L2 integrates PharosGraph narrative intelligence for programmatic voter targeting on CTV
Voter data platform L2 has integrated PharosGraph's narrative intelligence technology to enable political campaigns to target voters based on issue salience. The integration allows campaigns to execute programmatic ad buys across CTV and digital channels through existing DSP connections with Yahoo, LiveRamp, and The Trade Desk.
Key Takeaways
- Integration enables programmatic ad buys via Yahoo, LiveRamp, and The Trade Desk using L2 voter files.
- PharosGraph technology tracks and ranks how specific issues and candidates resonate with voters over time.
- Campaigns can now optimize creative assets based on real-time trending voter sentiment rather than just party affiliation.
- The deal follows a previous integration with AdImpact to track ROI based on issue-salient spending.
Why It Matters
This integration moves political advertising beyond static demographic data toward dynamic, issue-based activation. By linking narrative intelligence directly to major DSPs like The Trade Desk and Yahoo, campaigns can adjust CTV and digital spend in real-time as specific topics gain or lose traction with the electorate. This capability reflects a broader trend in the streaming ecosystem where granular sentiment analysis is becoming a prerequisite for high-stakes addressable advertising. As the political cycle intensifies, watch for how these salience-based models impact the allocation of programmatic budgets between traditional digital display and premium CTV inventory.
Additional Context
L2's integration of PharosGraph narrative intelligence arrives amid a broader wave of political ad-tech partnerships targeting connected TV inventory. In March 2026, The Trade Desk launched its OpenPath political advertising initiative, giving campaigns direct access to premium CTV supply without intermediary resellers, a move that directly benefits platforms like L2 that already route buys through The Trade Desk's DSP. The political CTV market has drawn significant investment, with AdImpact reporting that political advertisers spent $1.8 billion on connected TV during the 2024 cycle, a figure that analysts expect to grow substantially heading into the 2026 midterms.
On the data and identity side, LiveRamp has been positioning its RampID solution as the connective tissue for political addressability across fragmented CTV environments. In May 2026, LiveRamp announced a partnership with Yahoo to enable political advertisers to activate first-party voter data across Yahoo's DSP, creating a direct competitor to the L2-PharosGraph pipeline for campaigns seeking programmatic political buys. Yahoo itself has been aggressive in courting political spend, with the platform reporting a 40% year-over-year increase in political ad bookings for Q1 2026 compared to the same period in 2024. These moves underscore that the political programmatic ecosystem is consolidating around a small number of identity and activation layers.
From a technical standpoint, narrative intelligence models like PharosGraph's represent a shift from traditional voter segmentation toward real-time sentiment tracking. PharosGraph raised a $12 million Series A in late 2025 to scale its natural language processing models for political and advocacy use cases, with the company claiming its models can detect issue salience shifts within 48 hours of a triggering event. Mathieu Trepanier, who leads L2's product strategy, has framed the integration as a response to campaigns demanding faster feedback loops between messaging tests and media execution. The approach mirrors what commercial brands have adopted through tools like Brandwatch and Sprinklr, but applying similar narrative-tracking rigor to the compressed timelines of political campaigns where a single debate or policy announcement can reshape the media landscape overnight.
Read full article at mediapost.com
Enjoy our coverage?
Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.
Add as preferred source