Horizon Media pilots AI to spot cultural shifts mid-campaign
Horizon Media is piloting an AI-driven audience and media capability from Sightly and Vurvey Labs that flags cultural shifts while campaigns are live - and adjusts media decisions in response. The integration pairs Vurvey Labs' real-time human sentiment (People Model) with Sightly's cultural and contextual signals (Brand Mentality) inside the HorizonOS partner ecosystem.
Dentsu's Carat is also testing the capability. The move speaks to a clear mandate across agencies: stop tinkering and prove outcomes with faster reads on risk, fit, and creative direction.
How it works
The tool creates custom audiences informed by two streams: always-on human insight from target consumers and cultural/contextual signals from news, social, and brand risk indicators. When sentiment shifts, the system surfaces conflicts and recommends changes to targeting, creative, and placements.
"The challenge for agencies isn't access to more data, it's knowing which signals are worth acting on," said John Koenigsberg, executive vice president and head of platform partnerships at Horizon Media. "This is exactly why we launched HorizonOS - to foster a community of emerging technology partners who can innovate through intentional, curated collaboration."
Why this matters for marketers
- Act before performance slumps: Detect sentiment changes early instead of waiting for lagging indicators like lower conversion or higher CPAs.
- Sharper suitability calls: Merge brand safety, cultural nuance, and audience sentiment to avoid tone-deaf moments without over-blocking.
- Creative clarity: Know which messages still land with your audience - and which need reframing or pausing.
- Operational speed: Reduce the manual effort required to monitor culture, interpret signals, and brief teams mid-flight.
Example in practice
An athletic apparel brand launches a "no excuses" message that usually performs. A major college football program then faces allegations of abusive conditioning culture. Vurvey's People Model picks up a shift in emotional posture around "toughness," while Sightly's Brand Mentality layers in news coverage and risk indicators, then flags the conflict to the agency.
From there, the team can pivot: pause certain lines, swap to "support and progress" creative, update exclusions, and re-allocate budget into safer contexts and audiences.
What to do now
- Define sensitive territories: Map themes, phrases, and contexts that could turn risky for your category. Align this list with brand suitability frameworks like the GARM Brand Safety and Suitability Framework.
- Set decision thresholds: Establish sentiment and context triggers with clear SLAs for pausing, swapping creative, and reallocating spend.
- Pre-build creative alternates: Have "on-deck" variants with different emotional tones and CTAs ready for rapid rotation.
- Wire it into buying: Connect the signal layer to activation platforms so changes flow to targeting, exclusions, and bids automatically where possible.
- Loop in comms and legal: Coordinate brand messaging shifts with PR and crisis protocols to keep external narratives consistent.
- Measure beyond media: Track message fit, brand favorability, and earned conversation alongside performance KPIs.
- Run controlled tests: Pilot on a subset of campaigns to compare outcomes versus business-as-usual optimization.
- Mind governance: Document data sources, bias checks, and privacy practices across sentiment and contextual feeds.
Who else is testing
Dentsu's Carat is also piloting the Sightly-Vurvey integration. Separately, Dentsu introduced Generative Audiences earlier this year, combining people-based data with AI signals to simulate audience groups.
Risks and limits
Analysts are split on the long-term value of proprietary AI stacks. Gartner forecasts that about half of agency-built AI platforms could wind down or be obsolete by 2029, and that open-source platforms will support over 75% of enterprise AI deployments on the client side by 2028. Translation: build for interoperability and portability from day one.
Bottom line
Cultural context changes fast. Pairing live human sentiment with contextual intelligence gives marketers a practical way to protect brand equity, keep creative relevant, and move budget with intent - before performance tells you something's off.
Want playbooks, tools, and training on this topic? Explore AI for Marketing. For comms teams coordinating message shifts and crisis response, see AI for PR & Communications.
Your membership also unlocks:
