Man
Professional
- Messages
- 3,070
- Reaction score
- 601
- Points
- 113
Invalid traffic is false impressions or clicks made by a real person or a bot, either accidentally or intentionally, leading to an increase in advertising budget costs. For example, such traffic includes motivated click fraud orders, accidental clicks on your own ads, automated targeted botnet attacks, etc.
The problem of invalid traffic in contextual advertising is not new, and technologies for its detection do not stand still. Advertisers and advertising partners have access to advanced tools for monitoring, analyzing and stopping invalid clicks and fraudulent activity.
However, the flow of automated advertising fraud, for example, from various botnets such as Methbot (liquidated, brought its creators 3-5 million dollars per day), DrainerBot, SilentFade, continues to fall upon advertisers.
On the other hand, it affects not only the advertiser's budgets, but also the status of the partner on whose platform the advertisement is placed. His account, in case of suspicious activity and invalid clicks or impressions, can be blocked, even if there is no direct fault on his part.
Now let's look at what it is, what types there are, and what you can do to protect your reputation and budget.
Contents
1. About invalid traffic in Google Adwords
2. What are invalid clicks in Yandex.Direct
3. Types of traffic
3.1. Total Traffic (GIVT)
3.2. Slow to detect traffic (SIVT)
4. How invalid traffic affects advertising platform owners
4.1. What can you get banned for?
5. How to avoid encountering invalid traffic
5.1 How to identify invalid traffic in Google Adwords
5.2. How to detect invalid clicks in Yandex.Direct
There is a guarantee for clicks on Google Ads ads, when within two months (no more) from the moment of the pseudo-click you can return the money written off for it. Information about this is available in the "Payment" tab, label "Invalid activity". Such traffic does not affect the campaign CTR.
Offline filtering works next. It “catches” fraudulent or erroneous clicks that bypassed the online filter and for which money was written off. You can see it in the “Report Wizard” — “Invalid Clicks”. In this case, the money is returned to the advertiser.
Comparison table of two types of invalid traffic:
They are not real users and are used for their own purposes. They help to correctly interpret the results of work on the resource and improve the indicators, that is, they work for the benefit of the resource. And, most importantly, invalid general traffic does not interact with advertising placed on the site.
To detect and exclude it, a marketer or advertiser will not need any additional efforts, since it is determined automatically by the advertising systems themselves (if we are talking about Direct and Adwords).
It is more difficult to detect, as cybercriminals can use botnets that effectively imitate the actions of real users. To detect it, you have to resort to advanced analytics, services and specialists who specialize in analyzing and identifying fraudulent activities.
There are two ways in which invalidated traffic impacts advertising partners:
Similar methods of combating fraud are also available in other advertising systems, including Yandex.Direct. From this we can also conclude that protection from click fraud should be taken care of not only by one party — the advertiser — but also by the advertising platform.
► Spam[/B]
Mass fake account registrations on websites and spam comments. Spam causes a heavy load on the server, to the point that it becomes unavailable. It also leads to theft of data from comments, feeds and online chats and brings reputational risks.
► Audience Purchase[/B]
If you buy motivated clicks on specialized sites (buxes), this increases the risk of low-quality traffic, including that created by bots.
► Parsers[/B]
Good and quality content is worth a lot. Most likely, you have already seen your article more than once, which word for word was posted on other sites. It is quite possible that content parsers worked here. These are mass visits, one could say, a robot attack on the site. Due to which the site's bandwidth decreases and real users cannot access the site.
► Botnets[/B]
How ads are shown to real visitors: Browsers have cookies with certain information - the user's interests (see What is a fingerprint) - based on which the corresponding ads are displayed. Fraudsters, using programmed bots, collect such cookies on sites with high-quality content and artificially increase the cost of ad impressions.
Analyzing traffic patterns and behavior is the main way to prevent invalid clicks. Analyze location, segments, and channels. Advertising platform owners can use the Botfaqtor service to block bots.
► Don't click on ads on your own site[/B]
This is an appeal, rather, to novice advertising partners who think that they can increase their income this way. Self-clicks are easily identified and can lead to blocking an account in the advertising network.
► Avoid placing ads on low-quality resources[/B]
The worse the donor, the higher the risk of low-quality traffic. You should not resort to the services of unknown intermediaries or little-known advertising networks with dubious donors. It is advisable to familiarize yourself with the terms and guarantees in advance.
► Sudden surge in traffic[/B]
An increase in traffic is quite natural when the owner of an advertising platform makes efforts to improve the quality of content and UX on their resource, optimizes it for search engines. But if you begin to notice that today your site receives 5,000 visits, and tomorrow it is already 25,000, then there is a decline again, think about it. Analyze the visits and evaluate them.
► High bounce rate (~100%)[/B]
The bounce rate shows the percentage of users who visited the site and left it within a few seconds, i.e. did not find the information they needed. The lower this rate, the more accurately your resource meets people's requests. However, when the rate approaches 100%, sound the alarm. It is quite possible that such traffic can be considered invalid.
