Know which posts get read. No cookies needed.

• Overview. One card with your blog’s reads over the last 30 days. If nothing has been counted yet, it says so and links to the fix.
• Posts. A Views (30d) column beside every published post, so the posts that work stand out in the list you already use.
• Each post. Open a published post to see its views, unique readers, clicks and click rate, a daily chart that marks the peak day, the links readers pressed and the countries they read from.

A view is reported by the reader’s browser when the post is actually opened. Your site fetching a post is not a view, which keeps build steps, cached pages and search-engine crawlers out of your numbers.

Nothing is stored in the reader’s browser apart from a note, cleared when the tab closes, of which posts it has already reported. To tell one reader from another for a day, our server combines a random value that changes every day with your website, the reader’s IP address and their browser’s name and version, and keeps only a one-way hash of the result.

• The IP address and the browser string are never stored.
• The daily random value is deleted after 48 hours. After that, nobody, including us, can work out which hash belonged to whom.
• The hash is kept with each reader event for 90 days, then deleted. Daily totals, which contain no identifiers, are kept.
• The page address is stored without its query string, where campaign codes and sometimes email addresses end up.

Because the hash changes every day, a reader cannot be followed from one day to the next or from your site to another. The full account is in our privacy policy.

'use client'

import { useTrackView } from 'cmskite/react'

export function TrackView({ postId }: { postId: string }) {
  useTrackView(postId, { apiKey: process.env.NEXT_PUBLIC_CMSKITE_KEY ?? '' })
  return null
}

Clicks are counted on coded sites, with trackClick from the same package; the free blog link and the embed report views only. If every post reads zero, Why your views are zero finds the cause in a minute and has the snippet for plain HTML, PHP and WordPress too.

• Same network, same browser, same day: one reader. Two colleagues on the office Wi-Fi with the same browser version count once. Mobile networks often put many phones behind one shared address, so the same can happen there. That is the cost of not using a cookie.
• Returning readers are not recognised. Somebody who reads on Monday and Tuesday counts on both days, and there is no “returning visitor” figure.
• No funnels, sessions, scroll depth or time on page, and no list of individual visitors. There is no visitor to look up.
• Bots are left out when they say what they are: crawlers, link previews, curl, and automated browsers such as Playwright. A bot pretending to be an ordinary browser is counted.
• Ad blockers and privacy extensions can block the report, so some real readers are missed.
• The dashboard shows the last 30 days. Other ranges are available through the API.

New views appear within about a minute. If you need funnels or campaign attribution as well, a full analytics tool can run on your site alongside this one.

Reader numbers are their own permission. Owners, admins, developers and analysts have it; editors and viewers do not, unless somebody gives it to them. See Team and roles.

The same figures are in the API, with a read-only key: a project summary, the top posts, one post’s series and breakdowns, and totals for up to 50 posts at once. An AI assistant connected through MCP can read them too. The endpoints are in the analytics docs.

Reader numbers are the same on every plan, Free included: views, unique readers, clicks and countries for every post. What changes between plans is how much you can publish.