Google does a much better job in communicating with the search community these days. Whether it's announcing the latest updates, posting extensive documentation, or answering questions asked directly to team members, Google usually quickly explains what works and what doesn't in search. That said, I did some digging. There are some techniques that Google doesn't officially support, but that are at least to some extent effective. As a disclaimer (and to clear my conscience if I decide to implement something in this post): These tactics are not guaranteed to work, but at least partially. I know that. What does Google unofficially support? 1. Noindex directive in Robots.Txt file The meta noindex tag, applied at the page level, is a reliable staple for excluding pages from search engine indexes.
However, Google also respects most robots directives in robots.txt files that contain noindex. advertisement Continue reading below Officially, Google's John Mueller states that noindex robots.txt is "not trustworthy", but testing with DeepCrawl (disclaimer: I work for DeepCrawl) indexes the page. Do not create This method shows that it photo color correction services will continue to work. Compared to meta noindex, robots.txt noindex provides a cleaner and more manageable solution to tell Google that pages shouldn't be indexed. This way, robots.txt overrides URL-level directives, so there is less confusion about which directive overrides which directive. Robots.txt noindex is better than metanoindex because you can noindex a group of URLs by specifying a URL pattern instead of applying it page by page.
As with any robots.txt directive, you can use the Robots.txt tester in the Search Console to test your robots.txt noindex. Robots.txt tester 2. Regular tag inserted by JS At this year's Google I / O, Tom Greenaway said that Google handles rel = canonical only when the page is first fetched and is lost if it relies on rendering by the client. advertisement Continue reading below If true, this can be a problem for SEO professionals making changes through Google Tag Manager. This may be the best option when faced with long development queues and content management systems that are difficult to customize. Following this, John Mueller later reiterated Greenaway's claim on Twitter. (Currently) Process rel = canonical only in the first fetched and unrendered version.