Search engines traditionally learn about changes by crawling: they re-fetch pages at intervals. That is inefficient for both sides. IndexNow reverses the principle: the website reports what has changed, and the search engine can crawl in a targeted way.
How it works
The website generates a key and places it as a text file on its own server to prove ownership. When something changes, it sends the affected URL or a list of URLs together with the key to an IndexNow endpoint. Participating search engines share submissions with each other, so one submission to one endpoint is enough.
Who supports it
IndexNow was introduced in 2021 by Microsoft Bing and Yandex. Other search engines such as Seznam and Naver have joined. Google has tested the protocol but does not use it for indexing so far. For Google, the sitemap, internal linking and URL inspection in Search Console remain the tools of choice.
Why it still matters
Bing is not only a search engine in its own right but a data source for other services, including AI assistants with web search. Faster freshness at Bing can therefore also help visibility in AI answers. For websites with frequent changes, such as shops, job portals or programmatic SEO sites, the time to update drops noticeably.
Implementation
Many CMSs and plugins support IndexNow directly. In custom applications such as Next.js, the key is served as a static file and a list of changed URLs is submitted after a deployment, for example in a post-deploy step of the CI/CD pipeline. It is important to submit only URLs that actually changed and are indexable; indiscriminate mass submissions achieve nothing and may be ignored.
Limits
IndexNow speeds up discovery, not evaluation. A thin or duplicate page is not indexed because it was submitted. And it replaces neither a clean sitemap nor good internal linking. We set up IndexNow as part of technical SEO audits where frequent changes justify it.