Man
Professional
- Messages
- 3,068
- Reaction score
- 599
- Points
- 113
There are about 5 billion internet users worldwide. More than 2.1 billion of them regularly order goods online. So it’s no surprise that digital advertising is at its peak right now… And so is the level of fraud.
Marketers know which advertising models and formats attract the most attention from certain audiences. Pay-per-click or pay-per-action ads, video and audio formats, pay-per-impression advertising - the goal of each model is to generate the largest possible number of customers/buyers/users loyal to the brand.
And fraudsters know this too, so they create their traps, technologies and programs to deceive advertisers and steal their budgets. We will talk about click fraud in the CPM model.
What is the CPM advertising model, why you might encounter fraudulent impressions and how it works, what types exist and how to deal with them. Let's find out.
Contents
1. What is the CPM advertising model?
2. Clicks vs. Impressions
3. Generation of invalid impressions
4. How does impression fraud work?
5. Examples and methods of generating fake ad views
5.1. 1. Pixelation
5.2. 2. Overlaying Ads
5.3. 3. Autoview
5.4. 4. Bot traffic
5.5. 5. Click interception
6. Protection from advertising fraud
The CPM advertising formula looks like this:
CPM = (COST OF PLACEMENT / NUMBER OF IMPRESSIONS) * 1000
This model is common among advertisers who want to increase their brand recognition and authority. The strategy works well for new companies that have just appeared on the market. Placing CPM ads on multiple sites, both desktop and mobile, allows not only to increase the company's recognition, but also to increase the number of clicks in the CPC model in the future.
Theoretically, if an advertiser places pay-per-click ads, they only pay for those users who actually click on the ad. But in the case of impression advertising, any view of the ad located in the user's screen area is paid for.
Experts believe that less than half of all ad views in all programmatic advertising can be considered valid. Due to the generation of fake impressions, advertisers lose their budgets.
For example, the DrainerBot botnet was developed by a group of fraudsters to generate fake mobile ad views on Android devices. It became famous due to its scale - 10 million downloads of applications on Google Play infected with this malware.
The scheme of deception is as follows:
The simple nature of the deception, coupled with the high payout potential, creates an irresistible incentive for shady publishers and fraudsters. Without dedicated click fraud protection, the risk of impression fraud will be much higher, especially in programmatic advertising.
This can be achieved in several ways, but the most common is by creating a fake website.
Cybercriminals join an ad exchange to place ads. They then upload the ad to a single-pixel space on their resource. Bots and random users will generate impressions of real and hidden ads. The attackers will pass on these statistics and make money from each impression.
Along with other forms of impression and click fraud, pixelation can create serious problems for businesses. The main one is lost budget. As long as the number of impressions and cost per impression are lower than in CPC advertising, marketers often do not notice these expenses.
But what happens when fraudulent statistics distort the overall value of actual conversions? Advertisers may simply stop running their ads or adjust the product or its presentation to the audience based on the false statistics. This kind of misinformation within a marketing strategy can cause catastrophic damage.
Advertising fraud wastes budgets and often ruins promotional efforts. Sales targets are not met and leads are lost.
It is one of the most widespread forms of fraud, surpassing even credit card fraud. It accounts for 20% ($66 billion) of global advertising spend, affecting companies of all sizes, from small and medium to multinational.
Reported losses from this type of fraud in 2020 were $35 billion and are likely to reach $50 billion by 2025, according to Wfanet.org. North America and China suffered the highest losses in 2019, at $2.6 billion and $18.7 billion, respectively.
A fraudster can generate up to 20-30 requests every second over a 40-second period, crowding out real users and replacing them with bots. Statistics show that every fifth advertising site is visited exclusively by bots.
The characteristic signs of such an attack are:
Site owners should make sure that they are using an ad network that can protect their sites from fraud. There should be filters to filter out bot traffic based on user agent, referrer, IP address, time spent on the site, etc.
The use of such technology is possible with the help of HTML frames (iframes) - displaying some web pages inside others. That is, the fraudster simply creates several layers, one visible, and the second fraudulent, invisible to the user.
Probably many of you have visited sites with pop-up windows, where the close icon (X) was simply drawn. And when the user clicks on the "close window" icon, several more pages with advertised or malicious resources open in neighboring tabs. The real "close" button may be located elsewhere.
