AI Signals

AI Signals let you define a completely custom detection pattern using natural language. Instead of choosing from pre-built logic, you describe exactly what content you want to find — and AI evaluates every post or comment against your criteria.

​

When to use it

Use an AI Signal when none of the pre-built templates match what you’re looking for, or when you need highly specific detection logic. Common use cases:

• Tracking a niche topic that doesn’t fit any standard category
• Combining multiple intent signals into a single custom pattern
• Filtering content based on tone, sentiment, or context that keywords alone can’t capture

​

How to set it up

​

Write an effective prompt

Your qualification prompt tells the AI what to look for. The more specific you are, the better the results.

​

Include and exclude criteria

The most effective prompts have two parts: what to look for and what to ignore. This dramatically reduces false positives. Good prompt:

Qualify posts where a founder or CEO announces they are personally taking over a sales or GTM function that was previously handled by a team member or agency. Include posts expressing frustration with previous sales performance. Exclude general hiring posts or posts about delegating sales to someone else.

Weak prompt:

Find posts about sales.

​

Be specific about the subject

Tell the AI who the relevant person or company is in the context of the signal. Good prompt:

…When qualified, the relevant subject is the company receiving the funding.

This helps Clearcue correctly attribute the signal to the right person or company in your pipeline.

​

Use concrete examples

If certain phrases or patterns are strong indicators, mention them explicitly. Good prompt:

Include personal announcements (“Excited to share I’ve joined…”), company announcements (“Welcome our new…”), and congratulatory posts that confirm the appointment.

​

Define hard disqualifiers

List specific conditions that should always result in rejection, even if the post otherwise looks relevant. Good prompt:

Hard disqualifiers — immediately reject if any apply:

• The author is promoting their own agency or freelance practice
• The post is a guide, ranking, or listicle about choosing providers
• The post is a job listing for a permanent, in-house position

​

How AI qualification works

When you add an AI prompt to any signal, Clearcue runs a two-stage detection process:

1. Keyword matching — Your boolean search query or profile monitoring finds candidate posts
2. AI evaluation — Each candidate post is evaluated against your qualification prompt. Only posts that pass both stages are surfaced as signals.

This two-stage approach keeps detection fast (keyword search narrows the field) and accurate (AI filters out false positives).

​

Content vs signal qualification

There are two types of AI prompts: You can use one or both depending on your signal setup.