Every list of the most scraped websites comes from somebody’s private logs, so it is hard to check. We wanted one you can rebuild yourself. On 30 October 2026 we pulled the public statistics of the most popular ready-made scrapers on the Apify Store: 8,186 scrapers that together logged 239.8 million runs in their last 30 days. We matched each one to the website it targets and ranked the sites. If you collect from any of them, the residential proxies setup notes at the end show what usually works for each type of target.
For the wider trends behind these numbers, see the state of web scraping in 2026. This post is the data.
Key takeaways
- Instagram and LinkedIn take 43.5% of site-specific demand. Instagram leads with 25.3%, LinkedIn follows with 18.2%.
- Four sites make up 60.0%: Instagram, LinkedIn, TikTok and Google Maps.
- Social media and video account for 56.0% of demand and 60.2% of runs.
- Meta properties together take 35.9%, Google properties 19.6%.
- AI answer engines barely register: 0.1% of demand.
- Most big targets run smoothly. Seven of the top eight sites have a run success rate of 98.5% or higher. The lowest rates sit with regional and travel targets such as Naver, Taobao, ImmobilienScout24 and Google Flights.
How we measured it
The Apify Store API publishes, for every public scraper (an “Actor”), its number of users in the last 30 days and the outcome of its public runs in the same window: succeeded, failed, timed out or aborted.
- Collect. We read the first 10,000 places of the Store’s own popularity ranking on 30 October 2026, which returned 8,186 unique scrapers out of 104,320 listed.
- Match to a target. We assigned each scraper to a single website using its name and title, for example “Instagram Profile Scraper” to Instagram. A scraper that names several platforms went to the first one our rules matched, so shares between the big social networks are approximate.
- Rank. We ranked websites by 30-day active users summed across all scrapers for that site, and report runs and run outcomes next to them.
Four caveats matter:
- This is self-serve demand, not total traffic. Teams running their own scrapers do not appear here.
- Users are counted per scraper. One person using three Instagram scrapers counts three times, so sites with many specialised scrapers look bigger.
- 13.8% of demand is not tied to one site. Generic crawlers, RAG browsers, contact finders and lead databases are left out of the site ranking.
- A failed run is not proof of blocking. Runs also fail on bad input, site changes and timeouts set by the user.
The 15 most scraped websites of 2026
Share is each site’s slice of 30-day active users across all scrapers that target a single site.
| Rank | Website | Share of site-specific demand | Runs in 30 days | Runs per user |
|---|---|---|---|---|
| 1 | 25.3% | 51.0M | 239 | |
| 2 | 18.2% | 37.2M | 244 | |
| 3 | TikTok | 8.7% | 33.9M | 464 |
| 4 | Google Maps | 7.8% | 17.4M | 266 |
| 5 | 6.5% | 18.1M | 330 | |
| 6 | YouTube | 6.5% | 8.4M | 153 |
| 7 | Google Search | 3.6% | 13.6M | 447 |
| 8 | X (Twitter) | 3.5% | 17.5M | 594 |
| 9 | Facebook Ads Library | 3.1% | 4.8M | 186 |
| 10 | 2.8% | 1.9M | 84 | |
| 11 | Amazon | 1.6% | 2.3M | 169 |
| 12 | Indeed | 1.5% | 2.9M | 233 |
| 13 | Douyin | 0.8% | 0.10M | 15 |
| 14 | Threads | 0.7% | 0.74M | 132 |
| 15 | Naukri | 0.6% | 0.10M | 18 |
Three patterns stand out.
Social platforms dominate. Profiles, posts, comments and videos from Instagram, TikTok, Facebook and YouTube are the raw material of social media monitoring, influencer research and brand tracking, and that work runs every day.
X is run far more often per user than anything else in the top 15. 594 runs per active user in 30 days, against 239 for Instagram. TikTok (464) and Google Search (447) follow. A high ratio points to scheduled, repeated collection rather than one-off pulls.
Google Maps at number four is a lead-generation story. Its 152 scrapers are built to collect business listings, reviews and contact details. We cover the method in how to scrape local business data.
Scraping demand by category
Social media and video take more than half of demand, and the top three categories together take 82.3%.
| Category | Share of site-specific demand | Share of runs | Biggest targets |
|---|---|---|---|
| Social media and video | 56.0% | 60.2% | Instagram, TikTok, Facebook, YouTube, X |
| B2B and professional data | 18.3% | 16.8% | LinkedIn, Crunchbase |
| Maps and local business | 8.0% | 8.6% | Google Maps, Yelp |
| Search engines and SEO data | 5.2% | 6.5% | Google Search, Google Trends, Similarweb |
| E-commerce and marketplaces | 3.9% | 1.9% | Amazon, AliExpress and Alibaba, Taobao and Tmall |
| Ad intelligence | 3.5% | 2.5% | Facebook Ads Library, Google Ads Transparency Center |
| Jobs and talent | 2.7% | 1.8% | Indeed, Naukri, Glassdoor |
| Travel | 0.9% | 0.8% | Airbnb, Tripadvisor, Booking.com |
| Real estate | 0.6% | 0.6% | Zillow, Idealista |
| Reviews | 0.5% | 0.1% | Trustpilot, G2 |
| App stores and developer | 0.3% | 0.1% | Google Play, Apple App Store |
| AI answer engines | 0.1% | 0.0% | ChatGPT, Perplexity, Gemini |
Jobs data is larger than its row suggests: LinkedIn job scrapers sit under LinkedIn, in the B2B row.
