• Возможности
  • Тарифы
  • Вопросы
  • Блог
ВойтиНачать бесплатно
0%
Binetro

Защита

  • Защита от ботов
  • Защита от скликивания

Аналитика

  • Аналитика трафика
  • Отчёты

Подключение

  • Тильда
  • WordPress
  • Любой сайт

Начало работы

  • Как подключить
  • Тарифы
  • Блог

Документы

  • Политика конфиденциальности
  • Договор публичной оферты
  • Согласие на обработку данных
  • Политика использования cookie
  1. Blog
  2. Metrics and calculations

Robotness in Yandex Metrica: what the number is

Published October 24, 2026updated October 25, 2026Binetro editorial teamMetrics and calculations6 min read—
Contents
  1. Two figures people confuse with the budget
  2. Who the metric has already removed, and who it left in
  3. What not to switch on if you want to keep the history

Contents

  1. Two figures people confuse with the budget
  2. Who the metric has already removed, and who it left in
  3. What not to switch on if you want to keep the history

A high CTR isn't success yet

Binetro separates real visits from bot clicks and protects your ad campaign.

Protect your ads

Share

Robotness in Yandex Metrica is the share of visits the system classed as robots. It is not the share of your budget in Yandex Direct, and it is not a list of clicks you will be refunded. A Yandex Ads news post of 23 September 2021 introduced the metric as the share of robots in traffic. The help page, read on 24 October 2026, describes a different default mode from that post. What matters is the switch in your own report.

How the label “robot” differs from a person is covered in bot or human. This piece is what to do with the percent, and what not to expect from it.

Close-up of a server rack with blue indicator lights and no readable brand name
A system calculates the percent. A light on a rack does not say how much of that percent was charged in Direct. · Photo: Pexels License

Two figures people confuse with the budget

Suppose the site had 10,000 visits in a week, and robotness is 30%. That is 3,000 visits Metrica classed as robots under its rules. Beside that, a Direct campaign produced 400 charged clicks. Thirty percent of ten thousand is not thirty percent of four hundred. They are different baskets. The numbers in the example are round so the baskets are visible, not so that 30% becomes a norm. The guide has no norm.

  • Teaching example, count

Even robotness of the advertising source itself is still not the bill. The Metrica guide says outright that Metrica and Direct use different systems to decide who is a robot. The “Direct” report group in Metrica is built from clicks Direct’s antifraud let through. Other slices do not apply that filter. High robotness on the slice “whole site”, or on a slice that still contains clicks already removed, does not measure the budget. Where a charged click begins is set in what ad click fraud is. How not to treat the share as a sign of click fraud on its own is in signs of click fraud.

Who the metric has already removed, and who it left in

Robots that introduced themselves in the User-agent are dropped from ordinary statistics at once. You see them in the Robots report, by name. Behavior-based robots, the ones that mask themselves, are a separate line in that table and are not spelled out. The 2021 news post said the robotness metric appears in a report when you choose the mode “Data: with robots”, and it advised leaving the default at “Data: without robots”.

The help page on 24 October 2026 says otherwise. By default, statistics include visits of robots detected by behavior, and removing them from a report takes a segment. The two pictures are compared in the setting “Data: without robots”, by choosing “Compare with data that includes robots”. Until you have read the switch in your report, a remembered “default mode” is worth nothing. The news post describes the launch of the metric. The help page describes the screen that opens now. The gap is better left in the open.

Metrica can also class as robots the visits to addresses missing from the tag’s Address and Additional address fields. Before you trust the share, check the domain list. Otherwise robotness includes your own omitted address.

What not to switch on if you want to keep the history

The filter in the tag settings

The option “Filter robots by behavior” removes those robots from the source data, and it cannot be undone. After it is on, you can no longer change how the report displays robots. Yandex recommends it to people who pull raw data through the Logs API and want that data without robots. Otherwise the guide recommends leaving the option off and switching the mode inside the report.

The wording of the option was checked against the Metrica filters page. You do not need it to read the share. You need the “with robots” mode, or the comparison. Otherwise the denominator no longer contains the visits whose share you wanted to see.

A separate number is “Robotness Pro” in Metrica Pro. The package description calls it a flag in the Logs API and Data Streaming, from a more precise formula, for robot visits that may be tied to fraud. It is not the share in an ordinary report, and without the Pro package it is not in the interface. One cannot stand in for the other.

Zero leads at any robotness can still be a broken goal, or a real person who did not buy. The robot percent does not replace that check. The order is in clicks without leads.

A visit check on the site is not Metrica’s robotness. Even if the result is sent to the tag, it is a different mark: behavior, the browser environment, the network and the ad-click parameters, with an outcome to let the visitor through, to give the browser a short task, or to refuse the visit. It does not have to match Metrica’s share. Direct is a third system. You cannot subtract a robotness percent from ad spend under any of the three. If the check on the site does not answer, the visitor is let through, and that pass does not change Metrica’s robotness.

A high CTR isn't success yet

Binetro separates real visits from bot clicks and protects your ad campaign.

Protect your ads

Related

  • How to estimate budget lost to click fraud

    The loss is the share of spend you were charged, behind which there was no buyer. The invalid-clicks row is not part of that sum: you were not charged for it.

    5 min read
  • How to calculate cost per lead

    Cost per lead in the report is spend divided by leads. The conversion price on a Yandex Direct strategy is a different number: a limit you set, not the period's result.

    5 min read
  • What CTR is, and why there is no normal CTR

    There is no normal CTR. It is clicks divided by impressions, and one number for a whole campaign mixes search with the ad network. Compare like with like, over one period.

    5 min read
  • A seasonal drop or fraud: how to tell them apart

    The checked guides never name a season as the cause of a conversion drop. A drop you can see comes from a different report date, from campaign types you must not compare, or from a bounce threshold you can change.

    5 min read
All articles