Media and publishing is at a crossroads. AI is upending everything. SEO isn’t driving traffic back to websites. Perplexity is stealing clicks. Publishers need to find ways to share their content and monetize it. Fast. That was the catalyst for Code and Theory to build and launch ContextLens, an AI-powered system that creates a clear, compelling content experience rooted in data and simultaneously functions as a custom ad unit to reach audiences when they're most engaged.
The first publisher to use ContextLens was RealClearPolling, which has been the arbiter of election polling and data for over two decades. In anticipation of the 2024 election cycle, RealClearPolitics (RCP) sought to enhance user engagement on its RealClearPolling platform. The primary objective was to develop an innovative tool that could simplify complex polling data, provide real-time insights and create new revenue streams through advanced advertising opportunities.
Enter ContextLens.
Recognizing that readers often feel overwhelmed by dense polling data, RealClearPolitics identified a need for a solution that transforms raw numbers into accessible, compelling narratives. The strategy powering ContextLens leveraged artificial intelligence to anticipate user queries and deliver dynamic, context-rich visualizations. This approach not only enhances the user experience, but also offers businesses a platform for precise, contextually relevant advertising, opening new B2B monetization avenues.
ContextLens is a first-of-its-kind AI-driven anticipatory design system.
"ContextLens" is our umbrella term for two novel technologies in partnership: One is an AI agent with “read” access to a website’s back-end databases, and the other is a front-end mechanism that can inject components directly into a webpage. These combine into a tool that monitors which components are currently in the users’ viewport and can ask and answer questions. Together, it allows users to create their own homepage while obviating the need for any other page, but the one ContextLens is on.
ContextLens empowers people to ask for any other information the website offers via the AI agent hooked into the back end. No more click-click-clicking to get to what you need. You just ask for it, and it comes to you.
In the lead-up to the 2024 elections, amid a landscape riddled with misinformation and cyberthreats, ContextLens positioned RealClearPolitics as a leader in accurate, real-time, contextual data storytelling.
When infused within the RealClearPolling experience, ContextLens proactively analyzed the content a user is viewing, predicted subsequent questions and generated bespoke visual representations directly within the interface. Curious how the county-by-county polling in Wisconsin is different from the last election? ContextLens knows. Want to see how that compares to other swing states? How about a deep dive into favorability changes between the two Presidential candidates every month leading up the election? You ask and ContextLens delivers; transforming static data into interactive storytelling.
ContextLens appears as a subtle interface element, inviting users to engage with dynamic visualizations without navigating away from the page. This approach ensures an uninterrupted, enriched user experience.
Built with cutting-edge LLMs, RAG architectures and React Agentic Design, ContextLens translates user intent into database queries first and then translates those queries into bespoke visualizations.
Since ContextLens launched around the 2024 election, a bevy of publishers (15+) have inquired about how they can integrate it. ContextLens proved its ability to rearticulate advertising from a disruptive necessity into a value-added service, alignING the interests of users, advertisers and publishers, creating a harmonious ecosystem where everyone benefits.
On Election Day 2024, RealClearPolitics saw a series of record numbers connected to ContextLens:
As RealClearPolitics co-founder Tom Bevan put it: “This marks a bold step in how people interact with complex information on the web.”
