How Ad Servers Detect suspicious activity on a site or app

How Ad Servers Detect suspicious activity on a site or app: Online advertising now becomes a giant industry. Expenditure reached to $300bn. It is now very popular adverting system as effective and efficient.  The advertisers can know who, where, when and how people are seeing the advertisements. How many times that advertisement was seen by visitors. What was their reaction? Advertisers also can target the specific group of people to target for the product based on the buyers’ interests. So, there are a lot of advantages of online advertising.  But there are some disadvantages too.

Non-Human Traffic: The main disadvantage of online advertising is Non-human Traffic. It refers to robots, spiders, crawlers, and agents. Online advertising ecosystem works through software which maintains several protocols like – real-time bidding, target-matching, ad hosting and serving. Overall mechanism performed very fast, usually, it takes 2/10th of a second.  But some inventory owners misuse this advertising ecosystem by using the glitch. There is some software that works like a human and generates fake impressions as well as highjacked money from advertisers. You can get more details about non-human traffic from IAS. To overcome this, most of the ad-platforms and third-party vendors using several observing tools.

Non-Human Traffic Example:

Usage of automated bots that imitate user behavior
Suspicious behaviors of users; visiting sites, click on ads etc.
Buying unrelated traffic from unknown networks
Click Fraud, another type non-human behaviors, usually click on the ads.
Participating in pay-to-click programs
Receiving traffic from unwanted email or software applications

 

How Ad Servers Detect suspicious activity on a site or app

Every ad tag has a tracking pixel that collects data from visitors browsing cache and cookie.  From a single visit on a website, this tracking pixel can collect IP address, MAC address even some tracking pixel can collect Hard Disk number. This pixel collects visitor’s browsing behaviors, other social networking activities. So, these pixels record all of your buying, living data, interest, sex, ages etc.  Some servers offer the user to create free profile, email account. By this, they also can track its users.  If you log-in from your Desktop, they will collect your data. Once you log into the same account or profile, they will match the data and consider both the user as a single user.  Here is an example of Google’s ad targeting that matched your buying behaviors.

There are several medias work in this advertising ecosystem. From creative hosting to serving ads on other ad-server’s websites.

  • Creative Hosting: Sizmek, Moat is creative hosting platform which has own tracking pixel and serves to other ad exchange.
  • Ad Server: DFP, AOL, Yahoo is an example of ad-server (they also host creative) and use Sizemek/Moat’s tag
  • RTB: Through RTB (Real-time bidding), DSP, SSP, Networks are interconnected with each other and share the ad-viewing behaviors.
  • Third party tracker: There are some other agencies who also provide the data.

 

 

How Ad Servers Detect suspicious activity on a site or app

Decision criteria: If the tracking pixel data found some suspicious activity like 12 impressions in one minute from the same IP. If they found most of the visitor’s do not have the interest in that article but they visit or click on ads several times etc. finally, if this is a regular behavior on a site and then that put the site as blacklisted. All the servers share their findings. If a site is blacklisted on AOL, the site may be blocked on other ad-networks; like – Adkarma. They have just one difference, which is policy. different ad-networks or servers have different policies (though IAB setting a common rule for all), some allow up to 25% suspicious activity, some do not allow more 15%.

Suspicious activity tracking vendors:
If you are an advertiser or ad-network, you can use following services to get your true impression result. They also provide viewability data and all the matrix you need.

  • IAS: Integral Ad Science is one of the most reliable third-party ad tracker services and most of the ad-networks and ad-servers use IAS service. IAS provides viewability, suspicious, bounce rate data. They also can track from where a traffic came from and where they move, how long visitors are staying on that page and what actions the visitors made.

  • True URLs from DoubleVerify: DoubleVerify is another third-party ad tracking service and data provider. They can provide real traffic information. DoubleVerify is now partnered with Facebook. So they have bunch of data in their system.

  • Own tracking system of all Ad servers: Beside those third-party data providers, all the ad-servers have its own tracking system. So, they can also track all the relevant data. AOL, Yahoo, Google (DFP, DFA, DMA, DCM), Moat, Rocket-Fuel are the example of Ad-servers.
Ranking of Digital Inventories (APP, SITE, GAMING CONSOLE) required to get Approve
  • Site performance tracker: beside the ad-servers, there is some service provider in the digital ad ecosystem. they basically provide sites’ ranking and traffic source information. SimilarWeb and Alexa is a part of that. Most of the ad-server took data from them to check the ranking, traffic depth to approve publishers’ sites.
  • App performance tracker: There is another site, Mixrank, provide the ranking information of android app or IOS.

 

Summary
How Ad Servers Detect suspicious activity on a site or app
Article Name
How Ad Servers Detect suspicious activity on a site or app
Description
How Ad Servers Detect suspicious activity on a site or app. some allow up to 25% suspicious activity, some do not allow more 15%. suspicious activity like 12 impressions in one minute from the same IP
Publisher Name
Online Ad Tech

Leave a Reply

Your email address will not be published. Required fields are marked *