► Time spent on site is close to 0[/B]
Another way to recognize NT. The average session duration is calculated by dividing the total duration of all sessions (in seconds) by their number. If this indicator is close to 0, then this may be a characteristic sign of fraud - motivated (purchased) or automated (bots).
If in the first one you see suspicious mass transitions on ads lasting 1-2 seconds with the same type of behavior, then this is a reason to study the Yandex.Direct summary in more detail. If the number of clicks considered invalid matches the number of transitions that you analyzed in Metrica and considered suspicious, then there is nothing to worry about.
Please note ! The "Direct, summary" may not display clicks that did not pass the advertising service's protection system, i.e. money for them was not actually written off.
If you have found a behavior pattern yourself, set up a bid adjustment for this segment (minus 50-100%). If you have not found an obvious pattern, but are almost certain that it is fraud, go to Yandex.Audiences, create a look-alike segment and set a bid adjustment for it.
The problem of invalid traffic in contextual advertising is not new, and technologies for its detection do not stand still. Advertisers and advertising partners have access to advanced tools for monitoring, analyzing and stopping invalid clicks and fraudulent activity.
However, the flow of automated advertising fraud, for example, from various botnets such as Methbot (liquidated, brought its creators 3-5 million dollars per day), DrainerBot, SilentFade, continues to fall upon advertisers.
On the other hand, it affects not only the advertiser's budgets, but also the status of the partner on whose platform the advertisement is placed. His account, in case of suspicious activity and invalid clicks or impressions, can be blocked, even if there is no direct fault on his part.
Now let's look at what it is, what types there are, and what you can do to protect your reputation and budget.
Contents
1. About invalid traffic in Google Adwords
2. What are invalid clicks in Yandex.Direct
3. Types of traffic
3.1. Total Traffic (GIVT)
3.2. Slow to detect traffic (SIVT)
4. How invalid traffic affects advertising platform owners
4.1. What can you get banned for?
5. How to avoid encountering invalid traffic
5.1 How to identify invalid traffic in Google Adwords
5.2. How to detect invalid clicks in Yandex.Direct
About Invalid Traffic in Google Adwords
IVT (invalid traffic) — traffic in Adwords of random or artificial origin, which did not pass the Google filtering funnel and is considered invalid. Google removes such transitions or impressions from statistics — they are not paid. However, the advertiser will be able to view them in a detailed report.There is a guarantee for clicks on Google Ads ads, when within two months (no more) from the moment of the pseudo-click you can return the money written off for it. Information about this is available in the "Payment" tab, label "Invalid activity". Such traffic does not affect the campaign CTR.
What are invalid clicks in Yandex.Direct
Direct also has protection against invalid clicks, including intentional ones. Online and post filters are used for this. Directed artificial traffic is cut off automatically immediately, if it comes to bots. Payment is not debited.Offline filtering works next. It “catches” fraudulent or erroneous clicks that bypassed the online filter and for which money was written off. You can see it in the “Report Wizard” — “Invalid Clicks”. In this case, the money is returned to the advertiser.
Types of traffic
It can be divided into two types. General Invalid Traffic (GIVT), which includes search engine robots, visits from data center servers, and repeated erroneous clicks, and Sophisticated Invalid Traffic (SIVT), which is difficult to identify. The second type includes any actions by fraudsters who resort to complex tools and services to hide or disguise their fraudulent actions for clicking. For example, motivated (purchased) clicks through PTCs for earnings, interface masking, hacked browser extensions.Comparison table of two types of invalid traffic:
| General | Difficult | |
| Intention | Harmless | Fraudulent |
| How to determine | Easily determined using primary filtration and other standards | Advanced analysis, collaboration with other analytics services on specific parameters and manual verification are required |
| Examples | — data centers; — search engines and other robots; — unknown browsers or other web browsers; — accidental clicking and closing the page before it loads. | — bots and crawlers that imitate the actions of real users; — hacked devices; — invalid proxy traffic; — adware; — malware; — measurement manipulation; — false sites and ads; — cookie substitution. |
Total Traffic (GIVT)
ONT is filtered automatically, so marketers practically do not pay attention to it, since index bots and other crawlers do not threaten the overall traffic of the resource.They are not real users and are used for their own purposes. They help to correctly interpret the results of work on the resource and improve the indicators, that is, they work for the benefit of the resource. And, most importantly, invalid general traffic does not interact with advertising placed on the site.
To detect and exclude it, a marketer or advertiser will not need any additional efforts, since it is determined automatically by the advertising systems themselves (if we are talking about Direct and Adwords).
Hard to detect traffic (SIVT)
The generation of such traffic is used for the purpose of fraudulent enrichment and increasing profits from placing ads or interacting with them. But that's not all. Fraudsters use various techniques: hacking devices, manipulating location data, phishing, etc.It is more difficult to detect, as cybercriminals can use botnets that effectively imitate the actions of real users. To detect it, you have to resort to advanced analytics, services and specialists who specialize in analyzing and identifying fraudulent activities.
