Building an AI Search Visibility Intelligence Platform
DSC built an analytics platform that shows a communications agency how its clients appear in ChatGPT, Google AI Overview, and Perplexity. It brings brand visibility, cited sources, and competitor performance into one reporting view.
The Challenge
A communications and public affairs agency needed to measure how AI search platforms represented its client. The agency had raw prompt, response, and citation data from Scrunch, but no reliable way to turn it into client-ready reporting. Its team could not readily see how often the brand appeared in AI answers, which sources were cited, or how it compared with competitors.
Manual analysis would also be difficult to repeat for new clients. The agency needed a consistent set of measures and a reporting system it could reuse.
Key Pain Points
No standard measure of organic brand visibility across AI platforms
Limited insight into the domains and URLs cited in AI answers
No consistent benchmark against more than ten competitors
Raw data and manual reporting that were difficult to scale
Business Goal
Measure Organic Visibility
Track brand appearances without inflating results with prompts that explicitly name the client.
Understand Citation Influence
Identify the owned, competitor, media, social, and other sources shaping AI-generated answers.
Benchmark Competitors
Compare visibility and citation performance across ChatGPT, Google AI Overview, and Perplexity.
Scale Client Reporting
Automate monthly updates and make the platform reusable for additional agency clients.
Tech Stack
What We Did
Approach
DSC connected Scrunch to Snowflake through Airbyte, bringing in prompts, responses, brand citations, and brand-presence records on a schedule. More than 14 dbt models clean and organize the data, classify cited domains, and calculate the measures used in the dashboards.
The reporting rules separate organic visibility from prompts that mention the brand directly. Individual citations are counted, visibility and citation rankings remain distinct, and competitor filters do not change the client's own KPIs. A governed classification system groups citation sources into Company Owned, Competitor, Earned Media, Social Media, UGC, and Other.
DSC built six Tableau views covering executive KPIs, visibility trends, citations, prompts, website performance, and social media. The models and dashboards were designed so the agency could configure the same platform for another client.
Outcomes
95% Automated Monthly Reporting
Scheduled ingestion and dashboards replaced most manual data processing in the monthly reporting cycle.
Competitive Visibility Across Three Platforms
The agency can benchmark the client against more than ten competitors. The initial analysis placed the client eighth in visibility.
Actionable Citation Insights
The analysis found approximately 34% owned citation share for the client website, with its /blog path accounting for 42% of owned citations.
Faster Onboarding for New Clients
The initial build required more than 200 hours. DSC estimated 22 to 38 hours, or one to two weeks, to configure the platform for another client.
