Google has reportedly removed much of Twitter’s links from its search results after the social network’s owner Elon Musk announced reading tweets would be limited.
Search Engine Roundtable found that Google had removed 52% of Twitter links since the crackdown began last week. Twitter now blocks users who are not logged in and sets limits on reading tweets.
According to Barry Schwartz, Google reported 471 million Twitter URLs as of Friday. But by Monday morning, that number had plummeted to 227 million.
“For normal indexing of these Twitter URLs, it seems like these tweets are dropping out of the sky,” Schwartz wrote.
Platformer reported last month that Twitter refused to pay its bill for Google Cloud services.
Doesn’t sound like retaliation to me, it sounds like their scheduled web crawlers are finding that content they used to index is now no longer viewable and this removed from search results. Pretty standard. My guess is that there were 400 million URLs listed and as the crawler uncovers that they are no longer available, that number will keep dropping to reflect only content publicly viewable. If only 500 URLs are now publicly viewable (without logins) then that’s what they will index. Google isn’t a search engine for private companies (unless you pay for the service) they are a public search engine so they make an effort to ensure that only public information is indexed. Some folk game the system (like the old expertsexchange.com) but sooner or later google drops the hammer.
God, I hated expert sexchange so much. It was a blessing when stack overflow started.
I’m only an amateur but I’m happy to give it a go
I don’t think Twitter would rate limit the Google indexer, though.
It’s probably the increased bounce rate, as people click Twitter links in the search results, get Twitter’s login wall and click back to continue searching instead of creating an account.
I tried to access twitter by impersonating a googlebot. I was denied. The bots aren’t so much rate limited, but unable to access tweets as they don’t have a Twitter account.
Impersonating a Googlebot user-agent isn’t a perfect test though, since you’re supposed to implement IP verification too - either checking against a known list of Google crawler IPs or doing a PTR lookup. See https://developers.google.com/search/docs/crawling-indexing/verifying-googlebot
Question is whether or not Twitter has competent software developers implementing stuff correctly anymore. Anyone’s guess at this point.
Google also obfuscates their crawler. Or rather, they request a copy of the page with the Googlebot user agent, and then from a different IP and a random user agent. If the content between them is different, that raises red flags.