Yandex Metrica bot traffic: how to spot and stop fake visits

Yandex Metrica bot traffic shows up in the robot share: the portion of visits the system considers automated rather than human. A small share of bots is present on almost every website and is usually harmless. The trouble starts when that share jumps, your reports stop describing real customers, and leads and sales stay exactly where they were.

What the robot share actually tells you

Metrica judges each visit by the technical traits of the browser, behavior on the page and how repetitive the actions are. Some robots are identified strictly, from known signatures; others are flagged for suspicious behavior. The algorithm is not public and not perfect: a well-built bot can pass as a person, and an unusual real visitor, say someone with a rare browser, can be labeled a robot.

So treat the number as a compass, not a verdict. Watch it over time and compare it with what is happening in the business: an ad launch, a site update, a Telegram broadcast or an Instagram post.

The practical point is that a high bot share distorts everything else: bounce rate, time on site, page depth and conversion. Before judging a channel, make sure you are looking at data without robots and that their share did not change during the period you are analyzing.

Where bots on your site come from

Not every bot is hostile. Some are service traffic, some are plainly harmful.

  • search and service crawlers scanning your pages
  • uptime and speed monitoring tools that open pages on a schedule
  • competitor scrapers collecting prices and product ranges, for example from an online store catalog
  • scanners hunting for CMS vulnerabilities
  • paid bots that imitate human behavior
  • spam bots filling in website forms

What a bot looks like in your data

One sign proves nothing, but several together paint a fairly confident picture. Look for visits lasting a few seconds or all having exactly the same duration, a single pageview with no scrolling or cursor movement, odd screen resolutions and outdated browser versions.

Another signal is a spike at the same time of day. Real shoppers in Tashkent and the regions arrive in waves: morning, lunch, evening. A flat, even stream of visits at three in the morning is rarely human.

When it is normal and when to dig in

If the share is stable and real visitor behavior and lead numbers are unchanged, no special action is needed. These situations are different:

  • the robot share jumped although nothing changed on the site
  • many visits from one source with zero activity
  • a surge in direct visits or referrals from unfamiliar sites
  • growing branded search visits with no growth in leads

What to do when bot traffic spikes

Work step by step instead of blocking everything. First, pick one robot-handling mode in Metrica reports and keep it the same when comparing periods, otherwise the numbers will not be comparable. Then find the source: split traffic by source, region, device and landing page. Watch a few sessions in Webvisor, because bot behavior is usually obvious at once.

Next, check server logs for IP addresses, user agents and request frequency. Restrict suspicious traffic at the server or protection-service level, but leave search engine crawlers alone, or indexing will suffer. If you suspect someone is sending artificial traffic to your site on purpose, record dates, sources and report screenshots; you will need them if you contact search engine support.

Form spam and Telegram bot lead capture

Bots love forms. They clog the CRM, distract managers and corrupt Metrica goals. A captcha, hidden honeypot fields and a check on how fast the form was filled all help. Set the goal on a successful, validated submission, not on a button click.

For many businesses in Uzbekistan there is a simpler path as well: accept inquiries through a Telegram bot as well as a web form. The customer taps a button, shares a contact, answers a couple of questions, and the manager gets a notification in the app they already use all day. An automated spam script finds that dialogue harder to pass than an open form, though it is not a guarantee. markdev builds Telegram bots for business tasks and connects them to analytics, so the goal in Metrica becomes a real inquiry.

Do not buy traffic

Offers to "boost visits" or "improve behavioral factors" almost always mean bots. That traffic brings no customers, distorts analytics and risks penalties from search engines. If a contractor shows growth in visits without growth in leads, check the robot share and where those visits come from.

If you want to understand what is really happening with your site traffic and set up inquiry capture so bots do not spoil your statistics, talk through the task with the markdev team.

FAQ

What robot share in Metrica is considered normal?

There is no single norm, since almost every site has a small share of bots. Stability matters more. If the figure holds steady and leads and visitor behavior are fine, there is nothing to worry about.

Can bot traffic hurt my search rankings?

Artificial traffic, especially manipulation of behavioral signals, can lead to penalties even when the site owner is not involved. That is why spikes of suspicious visits should be documented and investigated.

Should I block all robots?

No. Never block search engine crawlers, or indexing will suffer. Restrict only suspicious traffic based on addresses, user agents and request frequency.

How do I protect a lead form from spam?

Use a captcha, hidden honeypot fields and a fill-time check, and tie the Metrica goal to a successful submission. You can also accept inquiries through a Telegram bot.

Let’s look at your task for free

Leave your contact — in a 15-minute call we’ll show where leads get lost and what to fix first.