With social media algorithms constantly changing, it’s time for news publishers to regain control of their audiences. Using AI and machine learning, publishers can program contextually relevant and personalized content down to the viewer level on their owned-and-operated sites for both breaking news and day-to-day operations. IRIS.TV and Gannett demonstrate how AI is helping to provide prescriptive insights into audience engagement and fueling future content strategies and video distribution on owned-and-operated sites.
Speakers:
Field Garthwaite, Co-Founder and CEO, IRIS.TV
Kara Chiles, Senior Director of Product Management, Gannett/USA TODAY NETWORK
Case Study - USA TODAY Networks: How A.I. is Revolutionizing News Publishing
1. How AI is Revolutionizing Publishing
Streaming Media East
Field Garthwaite, CEO & Co-Founder, IRIS.TV
Kara Chiles, Senior Director, Product Management, Gannett
2. Kara Chiles
Senior Director, Product Management
Gannett
KChiles@gannett.com
Field Garthwaite
CEO& Co-Founder
IRIS.TV
field@iris.tv
3. Overview
About Gannett: Video Product Goals
About IRIS.TV: Video Personalization and Programing
IRIS.TV Integration
Case Study: USA TODAY Network – Winter Olympics 2018
Next Steps: Election 2018
4. Gannett’s Video Product Goals
Product Health
• USA TODAY Network video creation and consumption continues to grow
and diversify
• Adoption of our proprietary player for system optimization
• Evolving solutions beyond autoplay with sound
• Continuing growth of video audience on mobile web
5. Goals with Personalization and Programming
• Test & Learn environment
• Enhanced user experiences across web and mobile
• Testing a new recommendation engine to validate lift
• Continue to grow and expand franchises
• Leverage national footprint to surface content from across the country
6.
7. IRIS.TV Integration
Step 1: Metadata Ingestion and Taxonomy Creation
Step 2: Video Personalization Player Plugin
Step 3: Setting up Programming Rules
Step 4: Launch!
8. Step 1: Asset IQ™
1. Content metadata from your CMSingested into Asset IQ™
2. Metadata cleanedand enriched
3. Videoassets mappedinto
clients unique taxonomy
4. Videoavailablefor
recommendation to your
audience
9. Step 2: Adaptive Stream™
1. Integrated video player plugin
2. Customize video player UI and events
11. Step 3: Setting up Programming Rules
Goal Highest video engagement while maintaining standards
Challenge All Gannett sites have access to all video content and third-
parties
Solution Whitelisting
• Ensure only local content and USA TODAY is
recommendable
• On USA TODAY, only content published on the site is
recommendable
12. Step 3: Setting up Programming Rules
Blacklisting
• No recommending videos in a designated category
Category Lock-Off
• Only recommending videos that match the category of the initial video
Recency Weight
• Recommend more recently published videos
13. Case Study: USA TODAY – Winter Olympics 2018
Major Tentpole Event
Opportunity:
• Global event with high-volume visual stories
• Increased network-wide unique traffic
• Reach casual and loyal audiences
• Heightened monetization potential
14. Case Study: USA TODAY – Winter Olympics 2018
Coordinating Complexity:
Email from the Halfpipe at 5AM
Production
• Daily original content from South Korea
• Local outlets produced hometown stories on
athletes
• Third-party content incorporated
15. Case Study: USA TODAY – Winter Olympics 2018
Programming
• Hundreds of videos and billions of playlist combinations
• Maintaining editorial standards and ensuring consistent integration at the article level
• Integrating new video templates to increase recirculation without duplicating
Monetization
• Maximizing video revenue without sacrificing user experience
16. Case Study: USA TODAY – Winter Olympics 2018
Production and Programming
● IRIS.TV is able to make most relevant
recommendations in real-time
● USA TODAY customized personalization
with programming rules
● Data-driven decision making
○ Asset and category-level performance
data to informed by content strategy
AugmentedIntelligence -CombiningA.I.withhumanexpertisetodrivevalue
17. Case Study: USA TODAY – Winter Olympics 2018
Monetization Priority
● Sponsored Pre-roll: The Farmers
Insurance campaign
● Six-second, autoplay sound-on spot
throttled for duration of exposure.
