Skip to main content
Lookalike signals let you create a new signal based on posts you’ve already found valuable. Instead of writing boolean queries and AI prompts from scratch, you provide example posts and Clearcue reverse-engineers the pattern.

When to use them

  • You found several posts that represent exactly the type of content you want to track, but you’re not sure how to express it as a search query
  • You want to find “more like this” without manually analyzing what the posts have in common
  • You want a faster way to set up an AI-enhanced signal with both a boolean query and a qualification prompt

How it works

1

Collect example posts

Find 3 to 5 posts that represent the pattern you want to detect. Copy their URLs.
2

Open the lookalike signal creator

In the signal creation flow, choose the lookalike option and paste your post URLs.
3

Clearcue analyzes the posts

Clearcue fetches each post, reads the content, and uses AI to identify what they have in common — topics, tone, intent patterns, and key phrases.
4

Review the generated signal

Clearcue produces two things:
  • A boolean search query that captures the common keywords and phrases across your example posts
  • A content qualification prompt that describes the pattern in natural language for AI to evaluate future posts against
You can edit both before saving.
5

Save and start monitoring

Once you’re happy with the query and prompt, save the signal. Clearcue starts scanning for new posts that match the pattern.

What you need

  • 3 to 5 post URLs — these must be public or accessible posts. The more examples you provide, the more accurately Clearcue can identify the common pattern.
  • Posts should share a common theme or intent. Mixing unrelated posts will produce a vague, low-quality signal.
Choose posts that are clearly representative of the pattern you want to detect. One off-topic post in your examples can skew the generated query and prompt.

What you get

OutputDescription
Boolean search queryA keyword query (e.g. ("AI" OR "automation") AND ("launched" OR "built")) generated from the common language in your example posts. This query runs continuously to find candidate posts.
Content qualification promptA natural-language description of the pattern (e.g. “Posts where founders announce launching an AI-powered product…”). AI uses this prompt to evaluate each candidate post and filter out false positives.
Both outputs are fully editable — treat them as a starting point and refine as needed.

Example

You notice three posts from founders announcing they’ve built an internal AI tool for their sales team. You want to find more posts like these.
  1. Copy the 3 post URLs
  2. Paste them into the lookalike signal creator
  3. Clearcue generates:
    • Query: ("built" OR "developed" OR "launched") AND ("AI" OR "machine learning") AND ("sales" OR "SDR" OR "outreach")
    • Prompt: “Qualify posts where a founder or technical leader announces building or launching an internal AI-powered tool specifically for sales or outreach. Include product launches, internal tool announcements, and ‘we built this’ stories. Exclude posts about third-party tools or general AI news.”
  4. Review, tweak if needed, and save
You now have a continuously running signal that finds similar announcements as they happen.

Handling failed URLs

If some of your post URLs can’t be fetched (e.g. the post was deleted or is private), Clearcue will still generate the signal as long as at least 2 posts are successfully retrieved. You’ll see which URLs failed and why.

Lookalike vs AI Signal

Both lookalike signals and AI Signals use AI to qualify content. The difference is how you create them:
Lookalike SignalAI Signal
Starting pointExample posts you’ve already foundA blank prompt you write yourself
Boolean queryAuto-generated from your examplesYou write it manually (or use a keyword search)
AI promptAuto-generated from your examplesYou write it manually
Best for”Find more like this""Find exactly this specific pattern”
You can always start with a lookalike signal and then fine-tune the generated query and prompt — combining the speed of auto-generation with the precision of manual control.