How to run a deep audience analysis on your products

Once your products are enriched, the Knowledge Vault can tell you something most ad tools charge agencies to find out: who is actually buying products like yours, and why.

That is what Deep Audience Analysis does. It runs as part of a social signals scan on a product, and it turns raw mentions into a readable picture of your buyer.

The short version:

• Open a product in the Knowledge Vault


• Scroll to Social Buzz and click Choose Platforms & Scan


• Enable Deep Audience Analysis when you scan (the scan costs 5 credits; deep analysis adds 15)


• Read the results in Audience Intelligence and Feature-Level Sentiment on the same page

Before you scan: add your competitors

The first time you scan a product, MerishAI asks you to add at least one competitor - a brand or store name, an optional website, and optional notes like "stronger in fragrance, weak in D2C".

This is not busywork. The AI watches for mentions of those competitors alongside your product, so the same scan also powers the Competitive Landscape section later. Add the 2-3 brands your customers actually compare you to.

Run the scan with Deep Audience Analysis on

From the product page, scroll to Social Buzz and click Choose Platforms & Scan. Pick your platforms, and make sure Deep Audience Analysis is enabled before you start.

The scan is product-specific - it searches for content about this one product, not your whole store, which keeps results focused. The dialog shows the exact cost before you start: 5 credits for the scan, +15 for deep analysis (20 total).

The scan searches social platforms for videos, reviews, and discussions about your product (a typical scan finds dozens of videos and stores the strongest as signals). With deep analysis on, it goes a step further: instead of just collecting what people said, it works out who they are.

Already scanned a product without it? You do not need to scan again - click Deep Analysis next to Scan Again and it runs on your existing signals.

What you get: Audience Intelligence

Once the scan completes, the Audience Intelligence section fills in:

• Age range, income level, gender split - the demographic core of your buyers
• Platform fit score - which platform your audience actually converts on, so you know where your video ads belong
• Purchase intent - how close to buying the conversation is
• Top purchase barriers - the objections holding people back (price, trust, availability)
• Where they shop and geographic locations - the channels and regions that matter

Below it, How to Win This Audience turns the data into concrete angles you can use in your next video.

What you get: Feature-Level Sentiment

The same deep scan also fills Feature-Level Sentiment: how people feel about specific product features, with mention counts and sentiment scores.

You will see your Top Loved Features - the things to lead your ads with - and Top Criticized Features, the objections to answer before the comments section does it for you. Market-specific factors like price and value are flagged separately, since they depend on local pricing and availability.

Put it to work

Audience data is only useful if it changes what you make. Inside MerishAI it feeds directly into the Creative Director: when you generate a video storyboard from the same product page, the audience profile shapes the hooks, the tone, and the platform format. An audience that skews 25-34 and lives on Instagram gets a very different film than one that buys on YouTube.

Scan first, then storyboard. The film comes out aimed at a person, not a demographic guess.

Next in the series: how to read everything a social scan collects - signals, creators, and competitor mentions - in the Social Buzz report.