AI Social Listening Dashboard — Know What Your Market Is Saying Before Your Competitors Do
Built by CircuitCore ·
About this build
A Claude and Perplexity powered workflow that monitors Reddit, Twitter/X, LinkedIn and niche forums daily for mentions of your brand, competitors and target keywords — classifies every mention by sentiment and urgency, filters out the noise and delivers a structured morning briefing with the mentions that actually matt
What was built
Built this after missing a Reddit thread where 40 people were discussing the product and most of the conversation was based on outdated information. By the time someone forwarded it the thread had run its course, the community had formed their opinion and the moment to correct the record had completely passed. The workflow runs every morning at 6am. Here is exactly what it does. Step 1 — Daily search sweep via Perplexity n8n triggers a series of Perplexity searches across four source categories every morning. Brand mentions — product name, company name and founder name searched across Reddit, Twitter/X, Hacker News, LinkedIn and relevant niche forums from the last 24 hours. Competitor mentions — the same search pattern applied to each of the top 3 competitors. Catches conversations where someone is directly comparing or switching. Industry keyword mentions — 5 to 8 keywords tied to the product category to catch broader conversations where the brand is not mentioned but the opportunity to add value exists. Crisis signals — brand name combined with words like scam, broken, cancelled, avoid, refund and security to catch emerging reputation problems before they compound. Step 2 — Signal extraction and classification via Claude Every raw Perplexity result gets passed to Claude which does four things simultaneously. Filters irrelevant noise — mentions clearly unrelated to the business context get removed before anything else. Classifies each remaining mention as positive, negative or neutral with a confidence score. Identifies emerging themes across multiple mentions rather than treating each one in isolation. A single negative comment is noise. Three negative comments about the same issue in the same week is a signal. Flags mentions that likely require a direct response — unanswered customer questions, complaints with engagement, or conversations where joining in would be genuinely useful. Step 3 — Trend logging in Airtable Each day’s top themes and sentiment scores get written to Airtable. This builds a running record so you can see whether sentiment is improving or declining week over week, whether a particular platform is generating disproportionate negative mentions and whether competitor mentions are accelerating in a way that suggests a new campaign or product launch. Step 4 — Morning briefing delivery By 8am a structured briefing lands in Gmail or Slack containing five sections. Overall sentiment summary for the last 24 hours. The 3 most important mentions that need attention today. Any emerging trend worth watching. One competitor insight from the previous day. A suggested action based on what was found. The whole thing runs without touching it. The market intelligence is just there when the day starts. What took the longest to get right: The noise filtering was the hardest part. Early versions were surfacing completely irrelevant mentions because the brand name shared words with unrelated topics. Fixed by adding context qualifiers to every search query — always combining the brand name with the product category and at least one additional identifier. The briefing length went through 4 iterations. The first version was too long — nobody reads a 600 word daily briefing consistently. The final format has 5 sections each under 3 sentences. Short enough to read in 90 seconds, specific enough to know immediately whether anything requires action. Results after 10 weeks: Caught a Reddit thread where a user was describing a bug as a scam and had already gathered 28 upvotes. Responded within 20 minutes. The thread ended with the user updating their post to say the issue was resolved and the response was fast. Without this build that thread would have been found the following week. Caught a positive mention on Hacker News where someone listed the product in a comment getting significant engagement — was able to join the thread authentically and add context that generated 3 new signups directly. The weekly Airtable data showed Reddit was generating 3 times more negative mentions than Twitter/X despite Twitter having higher total mention volume. That single insight completely changed where community engagement time was being invested. Will share the full n8n workflow JSON and Claude classification prompt in comments for anyone who wants to deploy this.