Ad quality - part 1

Related articles 

• Quality at Equativ - overview
• Ad quality - part 2
• Traffic quality - part 1
• Traffic quality - part 2
• Implementing ads.txt and app-ads.txt with Equativ
• Sellers.json and SupplyChain Object

Overview

Ad quality management has two main goals:

• ensuring publisher brand safety: preventing any misplacements of inappropriate ads in publishers’ content in order to protect the publishers’ brand and reputation (e. g. not showing gambling ads on a child-friendly website)
• preventing bad ads: protecting inventory providers from the detrimental impact on user perception/retention caused by harmful ads, i. e.:
  • intrusive or pervasive ads; e. g. auto-redirects, pop-ups, dialog boxes, audio-on ads, autoplay video ads, in-display video ads etc.
  • malicious ads served to commit fraud, either at the time of ad rendering or afterwards; e. g. ad fraud, user device infection, scams etc. 

Equativ has installed a wide range of processes and tools to ensure that all served ads are of appropriate quality:

• dynamic, automated creative scanning, categorization and blocking using the technology of the ad verification cybersecurity vendor Confiant
• verification of the content on advertiser domains declared by the buyer and blocking of ads to protect publishers from inappropriate ads

Creative scanning

Equativ uses the technology of the ad verification cybersecurity vendor Confiant to scan all programmatic (RTB) ads for intrusive, malicious ads and inappropriate ad content.
Any new, unknown creatives are scanned automatically and approved or rejected within 5 minutes.
Some malicious ads unfold their harmful impact only under certain conditions (e. g. only if the battery life falls below a specific threshold). Therefore, creatives are scanned at least 10 times under various conditions and configurations.

List of blocked malicious ads

Equativ blocks the following types of malicious ads by default:

• Ad stacking - without the publisher’s knowledge, multiple ads are stacked on top of each other; only the first ad is visible while the hidden ads behind it run as if they were visible
• Cloaking - a technique used by bad actors to evade detection by showing a different ad experience to ad quality scanners than to real viewers
• Criminal scams - sophisticated schemes in which a threat actor uses social engineering or other forms of deception to gain access to sensitive user information such as financial accounts
• Cryptocurrency mining - an ad that mines cryptocurrency in the viewer's browser using JavaScript or Web Assembly. This can cause harm by putting a strain on the viewer's device; It can also slow down the viewer's experience on the website
• Exploit kit - an ad drops a payload which attempts to install malicious code on the device
• Fake ad server - an ad serving domain that masquerades as a platform, ad network or DSP serves malvertisements at high volumes
• Fake download update - an ad that tells the user to update Flash (or other extensions/programs) and downloads a malicious program instead
• Forceful redirect - an ad that tries to redirect the viewer to another website without interaction
• High Risk Ad Platforms (HRAP) - a list of high-risk ad platforms based on an ongoing analysis of the continued malicious activity emanating from these sources, rather than one specific incident; an ad will be characterized as HRAP if a domain on the HRAP list delivered it or is associated with it
• In-banner video - an unscrupulous entity buys display inventory and then attempts to resell it to a video exchange at a significant markup (e. g. a video ad is served in a medium rectangle slot for banners); also known as video arbitrage
• Phishing scam - ads that attempt to trick the viewer into giving personal information through a phishing scam; e. g. a “You’ve won an Amazon gift card!” redirect
• Pixel stuffing - an ad fraudulently loads the full landing page in a hidden iframe triggering conversion pixels
• Unsafe click tracker - a way for malvertisers to direct users to unsafe landing pages usually used to collect data
• Unsafe/Malware landing page - the landing page associated with an ad is deemed unsafe or malicious

Blocking of ads

Blocking of advertiser categories

An ad is blocked by Equativ if its advertiser belongs to a sensitive advertiser category. This blocking applies by default to ads from open auctions only – ads of sensitive advertiser categories from deals are not blocked. 
You can also block advertiser categories yourself directly in EMP: set up a delivery rule with the effect “Block” and select the desired category in the field “RTB advertiser categories”.
To ensure the quality and safety of the inventory, Equativ categorizes advertisers through an integration with a third party vendor: new advertiser domains are submitted to the vendor to retrieve the category of the given advertiser. Additionally, Equativ’s quality team does manual reviews of advertiser domains, in particular in case of sensitive advertisers.

The table below contains the list of advertiser categories. Note that Sensitive categories (Sensitive=1) are blocked by default in open auctions.

Blocking of advertiser domains

An ad is blocked by Equativ if its advertiser’s domain contains inappropriate content. You can also block individual advertiser domains yourself directly in Equativ's user interface: set up a delivery rule with the effect “Block” and select the desired advertiser domain in the field “RTB+ advertisers”.

Blocking of creative attributes

The Open RTB specification specifies a list of creative attributes that can be blocked (see chapter “5.3 Creative Attributes” on page 50 in the Open RTB specification).
From these creatives attributes, Equativ blocks the following by default:

• 3: Expandable (Automatic)
• 8. Pop (e.g. Over, Under, or Upon Exit)
• 9. Provocative or Suggestive Imagery
• 10. Shaky, Flashing, Flickering, Extreme Animation, Smileys
• 12. Text Only
• 14. Windows Dialog or Alert Style
• 17. Adobe Flash

Blocking of malware in direct insertions

In some cases, fraudulent activities may also involve malicious code served through direct (non-programmatic) insertions (e. g. malicious code in the field “Customized script for the creative”). 
If Equativ has clear indications of such activities, the insertions are put offline and the account is locked until the investigation is completed. Further legal steps may be initiated.

Customizations of ad quality settings upon request

As a Equativ customer, you can request customizations of some ad quality settings for your individual network - get back to your main contact at Equativ who will process your request. However, note that any final decision on such requests rests with Equativ's quality team.

The following settings can be customized upon request:

• Blocking advertisers that never bought in Equativ's RTB - you can send a list of advertiser domains to Equativ's quality team who will add them to the database manually; domains must be sent in a csv file containing the domains in format test.com or test.co.uk (no subdomains allowed); once added, you can block the advertisers using a delivery rule (see chapter “Blocking advertiser domains” above)
• Bulk upload into a delivery rule - in case of large numbers of advertiser domains to be blocked; create a delivery rule and a csv file with all domains to be blocked (format: test.com or test.co.uk; no subdomains allowed); then, send the list and the delivery rule name to your main point of contact at Equativ who will add the domains to the delivery rule
• Allowing blocked advertiser categories - Equativ blocks sensitive advertiser categories in open auctions by default (see table in chapter "Blocking of advertiser categories" above); you can ask Equativ to unblock certain sensitive categories
• Blocking specific creatives - you can send individual creative Ids to Equativ and request to block them; to determine the creative Id - see chapters “Transparency report - web” and “Transparency report - in-app” in the Ad quality - part 2 article.

Ad quality FAQ

What does Equativ do to prevent the display of bad ads on my properties (websites, apps)?

 
Can I block bad ads myself without any intervention of Equativ's quality team?

I have seen a bad ad on my website and I want to block it. How can I do that?

My app is using Equativ's in-app SDK and I am seeing a bad ad I want to block. How can I do that?