Update: Enhanced Brand Suitability & Extended Language Support in Q1 2022

Open

Open

Welcome to the latest Verity product update from GumGum.

GumGum’s Verity is now delivering contextual classification and brand suitability solutions in 10+ languages! Plus, we’re accelerating our reach and scale, continuing to refine our Threat and Contextual models and working on key enhancements targeted for rollout in 2022. Keep an eye out for some of these upcoming changes.

In the News

GumGum recently acquired Playground XYZ, a global attention-based advertising platform!

Phil Schraeder, CEO, GumGum says “Our accredited contextual technology combined with Playground XYZ's proven attention measurement and optimization solution will allow advertisers to future proof their attribution strategies by moving past basic viewability metrics to better understand the true effectiveness of their digital campaigns.”

Read more about the acquisition here GumGum Acquires Attention Measurement And Optimization Platform Playground XYZ.

Upcoming Enhancements

In the upcoming month, Verity will add these enhancements:

Contextual and Brand Suitability solutions in 10+ languages

Verity already provides analysis for page and video content in the English, Japanese, French, German and Spanish languages.

We’ve recently extended our contextual and brand suitability reports to also analyse page and video content in:

• Portuguese

• Italian

• Dutch

• Danish

• Norwegian

• Swedish

Keep track of all the currently available languages across page, text, video, and audio in the Language Support Grid!

New Risk field in Threat category results

Verity’s Threat categories (GGThreat) are part of GumGum's custom brand safety taxonomy (e.g. GGT6 indicates content that contains Hate speech, harassment, and cyberbullying).  

To make it simpler for you to establish effective brand suitability strategies, we’re improving the data returned when Verity detects threats in page or video content. As of March 1, 2022, we’re adding a “risk” field to Verity threat category results. 

The new “risk” field returns one of these risk levels:

• HIGH

• MEDIUM

• LOW

These risk levels align closely with the risk levels defined in the GARM: Brand Safety Floor + Suitability Framework.

You can use the risk level data to dial in the right kind of content for specific brands or ad campaigns. For example, a childcare brand might have zero tolerance for any level of risky content in any category. Alternately, a pharmaceutical brand might be able to tolerate MEDIUM or HIGH-risk content in certain categories.

As a reminder, Verity’s Threat categories (GGThreat) results have returned a “confidence” score which previously has acted as a combination of confidence of presence and potential risk level. The “confidence” score will remain in the response, but should not be used with respect to risk level moving forward. 

GumGum recommends that partners reference the threat category risk levels (instead of the binary “safe”: “TRUE/FALSE” field) as this new release provides optimal levels of flexibility and control in managing brand suitability strategies and targeting. These risk levels closely align with the risk levels defined in the GARM: Brand Safety Floor + Suitability Framework. Partners can tailor a level of tolerance for risky content to a level appropriate for a specific brand or ad campaign.
Best practice recommendations might include:

• Always consider content unsafe where the threat risk level is HIGH

• For brands with some tolerance for risk, accept MEDIUM or LOW risk for some categories

• For more protection, always consider content unsafe where threat risk level is MEDIUM or HIGH

• For zero threat tolerance, always consider content unsafe where threat risk level is LOW, MEDIUM or HIGH

Change Notifications

Verity will be implementing the following changes in the upcoming months:

March 1, 2022 – New INSUFFICIENT_CONTENT status message

The Verity API will return an additional application error status message named INSUFFICIENT_CONTENT.

An INSUFFICIENT_CONTENT message is returned when Verity’s content extraction processes cannot extract sufficient relevant content from a page or video to adequately perform classification tasks across text and imagery.

The benefit of excluding insufficient content from Verity analysis is that classifications will only be made based on meaningful amounts of data, enabling increased accuracy across all classes. 

We'd love to hear from you! Hit us up in Slack or email verity-support@gumgum.com with any questions or feedback.