● Category lock-off to prioritize content
associated with campaign
18. Case Study: USA TODAY – Winter Olympics 2018
Record-setting returns:
● Over 50% video lift
● 13% increase in user in-stream retention for
Olympic vs regular content
19. Case Study: USA TODAY – Winter Olympics 2018
USA TODAY was able to achieve wall-to-wall coverage of the Olympics
• Product team can create engaging user experiences
• Editorial is able to focus on publishing high quality content
• Revenue teams are able to prioritize content with the best monetization
Owning your audience data informs smarter decisions
20. Next Steps: Mid-Terms 2018 and Beyond
Maximizing the strength of the
network
• Telling local election stories with
immediacy (live video, resources on
the ground)
• Telling a national story with layers
(primaries, debates, candidates,
voters, issues, election night)
Editor's Notes
Product Health:
USAT Network video consumption continues to grow (+208% YoY increase for Feb 2018) , particularly on local and social.
We’re in the final phase of full adoption of our proprietary Universal Web homegrown player.
Autoplay with sound has been a challenge for us due to negative user feedback thus we will be implementing changes that will have a positive impact on the user experience.
When it comes to on platform, mobile web has seen the most growth (+57% YoY increase for Feb 2018).
Test & Learn environment to extend experience for consumers and advertisers, focus on product optimization
Enhanced user experiences across web and mobile to increase franchise growth and focus on true-intent video consumers
Upcoming Highlights:
Testing a new recommendation engine to validate lift in continuous play and determine wider adoption
Continue to grow and expand franchises based on proven successes
Leverage our national footprint to surface compelling content from across the country
Goal: Get the highest video engagement while maintaining editorial standards
Challenge: All Gannett sites have access to video content from the entire network as well as third parties.
Solution: Whitelisting
For each newspaper, ensure only original local content and national content from USA TODAY is available for recommendations.
On USA TODAY, only content published on the site is eligible to be recommended
Global event with high-volume visual storytelling
Increased opportunity for network-wide unique traffic
Opportunity to reach casual and loyal audiences
High monetization potential
Production:
USA TODAY’s video team produced volumes of original content daily from South Korea
Local outlets produced hometown stories on athletes
Third-party content incorporated for breaking news
Programming:
Hundreds of original videos and billions of playlists combinations
Maintaining editorial standards and ensuring consistent integration at the article level for maximum reach
Integrating new video templates to increase recirculation without duplicating
Monetization:
Maximizing video revenue without sacrificing user experience
Production:
USA TODAY’s video team produced volumes of original content daily from South Korea
Local outlets produced hometown stories on athletes
Third-party content incorporated for breaking news
Programming:
Hundreds of original videos and billions of playlists combinations
Maintaining editorial standards and ensuring consistent integration at the article level for maximum reach
Integrating new video templates to increase recirculation without duplicating
Monetization:
Maximizing video revenue without sacrificing user experience
Production and Programming:
IRIS.TV is able to make most relevant recommendations based on real-time behavioral data
USA TODAY customized IRIS.TV A.I. with programming rules across desktop and mobile
Data-Driven decision making
Asset and category-level performance data to inform content strategy
Programming Rules and Data Insights
USA Today used IRIS.TV data to understand the value of each asset with respect to length and category.
For example, the team was able to gauge higher bounce rates on specific sports videos.
We were able to identify a specific length for their original sports videos. EX. Sports Pulse, versus third party content.
The team shorten those assets and saw increased retention over time.
Note: The team had a preroll campaign from Farmers that was targeting to all Olympics content(articles/videos). We assisted in fulling this campaign by the creation of the category lockoffs, ensure only Olympics content appeared on these pages.
Record-setting returns:
>50% video lift
13% increase in user in-stream retention for Olympic coverage vs regular content
USA TODAY was able to achieve wall-to-wall coverage of the Olympics
Product team can create engaging user experiences that balances automation with oversight
Editorial is able to focus on publishing high quality content
Revenue teams are able to prioritize content with the best monetization
Owning your audience data informs smarter decisions
Record-setting returns:
>50% video lift
13% increase in user in-stream retention for Olympic coverage vs regular content
USA TODAY was able to achieve wall-to-wall coverage of the Olympics
Product team can create engaging user experiences that balances automation with oversight
Editorial is able to focus on publishing high quality content
Revenue teams are able to prioritize content with the best monetization
Owning your audience data informs smarter decisions