Click hijacking itself is not the end goal of the attack. It is simply a means to trigger a third-party action. It leads to the main fraudulent process: stealing credentials, installing malware, getting likes on social media, generating clicks and ad views.
Statistics may look too positive and exceed all expectations. But do not delude yourself. When it comes to programmatic advertising, you first need to make sure that the webmaster is not deceiving you and that real users, not bots, are viewing the ads. Be careful! After all, your budget and business development prospects are at stake.
Using advertising platforms makes advertisers vulnerable to attackers who generate fake clicks in Yandex.Direct, Google Ads, and also use cheating in social networks. The largest popular platforms have their own filters to filter out bot traffic and other types of fraud, but only the most obvious cases. More complex and self-learning bot scripts can bypass such protection.
The news regularly reports the discovery of yet another malicious software designed to click on ads. What to do? How to protect your site or advertising campaign from click fraud?
Marketers know which advertising models and formats attract the most attention from certain audiences. Pay-per-click or pay-per-action ads, video and audio formats, pay-per-impression advertising - the goal of each model is to generate the largest possible number of customers/buyers/users loyal to the brand.
And fraudsters know this too, so they create their traps, technologies and programs to deceive advertisers and steal their budgets. We will talk about click fraud in the CPM model.
What is the CPM advertising model, why you might encounter fraudulent impressions and how it works, what types exist and how to deal with them. Let's find out.
Contents
1. What is the CPM advertising model?
2. Clicks vs. Impressions
3. Generation of invalid impressions
4. How does impression fraud work?
5. Examples and methods of generating fake ad views
5.1. 1. Pixelation
5.2. 2. Overlaying Ads
5.3. 3. Autoview
5.4. 4. Bot traffic
5.5. 5. Click interception
6. Protection from advertising fraud
What is the CPM advertising model?
CPM (cost per mille, or cost per thousand impressions) is an advertising model in which the advertiser pays for every thousand views of an ad by site visitors on one page.The CPM advertising formula looks like this:
CPM = (COST OF PLACEMENT / NUMBER OF IMPRESSIONS) * 1000
This model is common among advertisers who want to increase their brand recognition and authority. The strategy works well for new companies that have just appeared on the market. Placing CPM ads on multiple sites, both desktop and mobile, allows not only to increase the company's recognition, but also to increase the number of clicks in the CPC model in the future.
Clicks vs. Impressions
The problem with pay-per-impression advertising is that it is impossible to determine whether a site visitor actually viewed the ad or whether it was playing "in the background." A banner or autoplay video can simply be ignored in favor of the key content on the page, but the impression will still be counted.Theoretically, if an advertiser places pay-per-click ads, they only pay for those users who actually click on the ad. But in the case of impression advertising, any view of the ad located in the user's screen area is paid for.
Generating invalid impressions
Ad fraud is when a fraudster tries to deceive an advertiser with fake leads or traffic. One form of this is impression fraud.Experts believe that less than half of all ad views in all programmatic advertising can be considered valid. Due to the generation of fake impressions, advertisers lose their budgets.
For example, the DrainerBot botnet was developed by a group of fraudsters to generate fake mobile ad views on Android devices. It became famous due to its scale - 10 million downloads of applications on Google Play infected with this malware.
How Impression Fraud Works
Impressions are a cumulative indicator of how many times an ad has been viewed by site visitors, regardless of whether they even paid attention to it or not. Therefore, it is not difficult for fraudsters to deceive an advertiser. However, it is more difficult to verify the fact of fraud in this case than with clicks.The scheme of deception is as follows:
- The scammer creates a fake website.
- Connects the resource to the advertising platform for placing banner and video advertising.
- Massively places advertisers' ads on the pages of a fake website.
- Launches bots to crawl pages and generate fake impressions.
- Passes false metrics on traffic and views.
- Receives rewards for fraudulent views.
The simple nature of the deception, coupled with the high payout potential, creates an irresistible incentive for shady publishers and fraudsters. Without dedicated click fraud protection, the risk of impression fraud will be much higher, especially in programmatic advertising.
