Automate Share of Voice Tracking in AI Search for Insurance Carriers

Insurance marketing teams rely on tools like Semrush and Ahrefs for competitive intelligence. These platforms are excellent for tracking organic keyword rankings and backlinks. Their failure mode in the age of AI search is that they are architected to see a world of ten blue links, not a single, synthesized paragraph. They can tell you if you rank #1 for "term life insurance quotes," but not if Google's SGE answer quoted a competitor's rate information directly above your link.

Consider an insurance carrier trying to measure visibility for "best home insurance for hurricane-prone areas." A marketing analyst manually searches this query on three different AI engines, copies the text, and notes which of their five key competitors were mentioned. This process takes at least 15 minutes per query. To track 50 core keywords weekly, the analyst spends over 12 hours on error-prone data entry, creating a spreadsheet that is outdated the moment it is finished.

Brand monitoring tools like Mention are not the solution either. They track brand name occurrences across the web but lack the essential context of the user's search query. They cannot differentiate a mention for "auto insurance claims process" from one for "commercial fleet insurance." Without the search intent, the data is just noise.

The structural problem is that these off-the-shelf tools are built to parse structured HTML, not to understand the semantic content of a large language model's output. They cannot deconstruct a generated answer, attribute claims to their sources, and map that visibility back to the specific query that triggered it. This leaves insurance carriers blind to how their brand is actually being represented in the new front page of search.

The first step is to define the keyword universe and competitive set. We would work with your marketing team to identify the 50 to 200 non-branded queries that drive high-intent customers for your specific insurance products. We then create a definitive list of your brand and up to 10 competitors to monitor, establishing the baseline for the tracking system.

Our approach adapts the technology from our own four-stage AEO pipeline. A scheduled job, running on GitHub Actions, triggers a Python script that queries AI search engine APIs, like the Brave Search API, for each target keyword. The raw text from the AI-generated answer is captured. We then use the Gemini Pro API to parse this unstructured text, identify every mentioned insurance carrier, and write the structured results—query, carrier, and context—to a Supabase database.

The delivered system is a real-time dashboard that you own completely. It visualizes your share of voice percentage for each keyword against your competitors, updated every 24 hours. You can spot trends instantly. For instance, you could see your SOV for "small business liability insurance" drop from 30% to 5% a week after a competitor publishes a new, highly-cited article, giving your content team an immediate, data-driven signal to respond.