Publicis Media UK Programmatic Brand Safety Protocols

Publicis Media is dedicated to helping our clients navigate this continually evolving space with brand safety in mind.  This document details the measures we have put in place to promote a safe and transparent environment for our clients advertising messages.

Publicis Media utilises the following tools/technologies and processes intended to ensure ad misplacement is minimised when served through the technology platforms:

1. Independently verified pre-bid solutions, accredited in accordance with JICWEBS approved principles
2. Excluded inventory sources
3. Appropriate post-bid firewall solutions
4. Appropriate levels of transparency on URLs/Domains/Sub-Domains after delivery
5. Terms and conditions with technology and publisher vendors requiring the vendor to perform quality inventory assessments based on agency and client instructions (e.g., verify ad impressions, frequency caps, video specs, NHT, viewability, pre-bid tools)

Publicis Groupe Verified is our certification process that indicates a vendor has met what we believe to be industry leading standards for brand safety, consumer privacy and client data protection across display, video, mobile and social programmatic media.

As part of the Verified process, ad servers, data providers, DSPs, inventory partners and brand safety vendors undergo a multi-pronged evaluation and platform performance test (where applicable).  Upon successful verification, all verified vendors are assigned a Tier ranking (Tier 1-3) and receive the Verified seal for external use as well as a Verified scorecard that provides scores, commentary and competitive benchmarks across features (please note, brand safety vendors do not receive scores/tiers).

Publicis Groupe Verified gives us information that helps us to remove vapourware and un-grounded commitments to service and functionality within the technology space.  It verifies the original source of the data, the permissible uses and restrictions relating to such data, and the quality and safety of the inventory that we buy.  We have established stringent metrics and targets addressing performance and brand safety issues in the industry (e.g., non-human traffic guarantees, viewability requirements, etc.)

Data

Data providers are put through a legal and technical vetting process to assess capabilities, brand safety and privacy compliance.

Media

We endeavour to buy only non-blind, vetted URLs.  Before we even begin to buy from ad exchanges or supply sources we request that they deliver to us every URL that they hold within their business.

Our automated URL vetting is conducted by a third party technology called Grapeshot.  The combination of manual and automated processes is necessary as an automated solution cannot spot every offending URL.

Technology

Technology platform providers go through a 200-1200+ (depending on type of vendor) point inspection process (the Verification process mentioned above) and pilot testing to validate existence of mission-critical criteria and platform functionality.

The above process is applied to the applicable campaigns within Publicis Media unless agreed otherwise with the advertisers.

As part of its dedication to equipping Publicis Groupe agencies with safe, high performing, innovative technology solutions, Publicis Media continues to re-evaluate vendors even after the initial verification process is complete. This creates incentive for vendors to enhance and/or improve upon their offerings to meet the evolving demands of agencies and clients.

Publicis Groupe Verified Blacklists

There are a number of vendors available for blacklisting who will screen your site lists. We use Grapeshot as standard but always take a human view of the lists too, spot checking domains that have been passed as safe.

Vendors such as Google can also often provide lists of sites to blacklist.

Blacklisting is an effective tool for brand safety. It filters out the identified unsafe environments.  However, it does not protect brands from inappropriate content that appears within sites that are otherwise deemed safe.  A good example of this is large portal sites (YouTube, Yahoo, etc.) and news sites (Mail Online, The Guardian, etc.) where a story may be published that we would not want to appear against, in a domain that is otherwise considered safe.  Applying category blocking and negative keywords can help reduce this risk (see sections below).

Our process is designed to avoid ads being served to sites with high levels of volatile user generated content as well as sites that contain the following:

• Nudity
• Adult content
• Hate speech
• Violence
• Profanity
• Extreme content
• Illegal file sharing
• Illegal drugs
• Bizarre content
• Immaturity
• Click farms
• Parked domains

Any sites identified that contain any of the identified categories above will be  added to the blacklist.  Some of the above listed categories are subjective in nature (i.e, one client’s “bizarre content” is another client’s “target audience”), and we can work with clients to modify blacklists as needed.  

We contract with Grapeshot who screen site lists for identified unsafe sites.  In addition, our programmatic teams also add in sites that appear potentially fraudulent due to performance anomalies, misleading domain names, incorrect DSP categorization or show current negative news topics.

Furthermore, we work with the Federation Against Copyright Theft (FACT) and the London Metropolitan Police Intellectual Crimes Unit’s Infringing Website List (IWL) to supplement the blacklists we generate.

As standard practice we use our own Publicis Groupe Verified blacklist on campaigns.  The blacklists are available for all of our clients upon request.

Client Blacklists and Whitelists

In some cases where there are specific client or campaign requirements, we will take the manually curated list of sites which a client has deemed acceptable, and upload this to the programmatic platform and run campaigns against only those domains while still benefitting from pre-bid solutions.  Publicis Media has a global verified whitelist which client whitelists can be referenced against before use.

Clients can also provide Publicis Media with their own blacklist. This is particularly popular in the video and mobile space.

Whilst blacklisting and whitelisting is considered an effective tool for brand safety, by filtering out unsafe environments, it may not always filter out inappropriate content that appears within sites that are otherwise deemed safe.  Applying category blocking and negative keywords can help reduce this risk (see sections below).

