Your traffic is up, your conversion rate is down and your bounce rate keeps creeping higher. In the monthly report it looks as if your website has started to underperform. Often, something else is going on: a growing share of your visitors isn’t human at all. In Leadinfo’s data, the share of non-human traffic rose from 4.3% in the first quarter to 12.6% in the third quarter of 2026. That is roughly 1 in 8 visitors. Crawlers, scrapers and AI agents read your pages, sometimes fill in a form and then disappear. They count towards your sessions, but they never buy anything. In this article you’ll learn what bot traffic does to your analytics, why your dashboard doesn’t flag it and how to make your reporting reliable again.
What bot traffic is, and why it is growing so fast
Bot traffic is anything on your website that isn’t a person behind a screen. Think of search engine crawlers, monitoring tools that check whether your site is up, scrapers that copy your content or prices and, more recently, AI agents that fetch pages to answer a question for their user.
Bots used to be easy to spot. They didn’t run JavaScript, identified themselves with a recognisable name and often never showed up in your analytics in the first place. Google Analytics 4 also excludes traffic from known bots automatically.
The new generation behaves differently. An AI agent comparing suppliers for someone opens your site in a real browser, loads your scripts and clicks through to a second page. To your analytics, that looks like a visitor. B2B websites with detailed explanations, pricing pages and comparisons are particularly attractive to these agents, because that is where the answer they are looking for sits.
A share that has nearly tripled in two quarters fits that picture. No extra buyers are arriving. Extra readers are arriving who will never buy.
If you want to learn how to tell a person from a machine on individual visits, read AI traffic: bot or real visitor?. This article answers a different question: what all that traffic does to your reporting.
What non-human traffic does to your analytics
Sessions and users
The first effect is the most deceptive one: your traffic grows. Every bot that slips through the filter counts as a session and often as a new user. A rising line in your report feels like success, even when the number of people visiting your site has stayed the same.
Bounce rate and engagement
In Google Analytics 4, a session counts as engaged when it lasts longer than ten seconds, includes two or more page views or triggers a conversion. Everything else is a bounce. Many bots fetch a single page and leave straight away. As a result, your engagement rate drops and your bounce rate rises, while your human visitors behave exactly as they did before.
Conversion rate
Your conversion rate is a fraction, and bots only sit in the denominator. A worked example: say you have 2,000 human sessions and 40 enquiries. That is 2%. Add 300 bot sessions and you are measuring 40 enquiries across 2,300 sessions, so your conversion rate falls to 1.7%. Nothing went wrong on your website. You are simply dividing by the wrong number.
Forms and leads
Sometimes the effect works the other way. Bots that fill in forms make your conversion figures look better than they are. Sales feels the consequences: leads with fake names, nonsense in the message field and time spent following up on contacts that lead nowhere.
Sources and channels
Finally, your channel mix shifts. Bot traffic often arrives as Direct or Unassigned, from countries where you don’t operate or from cities that host large data centres. You see a channel grow without having done anything for it.
Why your dashboard doesn’t raise the alarm
If this much is changing, why doesn’t anyone notice? There are four reasons.
Alerts watch for drops. Most notifications in analytics and BI tools fire when traffic or conversions fall. Rising traffic triggers nothing, because that is what everyone wants to see.
The rise is gradual. There is no single day with an obvious spike. The share grows quarter by quarter, and each month looks much like the last. A few extra points of bounce rate are easily put down to seasonality or a new campaign.
The filter gives a false sense of security. Many marketers assume bots are already filtered out. That holds for known bots, but not for agents that behave like an ordinary browser. Whatever slips through never appears anywhere on its own.
Nobody owns it. Marketing reports the numbers, IT manages the website and sales receives the leads. As long as nobody compares those three views, the issue falls through the cracks.
The result is a familiar feeling: you can see there is traffic in your analytics, but you can’t do anything with it. If you want to know what Google Analytics does and doesn’t tell you about who visits your site, that is a good place to start.
How to make your reporting reliable again
You don’t need to identify every bot to trust your numbers again. Five steps will get you back in control.
- Set a baseline. Put sessions and engaged sessions side by side, per month, for the past twelve months. If the gap between the two lines is widening, your unengaged traffic is growing.
- Find the suspicious segments. Look at Direct and Unassigned traffic, sessions without engagement, new countries or cities and spikes in the middle of the night. That is where most bot traffic sits.
- Report on two bases. Show your conversion rate on all sessions and on engaged sessions. If the second stays stable while the first drops, you know your website isn’t performing worse.
- Protect your forms. A hidden extra field that only bots fill in, proper validation and a quick check before a lead goes to sales will keep your pipeline clean.
- Explain it in your report. Add a short note to your figures. That way, nobody in the management meeting draws the wrong conclusion from a falling conversion rate.
These steps also change the question you ask. No longer: how much traffic did we get? But: how much of that traffic had anything to do with our target audience? Volume says little as long as you don’t know who is behind it. What matters is lead quality, not the tallest line on your dashboard.
A session is not a company
Behind every B2B report sits one question: which companies visited, and are the right ones among them? A session count can’t answer that. A bot has no buying intent. A company does.
98% of your visitors leave without converting. Part of that group was never human. The rest is what matters: the organisations that looked at your offer, compared it and left again without filling in a form.
That is exactly what company identification shows you. Instead of a number, you see a company name, a sector and the pages they viewed. If you want to know more, here is how to identify the companies behind your traffic. Leadinfo works with 100% company identification without cookies, so visitors who decline your cookie banner stay visible too, with business data only and no personal data.
That gives you a second yardstick next to your session count. If your traffic rises but the number of companies that match your target audience doesn’t, you know where to look. If both rise, your growth is real. With a list of companies, marketing and sales have something they can actually work on together.
Getting started
Rising sessions alongside a falling conversion rate aren’t necessarily bad news about your website. Tomorrow, open your analytics and put sessions and engaged sessions side by side, per month, for the past year. Note where the gap widens and add that chart to your next monthly report with one sentence of explanation. If you also want to see which companies are really behind your traffic, add Leadinfo to your existing measurement, for example through the Google Analytics + Leadinfo integration. Start with one month and compare the number of companies from your target audience with your session count.
Frequently asked questions
How do I spot bot traffic in Google Analytics?
Look for sessions without engagement, unusual spikes in Direct or Unassigned traffic and visits from countries or cities where you have no customers. Also compare sessions with engaged sessions per month. If the gap widens while your campaigns stayed the same, the extra traffic is most likely not coming from people.
Doesn’t Google Analytics filter out bots automatically?
Google Analytics 4 automatically excludes traffic from known bots and spiders. That works for bots that identify themselves as bots. AI agents and scrapers that use an ordinary browser, load your scripts and click through often fall outside that filter. They are counted as normal in your sessions, your bounce rate and your conversion rate.
Why is my conversion rate dropping while my traffic is rising?
When your traffic rises but your number of enquiries stays the same, the base of your calculation gets bigger. If that growth comes from bots, your conversion rate drops without your website performing any worse. So also calculate your conversion rate on engaged sessions. If that stays stable, the problem lies in your traffic, not in your site.
