Real visitors vs bots

Most website analytics count more visitors than a site actually had. A search engine fetching your pages, a security scanner trying /wp-login.php, a script with a browser's name pasted into it: each one looks like a visit to a counter that trusts what it is told. On our own marketing site we measured it in November 2026, and about 85 of every 100 requests that passed every ordinary check were not people.

Your Insights never counted those, and now they say so. The Real visitors vs bots card shows the two halves side by side, so the number you make decisions on comes with the number that was kept out of it.

How a page view is counted

A page view on Insights needs a real browser to run the small script on your page and stay on it for four seconds, or touch it before then. A request that fetches the HTML and leaves, as every crawler does, never fires it. So Visits and Page views on the analytics dashboard were always people; what was missing was the other side of the ledger.

How bots are recognized

Every request your site answers is looked at before it is counted, and it is labeled a bot when:

• it names itself: Googlebot, Bingbot, Yandex, the AI crawlers, the uptime checkers and SEO tools all say who they are;
• it is a script (a Python or Go HTTP library, a command-line tool) or a scanner asking for pages only attackers ask for;
• it comes from a datacenter address rather than a home, office or mobile connection. Real people almost never browse from a server; crawlers almost always do;
• it is a fake browser: it claims to be Chrome but is missing the headers every Chrome sends, or arrives "from" a page on your site that it never loaded;
• it behaves like a crawler: dozens of pages in a minute, or the same address reading many unrelated sites in a day.

Each label is a reading, not a judgment on the visitor. A bot is never blocked by this; it is served the page and left out of your numbers.

Reading the card

• Real page views is the same number the chart above it uses.
• Bot requests kept out is how many requests in the period were labeled as above. They are not in any figure on the page, not in the funnel, not in Top pages.
• The bar and the sentence give the share. On a quiet site, bots are usually the majority; that is normal, not a problem. A new site with ten real visitors and three hundred bot requests is a new site that search engines have found.
• Who they were names the crawlers behind the bot count and how many requests each made. This detail is on paid plans; the count itself is on every plan. See Plan limits.

What to do about them

Usually nothing. Googlebot and Bingbot in the list is what you want: it means your pages are being read for search. A security scanner probing for WordPress files on a site that has none gets a "not found" and moves on. A datacenter fetch costs you nothing.

Two things are worth a second look:

• A crawler you do not recognize making hundreds of requests a day. Most are SEO tools and AI crawlers reading the web. They cost you nothing and are already outside your numbers. Your site's robots.txt is managed by Webkio for every site, so there is nothing to edit; if a crawler is genuinely causing you a problem, tell us and we will look at it platform-wide.
• A sudden jump in real page views with no change in the rest of the page. Check Where visitors came from and the conversion funnel. Real traffic brings a few messages or orders with it; a spike that brings nothing is worth watching.

The Monday note

Once a week, on Monday morning, the bell in your dashboard gets a one-line note for each live site: how many real visitors it had last week and how many bot requests were kept out. It is a note in the app, not an email, and it links to Insights. A site with nothing to report gets no note.

Why the numbers differ from other tools

If you also run Google Analytics or a similar tool, expect its visitor count to be higher. Script blockers and consent banners hide some real people from it, which pulls it down, but it also counts any client that runs JavaScript, including headless browsers, which pulls it up, and it does not look at the requests that never ran the script at all. Neither number is "the truth"; this one is the smaller and more conservative of the two, and the one your messages, bookings and orders will match. For what to measure and why, see What to measure on a small business website and Where website visitors drop off.

Frequently asked questions

Can I block the bots?

Not from this card, and mostly you should not want to: search engines need to crawl you to rank you. Your site's robots.txt is managed by Webkio, the same for every site, and abusive scanners are handled by the platform's security rules. Password protection or country rules in Site status & visibility are for people, not bots.

Why is my bot share so high?

Because bot traffic on the web is roughly constant and human traffic is not. A site with few real visitors has the same crawlers as a busy one, so the share looks alarming while the count is ordinary. Watch the real number; the share falls as it grows.

Do bots affect my Site Health score or page speed figures?

No. Page speed is measured from real browsers only, and the Site Health score reads your pages, not your visitors.

Are my own visits counted as real?

Yes, like anyone else's: a real browser on your page for four seconds is a real page view, whoever is behind it. On a new site your own checks can be a visible share of the number, so look at the trend rather than the first week's total, and prefer the builder's preview when you only want to look.

Does "kept out" mean the request was refused?

No. Every request was answered; "kept out" means it was not added to your analytics. Blocking is a separate matter, handled by the platform's security rules for genuine attacks.

Which plans see the crawler names?

Every paid plan. The Free plan sees the count and the share. See Pricing.