Examples and methods of generating fake ad views
The most well-known fraudulent methods are ad pixelation, ad overlay, and bot traffic generation. In addition, fraudsters can also deceive advertisers who advertise mobile applications.1. Pixelation
Pixel stuffing is a scheme where a dishonest webmaster “stuffs” a page of his site with ads that fit into 1×1 pixels. The site visitor simply won’t notice them on the page. However, the view will still be counted.This can be achieved in several ways, but the most common is by creating a fake website.
Cybercriminals join an ad exchange to place ads. They then upload the ad to a single-pixel space on their resource. Bots and random users will generate impressions of real and hidden ads. The attackers will pass on these statistics and make money from each impression.
Along with other forms of impression and click fraud, pixelation can create serious problems for businesses. The main one is lost budget. As long as the number of impressions and cost per impression are lower than in CPC advertising, marketers often do not notice these expenses.
But what happens when fraudulent statistics distort the overall value of actual conversions? Advertisers may simply stop running their ads or adjust the product or its presentation to the audience based on the false statistics. This kind of misinformation within a marketing strategy can cause catastrophic damage.
Advertising fraud wastes budgets and often ruins promotional efforts. Sales targets are not met and leads are lost.
2. Overlaying ads
Ad stacking is a fraudulent scheme where one ad is placed on top of another. And so on - in several layers. The site visitor sees only the top one. However, the view is counted for all ads.It is one of the most widespread forms of fraud, surpassing even credit card fraud. It accounts for 20% ($66 billion) of global advertising spend, affecting companies of all sizes, from small and medium to multinational.
Reported losses from this type of fraud in 2020 were $35 billion and are likely to reach $50 billion by 2025, according to Wfanet.org. North America and China suffered the highest losses in 2019, at $2.6 billion and $18.7 billion, respectively.
3. Auto view
Automatic display of ads is typical for mobile devices, when ads are viewed in the background. The user's device is infected with malware, which gets on the phone with some malware-infected application, launches the browser and goes to the fraudster's sites where the ads are posted.4. Bot traffic
Bot traffic is the generation of transitions and visits to website pages by automated scripts. Fraudsters create a whole network of fake resources. They register on some advertising platform and place ads on all their websites. Then they drive bots to visit all the pages of the site and generate fake traffic and views of advertisers' ads.A fraudster can generate up to 20-30 requests every second over a 40-second period, crowding out real users and replacing them with bots. Statistics show that every fifth advertising site is visited exclusively by bots.
The characteristic signs of such an attack are:
- sudden surge in traffic;
- slow page loading;
- identical User Agent parameters;
- another geolocation is a region that the advertiser does not target at all.
Site owners should make sure that they are using an ad network that can protect their sites from fraud. There should be filters to filter out bot traffic based on user agent, referrer, IP address, time spent on the site, etc.
5. Click interception
Clickjacking is a fraudulent technique in which a user's click on an interface element causes an action other than the intended one. For example, a website visitor clicks on the "Get a promo code" button, but in fact there is another button hidden there with the "subscribe to newsletter" functionality.The use of such technology is possible with the help of HTML frames (iframes) - displaying some web pages inside others. That is, the fraudster simply creates several layers, one visible, and the second fraudulent, invisible to the user.
Probably many of you have visited sites with pop-up windows, where the close icon (X) was simply drawn. And when the user clicks on the "close window" icon, several more pages with advertised or malicious resources open in neighboring tabs. The real "close" button may be located elsewhere.
Click hijacking itself is not the end goal of the attack. It is simply a means to trigger a third-party action. It leads to the main fraudulent process: stealing credentials, installing malware, getting likes on social media, generating clicks and ad views.
Protection against advertising fraud
A huge number of ad views after the campaign launch may actually be a scam. If there is an element of fraud, then in the long run the advertiser will get zero conversion and damage to the advertising budget.Statistics may look too positive and exceed all expectations. But do not delude yourself. When it comes to programmatic advertising, you first need to make sure that the webmaster is not deceiving you and that real users, not bots, are viewing the ads. Be careful! After all, your budget and business development prospects are at stake.
Using advertising platforms makes advertisers vulnerable to attackers who generate fake clicks in Yandex.Direct, Google Ads, and also use cheating in social networks. The largest popular platforms have their own filters to filter out bot traffic and other types of fraud, but only the most obvious cases. More complex and self-learning bot scripts can bypass such protection.
The news regularly reports the discovery of yet another malicious software designed to click on ads. What to do? How to protect your site or advertising campaign from click fraud?
