python

import requests

headers = {
    'authority': 'scrapeme.live',
    'dnt': '1',
    'upgrade-insecure-requests': '1',
    'user-agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_4) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/83.0.4103.61 Safari/537.36',
    'accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9',
    'sec-fetch-site': 'none',
    'sec-fetch-mode': 'navigate',
    'sec-fetch-user': '?1',
    'sec-fetch-dest': 'document',
    'accept-language': 'en-GB,en-US;q=0.9,en;q=0.8',
}

response = requests.get('https://scrapeme.live/shop/Ivysaur/', headers=headers)

You can create similar header combinations for multiple browsers and start rotating those headers between each request to reduce the chances of getting your web scraping blocked.

Use a headless browser like Puppeteer, Selenium or Playwright

If none of the methods above works, the website must be checking if you are a REAL browser.

The simplest check is if the client (web browser) can render a block of JavaScript. If it doesn’t, then it pretty much flags the visitor to be a bot. While it is possible to block running JavaScript in the browser, most of the Internet sites will be unusable in such a scenario and as a result, most browsers will have JavaScript enabled.

Once this happens, a real browser is necessary in most cases to scrape the data. There are libraries to automatically control browsers such as

1. Selenium
2. Puppeteer and Pyppeteer
3. Playwright

Anti Scraping tools are smart and are getting smarter daily, as bots feed a lot of data to their AIs to detect them. Most advanced Bot Mitigation Services use Browser Side Fingerprinting (Client Side Bot Detection)  by more advanced methods than just checking if you can execute Javascript.

Bot detection tools look for any flags that can tell them that the browser is being controlled through an automation library.

1. Presence of bot specific signatures
2. Support for nonstandard browser features
3. Presence of common automation tools such as Selenium, Puppeteer, Playwright, etc.
4. Human-generated events such as randomized Mouse Movement, Clicks, Scrolls, Tab Changes etc.

All this information is combined to construct a unique client-side fingerprint that can tag one as bot or human.

Here are a few workarounds or tools which could help your headless browser-based scrapers from getting banned.

Beware of Honey Pot Traps

Honeypots are systems set up to lure hackers and detect any hacking attempts that try to gain information. It is usually an application that imitates the behavior of a real system. Some websites install honeypots, which are links invisible to normal users but can be seen by web scrapers.

When following links always take care that the link has proper visibility with no nofollow tag. Some honeypot links to detect spiders will have the CSS style display:none or will be color disguised to blend in with the page’s background color.

This detection is obviously not easy and requires a significant amount of programming work to accomplish properly, as a result, this technique is not widely used on either side – the server side or the bot or scraper side.

Check if Website is Changing Layouts

Some websites make it tricky for scrapers, serving slightly different layouts.

For example, in a website pages 1-20 will display a layout, and the rest of the pages may display something else. To prevent this, check if you are getting data scraped using XPaths or CSS selectors. If not, check how the layout is different and add a condition in your code to scrape those pages differently.

Avoid scraping data behind a login

Login is basically permission to get access to web pages. Some websites like Indeed do not allow permission.

If a page is protected by login, the scraper would have to send some information or cookies along with each request to view the page. This makes it easy for the target website to see requests coming from the same address. They could take away your credentials or block your account which can, in turn, lead to your web scraping efforts being blocked.

Its generally preferred to avoid scraping websites that have a login as you will get blocked easily, but one thing you can do is imitate human browsers whenever authentication is required you get the target data you need.

Use Captcha Solving Services

Many websites use anti web scraping measures. If you are scraping a website on a large scale, the website will eventually block you. You will start seeing captcha pages instead of web pages. There are services to get past these restrictions such as 2Captcha or Anticaptcha.

If you need to scrape websites that use Captcha, it is better to resort to captcha services. Captcha services are relatively cheap, which is useful when performing large scale scrapes.

How can websites detect and block web scraping?

Websites can use different mechanisms to detect a scraper/spider from a normal user. Some of these methods are enumerated below:

1. Unusual traffic/high download rate especially from a single client/or IP address within a short time span.
2. Repetitive tasks performed on the website in the same browsing pattern – based on an assumption that a human user won’t perform the same repetitive tasks all the time.
3. Checking if you are real browser – A simple check is to try and execute javascript. Smarter tools can go a lot more and check your Graphic cards and CPUs 😉 to make sure you are coming from real browser.
4. Detection through honeypots – these honeypots are usually links which aren’t visible to a normal user but only to a spider. When a scraper/spider tries to access the link, the alarms are tripped.

How to address this detection and avoid web scraping getting blocked? 

Spend some time upfront and investigate the anti-scraping mechanisms used by a site and build the spider accordingly. It will provide a better outcome in the long run and increase the longevity and robustness of your work.

How do you find out if a website has blocked or banned you?

If any of the following signs appear on the site that you are crawling, it is usually a sign of being blocked or banned.

• CAPTCHA pages
• Unusual content delivery delays
• Frequent response with HTTP 404, 301 or 50x errors

Frequent appearance of these HTTP status codes is also indication of blocking

• 301 Moved Temporarily
• 401 Unauthorized
• 403 Forbidden
• 404 Not Found
• 408 Request Timeout
• 429 Too Many Requests
• 503 Service Unavailable

Here is what Amazon.com tells you when you are blocked.

You may also see response or message from websites like these from some popular anti-scraping tools.

or

or

or

A comprehensive list of HTTP return codes (successes and failures) can be found here. It will be worth your time to read through these codes and be familiar with them.

Summary

All these ideas above provide a starting point for you to build your own solutions or refine your existing solution. If you have any ideas or suggestions, please join the discussion in the comments section.

Thank you for reading.