E-commerce looks small, and that is misleading. Price and catalogue monitoring at scale is usually built in-house or bought as a feed, so it rarely shows up as marketplace demand. Amazon still ranks eleventh on its own.
Websites with the lowest scraping success rates
Popular is not the same as easy. The table lists the targets with the lowest run success rates among those with at least 50,000 runs in 30 days.
| Website | Run success | Failed or timed out | Aborted | Runs in 30 days |
|---|---|---|---|---|
| Naver | 79.0% | 0.3% | 20.7% | 50,991 |
| Taobao and Tmall | 81.4% | 18.6% | 0.0% | 140,407 |
| Google Trends | 89.0% | 5.7% | 5.2% | 119,170 |
| ImmobilienScout24 | 89.9% | 10.1% | 0.0% | 132,649 |
| Google Flights and Hotels | 92.1% | 7.7% | 0.2% | 149,021 |
| Google Play | 94.7% | 1.6% | 3.8% | 69,753 |
| Trustpilot | 95.9% | 2.5% | 1.5% | 126,576 |
| 96.2% | 3.0% | 0.8% | 1,943,609 | |
| YouTube | 97.0% | 2.8% | 0.2% | 8,432,062 |
| Apple App Store | 97.1% | 0.5% | 2.4% | 54,734 |
The list has a clear shape:
- Regional platforms fail most. Taobao and Tmall, ImmobilienScout24 and Naver all serve one home market. Access from outside that market is a common source of trouble, though the data does not record the cause of each failure.
- Naver is an outlier of a different kind. Almost all of its unsuccessful runs were aborted, meaning stopped before they finished, not failed. The data does not say why.
- Travel pricing is hard. Google Flights and Hotels failed or timed out on 7.7% of runs. Our guide to scraping flight and hotel prices explains why location and session matter there.
- Scale does not mean difficulty. YouTube and Reddit sit on this list with millions of runs, yet most big social targets stay above 98.5%.
A successful run is not always correct data. A site can return an empty or stale page with a normal status code, so check the content, not only the outcome.
Other findings
AI answer engines barely register. Scrapers for ChatGPT, Perplexity and Gemini drew 1,042 active users and 82,543 runs, about 0.1% of demand. That says more about this marketplace than about AI search traffic in general.
Meta and Google dominate. Instagram, Facebook, the Facebook Ads Library, Threads and Facebook Marketplace together take 35.9% of site-specific demand. Google properties (Maps, Search, YouTube, Trends, the Ads Transparency Center, Flights and Hotels, News, Play, Jobs, Shopping and Scholar) take 19.6%.
Chinese platforms add up. Douyin, Taobao and Tmall, Xiaohongshu (RedNote), Bilibili and Weibo together take 1.8% of demand, more than Amazon’s 1.6%.
Proxy setup by target type
The dataset does not record which proxies each run used, so these are general notes, not findings. The right setup depends on how a site defends itself:
| Target type | Examples | What usually blocks you | Setup to consider |
|---|---|---|---|
| Social platforms | Instagram, TikTok, X, Reddit | Per-IP rate limits, login walls on repeated requests | Rotating residential IPs, low requests per IP, public pages only |
| Professional and jobs data | LinkedIn, Indeed, Naukri | Aggressive rate limits, sign-in prompts | Rotating residential IPs, slow pacing, logged-out listings |
| Maps and local | Google Maps, Yelp | Results change by city | Residential IPs with city targeting |
| Regional portals | Naver, Taobao, ImmobilienScout24 | Foreign traffic looks suspicious | Residential IPs in the target country, sticky sessions for multi-page flows |
| Travel and e-commerce | Google Flights, Amazon | JavaScript rendering, fingerprinting, IP reputation | Residential IPs plus a real browser, one sticky session per search |
A note on what you collect
A site being heavily scraped does not make every kind of collection from it lawful. Public, logged-out pages are a different case from content behind a login, and profiles on social and professional networks usually hold personal data, which brings privacy law into play. Read is web scraping legal before building on any of these targets, and ask a lawyer about your specific project.
The bottom line
In this 30-day snapshot, Instagram, LinkedIn, TikTok and Google Maps take 60% of site-specific scraping demand, and social and video platforms take more than half on their own. The large targets mostly run smoothly. The lowest success rates come from regional platforms and travel pricing, where location and session handling decide whether a run works.
Data note: figures come from public Apify Store statistics pulled by Shifter on 30 October 2026, covering 8,186 public scrapers and 239.8 million runs in each scraper’s previous 30 days. They describe activity in this dataset, not all web scraping on the internet. Run success rates describe run outcomes in the dataset and are not a definitive measure of how hard a website is to scrape.