IndexNow is a simple protocol that lets you instantly tell search engines a URL has been added, updated, or deleted, instead of waiting days for a crawler to notice on its own. You ping a single endpoint with your changed URLs and a key that proves you own the site, and participating engines — Microsoft Bing, Yandex, Seznam, Naver and others — pick up the change within minutes. Google does not currently consume IndexNow, but the protocol still shortens the discovery gap on a meaningful share of AI and search traffic.
Why waiting for the crawler is a problem
Search engines discover changes by recrawling your pages, and recrawl frequency depends on how important and how fresh an engine thinks a page is. A brand-new page on a low-authority site can sit undiscovered for days or weeks. That lag hurts most in exactly the situations where speed matters: a product that just came back in stock, a news update, a corrected price, or a page you removed and no longer want served. AI answer engines that rely on Bing's index — including Microsoft Copilot — inherit that same delay, so a stale index can mean an assistant confidently citing information you already fixed.
IndexNow flips the model from pull to push. Rather than hoping a bot returns soon, you announce the change the moment it happens. It does not guarantee indexing or ranking — the engine still decides whether the page deserves a place in its index — but it removes the discovery bottleneck, which is often the slowest step.
How IndexNow works, step by step
The protocol has three moving parts, and all of them are deliberately minimal:
- Generate a key. Create a random hexadecimal string (at least eight characters). This is your shared secret that proves you control the domain.
- Host the key file. Place a text file at the root of your site — for example https://example.com/1a2b3c4d5e.txt — whose only contents are the key itself. When you submit URLs, the engine fetches this file to confirm ownership.
- Submit URLs. Send a request to an IndexNow endpoint listing the changed URLs and your key. You can submit a single URL with a simple GET request, or batch up to 10,000 URLs in one JSON POST.
A batch submission is just a small JSON body containing your host, your key, the key file location, and a urlList array. Because the participating engines share submissions with each other, you only need to notify one endpoint and the change propagates across the network. A 200 response means the submission was accepted; a 403 usually means the key file could not be verified.
Setting it up without breaking anything
The cleanest approach is to fire IndexNow automatically from your publishing workflow. Whenever your CMS saves, updates, or deletes a page, have it POST the affected URL to the endpoint. Most major platforms already offer this: Cloudflare can enable IndexNow at the edge with one toggle, and popular CMS plugins handle key generation and submission for you. If you deploy statically, a short script in your build pipeline can submit every changed URL after each publish.
A few rules keep you out of trouble. Only submit URLs you genuinely changed — spamming unchanged pages can get your submissions throttled or ignored. Keep the key file reachable and unchanged, since a rotated key with a stale file will fail verification. And treat IndexNow as a complement to, not a replacement for, your XML sitemap: the sitemap remains the canonical map of your site for engines that do not support instant push, while IndexNow handles the urgent, real-time signal.
Where IndexNow fits in your crawl strategy
Think of your crawl setup as layered. Your robots.txt file controls which bots may access what. Your sitemap tells engines the full shape of your site and when pages last changed. IndexNow adds a real-time nudge on top, so the moment something important changes, the engines that support it hear about it immediately rather than on their next scheduled pass. Together these three mechanisms shrink the window between a change on your server and that change appearing in results and AI answers.
For large or fast-moving sites, that window is not a rounding error — it is the difference between an assistant citing your current information and citing a version you retired last week. Understanding how search engines crawl and index websites makes it obvious why pushing changes beats waiting for a pull.
Want to see whether engines and AI crawlers can actually reach and read the pages you are trying to get indexed? Run the free CheckMy.site scanner to check your robots rules, sitemap, and bot access in seconds before you rely on IndexNow to speed things up.