Category blocking

Category blocking is available on all programmatic platforms used by Publicis Media.  This allows for brand safety protections to be set at an account or campaign level, generally against a number of potential categories (e.g., Violence, Drugs & Alcohol, Adult/Mature content, Politics, Language, Extremism, Illegal Streaming, etc.) in order to come up with customised rules for an individual brand (i.e. some brands may not mind being against bad language or content with Alcohol in it).

By default we apply the highest level of category blocking for campaigns.  This would only be changed on consultation and approval from a client.

Category blocking is a very effective tool for brand safety and generally will catch inappropriate content, however, it is only as effective as the technology categorising the pages where inventory is available. If a piece of content is newly published or written with misleading headlines or meta-tags, the technology may not be able to categorise it correctly, or may not yet have seen it before an ad is served.  This is particularly difficult with video content as most technology cannot ‘view’ the video to check if it is appropriate, and instead, relies on the headline and tags associated with it.

Negative keyword lists

Negative Keyword lists contain words associated with content we would not want a brand to appear against.  These are usually curated manually (or some platforms can provide a standard one) and applied within the platform against each campaign.

We use a standard list across campaigns and clients that contain pornographic keywords, explicit adult keywords and those associated with illegal activity.  This is applied as standard to campaigns and is updated on a regular basis.

Violent news list: this contains keywords related to topical news content that we would generally want to avoid and is updated regularly but also immediately after any large scale news event such as a terrorist attack. Some keywords are intended to be permanently on the list (Bombing, terror attack, ISIS, etc.). Sometimes keywords may be added to this that at other times would not be included, for instance after the Paris terror attacks in November 2015, terms such as Paris and Stade de France were added but later removed once this news story became less prominent.  This list is often client and vendor specific, and we would need to discuss its coverage/application regarding any given client, campaign or vendor.

Bespoke list: a client may also have a bespoke list they wish to apply specifically to their campaigns.

As with Category Blocking, negative keyword targeting can be very effective but is limited by the technology reading the URLs and categorising their content and is particularly limited with video.

Blocking tags

There are a number of third-party technology vendors who can apply a further level of brand safety protections. Integral Ad Science, and Double Verify are vendors we work with.

These technology vendors ‘wrap’ their tracking around the creative tags served by a client’s ad server before these are uploaded into the buying platform.

Their blocking tags ‘read’ the impression being bought and block the buying platform from displaying impressions in unsafe environments.

These vendors can also often block fraudulent impressions being served.  When an ad server has the ability to block a URL at the ad server level, we will ensure this is turned on for URL’s deemed inappropriate.  Of course, there are exceptions and limitations with every technological solution and we will discuss those with each client as applicable (e.g., some publishing partners do not allow the use of some of these technologies).

Pre-bid technology

Integral Ad Science, DoubleVerify and Moat, can also integrate with buying platforms at a pre-bid level.  Ensuring that we only bid on impressions that have previously been judged safe, non-fraudulent, and viewable through the historical data available to them from every bid request that they see.

These external vendors are independent of the platforms and publishers used.  They see the full volume of bid requests which allows them to accurately build a picture of the safe and unsafe inventory available.

If a client finds their advertisements in locations deemed inappropriate Publicis Media will take steps to ensure these URLs are removed as quickly as possible (e.g., we would request the publisher or technology vendor remove the offending ads within an hour of receiving notice). As this is a real-time programmatic buy the instant the URL is discovered and added to the Publicis Media blacklist all advertisements after this point will not be served on that site until it is removed from the blacklist.

The process of finding the URL within reports can generally take up to an hour but the application of the URL to the blacklist is instantaneous and ensures no further ads are placed on the site.  All take downs will be processed within a 24 hour time period.

The details of this process are outlined in Publicis Media’s Digital Trading Terms (available upon request) which we hold with all of our clients.

Publicis Media works with JICWEBS and the DTSG on the formulation of rules and regulations applied to the industry.  We also work with the IAB UK, IAB EMEA and IAB US.

Publicis Media has been awarded the DTSG Good Practice Principle Seal of Compliance following verification by independent industry auditor ABC in the UK since 2015.  The intention of the Principles is to significantly reduce the risk of the misplacement of display advertising on digital media properties, uphold brand safety and protect the integrity of digital advertising.  Compliance denotes that we have responded to the industry’s needs and are proactively demonstrating our support for industry agreed standards, we have opened up our business to an independent, industry auditor (ABC) and allowed them to see our business and the processes we follow, we abide by the six Principles which the industry has agreed are key to trading in a manner that protects brand reputation online and we are transparent about how ads will be delivered prior to delivery.

In addition, Publicis Media works with the TAG’s Domain Fraud Threat List in the US and is a registered TAG member company.  We have completed both the TAG “Certified Against Fraud” program and the TAG “Inventory Quality Guideline” requirements.

Publicis Media incorporates TAG’s standards and guidelines, and seeks to exceed TAG quality, brand safety and ad fraud requirements for inventory, SSP’s, DSP’s, ad servers, data providers and third party verification partners. As such, TAG has actively engaged with Publicis Media to support their efforts in strengthening current TAG guidelines via the brand safety standards available with Publicis Groupe Verified.

Through our memberships with industry bodies we continuously campaign for greater transparency of ad inventory, more stringent and rigorous processes to avoid fraud and an industry aligned approach to the removal of criminal and infringing websites.