The place assistants reach for when a question sounds like asking a person rather than reading a brochure. Comparison and recommendation prompts pull from here constantly.
Social Insights
Find out which social channel is teaching AI about you
Assistants cite Reddit threads, LinkedIn posts and YouTube videos when they answer questions in your category. Social Insights separates your AI visibility by channel, shows how each one trends against competitors, and names the conversations you are missing from.
The blind spot
Your site is not the only thing the models read
Most AI visibility tools show you citations as one list. That hides the pattern that actually matters, because a citation from your own pricing page and a citation from a Reddit thread call for completely different work.
When a buyer asks an assistant which tool to pick, the sources behind that answer are frequently discussion rather than documentation. A three-year-old Reddit comment can carry more weight than the page you shipped last week. You can hold every keyword ranking in your category and still lose the answer, because the model was never reading your site to begin with.
Social Insights splits your citations by the channel they came from, so you can see which platform is doing the talking, whether it is talking about you or about a competitor, and which direction that has moved over the last 90 days.
Channels
Three channels, tracked separately
Each one behaves differently in AI answers, so each one gets its own trend line, visibility score, top citations and gap list.
Where B2B credibility gets built. Posts, company pages and long-form articles show up behind answers about vendors, categories and who is credible in a market.
YouTube
Demos, reviews and walkthroughs. Assistants cite video pages when the question is how something works or whether it is worth buying.
Filter any channel by engine and time window. Perplexity leaning on Reddit while ChatGPT leans on LinkedIn is a real and common pattern, and it changes which platform you invest in first.
What you get
Four views per channel
Each answers a different question, and each one leads somewhere you can act.
Citation trend
Your citations from this channel over time, plotted against your tracked competitors. Shows whether you are gaining ground on a platform or quietly losing it.
Channel visibility score
The share of your tracked prompts where AI answers cited you from this channel, scored out of 100, with a bar per competitor next to it.
Top citations and core anchors
The specific URLs models reference most, with citation counts, share of the channel, the engine driving each one, and a flag on the assets that carry disproportionate weight.
Citation gap radar
Prompts where competitors were cited from social and you were not, named individually, each linked to building presence for that conversation.
On your dashboard
The headline sits on your diagnosis page
You do not have to go looking for it. Three cards on the main diagnosis view carry the summary and link into the detail.
- 1
Social citation share
Total social citations with week-on-week movement, split by channel, and which platform is currently your biggest citation driver.
- 2
Top AI anchor asset
The single social URL models cite most about you, with its channel, its citation count, and whether it qualifies as a core anchor.
- 3
Social citation gaps
How many high-value prompts have competitors cited from social, the share of your tracked prompts affected, and a risk band.
Not social listening
Mentions and citations are different numbers
Both are useful. Only one of them tells you what an assistant will say about you tomorrow.
Social listening counts mentions
How often your name appears on a platform, and how people felt about it. Measures conversation volume, weighted towards whatever is loud right now.
Social Insights counts citations
How often AI engines reach for that platform when answering your prompts, and which exact URLs they reach for. A quiet thread three models keep quoting outranks a viral post no assistant has ever referenced.
Where it goes next
Every gap has somewhere to go
A gap on the radar links straight into building presence for that prompt. A citation that is working against you hands off to the off-page reputation workflow, where the thread can be tracked and answered at the source. An anchor asset that is winning tells you what to make more of.
The measurement is worth having on its own. What makes it worth checking weekly is that each row ends somewhere other than a chart.
Frequently Asked Questions
Social Insights, answered
A breakdown of your AI visibility by the social channel that fed it. Targetlytics reads the sources AI engines cite when answering your tracked prompts, separates out the ones that come from LinkedIn, Reddit and YouTube, and shows how each channel is trending against your competitors. It answers a question a site-only view cannot: which social platform is teaching the models about your category.
Because answer engines lean on discussion, not just documentation. When someone asks an assistant for a recommendation, the sources behind that answer are often a Reddit thread, a LinkedIn post or a YouTube video rather than a vendor page. If competitors own those threads and you do not, you lose the answer without ever losing a ranking.
LinkedIn, Reddit and YouTube today. You can filter each of them by AI engine, so you can see whether Perplexity is pulling from Reddit while ChatGPT leans on LinkedIn, and by time window across the last 7, 30 or 90 days.
A single social URL that AI engines cite repeatedly when answering questions about your category. It is the piece carrying disproportionate weight in what the models believe about you. Knowing which one it is tells you what to protect, what to keep current, and what shape to copy when you build the next one.
A list of your tracked prompts where AI engines cited a competitor from a social source and did not cite you. Each row names the prompt and the competitors that were cited, then links straight to building presence for it. These are the specific conversations you are absent from, not a generic share-of-voice number.
Social listening counts mentions on the platform. This counts citations from the platform inside AI answers. A thread with modest engagement that three models keep quoting matters more here than a viral post no assistant has ever referenced, and only the citation view can tell those two apart.
No. The measurement works from the sources AI engines cite when they answer your tracked prompts, so it reads public content and needs no account access, posting permission or API token from any social platform.
See which channel is shaping your AI answers
Start with a free visibility audit. It shows where AI engines are sourcing answers about your category, before you connect anything.
Get a free visibility audit