How Invalid Traffic Affects Ad Platform Owners
According to an analysis conducted by the IAB and Ernst & Young in 2015, invalid site visits cost the American contextual advertising industry $4.6 billion. This amount includes not only the fraud itself, which caused the loss of advertising budgets, but also the cost of special tools and services to combat fraud.There are two ways in which invalidated traffic impacts advertising partners:
- Reputational damage. Advertisers have access to tools to accurately determine the quality of traffic. If they see that a site they place ads on systematically sends fraudulent traffic, they can lower the advertising rate for that site or even blacklist the partner.
- Account suspension or blocking. If we look at the Google support forum, we will see a sea of messages from owners of advertising platforms whose accounts were blocked due to invalid transitions. Before the account is suspended, the check for click fraud in the Google advertising network goes through many stages and filters, including automatic filters, online and offline monitoring, manual analysis, botnet detection. And it will be problematic to restore such an account.
Similar methods of combating fraud are also available in other advertising systems, including Yandex.Direct. From this we can also conclude that protection from click fraud should be taken care of not only by one party — the advertiser — but also by the advertising platform.
What can you get banned for?
Invalid traffic not only affects the income of advertising platform owners, but also reduces the cost of real impressions up to blocking, as we indicated above. Here are five reasons why you should monitor the traffic on your resource.► Spam[/B]
Mass fake account registrations on websites and spam comments. Spam causes a heavy load on the server, to the point that it becomes unavailable. It also leads to theft of data from comments, feeds and online chats and brings reputational risks.
► Audience Purchase[/B]
If you buy motivated clicks on specialized sites (buxes), this increases the risk of low-quality traffic, including that created by bots.
► Parsers[/B]
Good and quality content is worth a lot. Most likely, you have already seen your article more than once, which word for word was posted on other sites. It is quite possible that content parsers worked here. These are mass visits, one could say, a robot attack on the site. Due to which the site's bandwidth decreases and real users cannot access the site.
► Botnets[/B]
How ads are shown to real visitors: Browsers have cookies with certain information - the user's interests (see What is a fingerprint) - based on which the corresponding ads are displayed. Fraudsters, using programmed bots, collect such cookies on sites with high-quality content and artificially increase the cost of ad impressions.
How to avoid encountering invalid traffic
► Evaluation and knowledge of your visitors[/B]Analyzing traffic patterns and behavior is the main way to prevent invalid clicks. Analyze location, segments, and channels. Advertising platform owners can use the Botfaqtor service to block bots.
► Don't click on ads on your own site[/B]
This is an appeal, rather, to novice advertising partners who think that they can increase their income this way. Self-clicks are easily identified and can lead to blocking an account in the advertising network.
► Avoid placing ads on low-quality resources[/B]
The worse the donor, the higher the risk of low-quality traffic. You should not resort to the services of unknown intermediaries or little-known advertising networks with dubious donors. It is advisable to familiarize yourself with the terms and guarantees in advance.
How to Identify Invalid Traffic in Google Adwords
In addition to using special services to detect fraud, advertisers also have access to primary free analysis tools within Google Analytics:► Sudden surge in traffic[/B]
An increase in traffic is quite natural when the owner of an advertising platform makes efforts to improve the quality of content and UX on their resource, optimizes it for search engines. But if you begin to notice that today your site receives 5,000 visits, and tomorrow it is already 25,000, then there is a decline again, think about it. Analyze the visits and evaluate them.
► High bounce rate (~100%)[/B]
The bounce rate shows the percentage of users who visited the site and left it within a few seconds, i.e. did not find the information they needed. The lower this rate, the more accurately your resource meets people's requests. However, when the rate approaches 100%, sound the alarm. It is quite possible that such traffic can be considered invalid.
► Time spent on site is close to 0[/B]
Another way to recognize NT. The average session duration is calculated by dividing the total duration of all sessions (in seconds) by their number. If this indicator is close to 0, then this may be a characteristic sign of fraud - motivated (purchased) or automated (bots).
How to detect invalid clicks in Yandex.Direct
Use the Yandex.Metrica tool to identify fraudulent traffic. Analyze visits in Webvisor and compare the data in the "Direct, summary" section.If in the first one you see suspicious mass transitions on ads lasting 1-2 seconds with the same type of behavior, then this is a reason to study the Yandex.Direct summary in more detail. If the number of clicks considered invalid matches the number of transitions that you analyzed in Metrica and considered suspicious, then there is nothing to worry about.
Please note ! The "Direct, summary" may not display clicks that did not pass the advertising service's protection system, i.e. money for them was not actually written off.
If you have found a behavior pattern yourself, set up a bid adjustment for this segment (minus 50-100%). If you have not found an obvious pattern, but are almost certain that it is fraud, go to Yandex.Audiences, create a look-alike segment and set a bid adjustment for it.
