Skip to main content
Every signal starts with a behaviour — something a person or company does that tells you they’re worth talking to. Pick one from the library, tell Clearcue what to watch, and choose how it’s detected.

How a signal is put together

Each behaviour in the library offers one or more sources — different ways of catching the same behaviour. You can enable any combination. Most behaviours can also be given social research and web research, which work through your tracked companies one by one instead of waiting for something to surface.

Research

What research adds to a signal, which behaviours support it, and how to configure it.
You don’t have to pick from the library. Describe the behaviour in your own words and Cue AI sets up the signal for you — it picks the behaviour, drafts the configuration, and shows it to you before anything is created.

The library at a glance


Engaging with your brand

Detects: People warming up to your own brand and content. Sources: Social Configure: Add your own profiles and company page. Enable the options you want.
Views your team’s profiles and Connects with your team need the LinkedIn login for that profile, because only its owner can see who viewed or connected. Clearcue walks you through connecting it after you save.
You can also import connections from a LinkedIn export. Request your data from LinkedIn, then upload the Connections.csv file to the signal. It brings your existing network in one go, and you can upload a fresher export whenever you like. Importing is also the way to use this signal without connecting a LinkedIn login at all. Your network is then exactly as current as your last upload. Example: Add your CEO’s profile and your company page, enable engagers and commenters, and you have a running list of people already aware of you.

Engaging with competitors

Detects: People interacting with your competitors’ brands. Sources: Social Configure: Add competitor profiles or company pages. Example: Add three competitor pages. You get everyone engaging with them — people already interested in your category and actively comparing.

Engaging with Top Voice or Brand

Detects: People and companies interacting with a selected industry voice or brand. Sources: Social Configure: Add the profile or page of the voice you want to watch. Example: Track a well-known analyst in your category. Their audience is your audience, engaging in public.

Hiring for a role

Detects: Companies actively hiring for roles you sell to or sell through. Sources: Social, Job board · Research: social and web Configure: The role or function you care about.

Job board filters

The job-board option matches real open roles rather than posts about hiring, with fine-grained control over who you catch. Every field is optional — fill in the ones that matter.

Advanced job title search

Full operator reference for the title field.

Boolean search

Operators shared by every boolean field in Clearcue.
Example: Title "VP Sales" OR "Head of Sales", location United States, seniority Director/Executive, agencies excluded — US companies appointing senior sales leadership.
Run both options together. The social one catches the excitement around a hire; the job board one catches every role that actually opened, including the quiet ones.

Starting a new role

Detects: People who have just joined a company or moved into a new role. New joiners have fresh discretion over tooling. Sources: Social, Database · Research: social and web Configure: The role or function to watch. The database option takes its own filters.

Database filters

These two options answer the same question from different angles. The announcement only exists if the person chose to post; the database option doesn’t depend on anyone posting at all. Run both.
Example: Social option on “VP of Engineering”, plus the database option with company size 201-500 — every new engineering leader at mid-market companies, whether or not they posted.

Announcing funding

Detects: Companies announcing a funding round at a stage you care about. Sources: Social · Research: social and web Configure: One or more round types — Angel, Pre-seed, Seed, Series A, Series B, Series C, Series D+. AI qualification filters for genuine announcements and drops investing advice, fund news, and general finance commentary. Example: Select Series A to reach companies about to scale their team and tooling. Add web research and you also catch the coverage that never appears on LinkedIn.

Opening a new office

Detects: Companies opening an office or expanding to a new location. Sources: Social · Research: social and web Configure: A city, region, or country. Announcements name cities, so a country is expanded into its main business hubs. AI qualification filters for real office openings and drops remote-work discussion and coworking promotion. Example: Enter “Germany” to track companies setting up locally — they need local vendors, partners, and services.

Attending an event

Detects: People going to, speaking at, sponsoring, or exhibiting at a specific event. Sources: Social There are two ways to track an event. Pick the one that matches what the event has — you don’t need both. Configure (page): The event’s LinkedIn page, plus the edition year, start and end dates, and hashtag. Configure (keywords): The event name and the terms posts use for it — the acronym with its edition year, the hashtag, the full name, the organiser’s brand. You can also name the organiser, and flag anything else the name could be confused with.
AI qualification narrows the results to real participation — attending, speaking, sponsoring, exhibiting, or inviting people to meet — and drops the organiser’s own promotion, news recaps, and posts that merely name-drop the event.
Speakers, sponsors, and exhibitors have committed budget to the event. They signal stronger intent than attendees, and Clearcue records which one each person is.
Example: Track next year’s edition of your industry’s main conference a month out, and reach attendees before their calendar fills.

Engaging with lead magnets

Detects: People grabbing free resources — guides, templates, checklists — in exchange for a comment or a DM. Sources: Social Track lead magnets two ways: Example: Track “cold outreach” by topic. Everyone commenting to get the guide has told you what they’re working on.

Looking for an agency

Detects: People looking to hire an agency, consultancy, or freelancer. Sources: Social · Research: social and web Configure: The service or discipline being sought. AI qualification keeps only posts where the author is the buyer, and rigorously drops agencies advertising themselves, recruiters, job listings, and listicles about choosing a provider. Example: Enter “web development” to find business owners actively looking — real buying intent, not browsing.

Recommending a tool or service

Detects: People recommending or endorsing a specific tool or service. Sources: Social · Research: social and web Configure: The tool or service name. Example: Enter a competitor’s name to find people publicly endorsing them. Join the thread and introduce your alternative while the topic is live.

Making a first hire

Detects: A tracked company’s first-ever hire in a country, in a department, or in both at once. Sources: Database A first hire is a company entering territory it has never been in. The first sales hire in the US, the first engineer in Germany — each one means a new budget owner and a new set of tools to buy. Under Specific criteria, set a country, a department, or both. With neither, the signal would fire on every arrival at every company you track.
Setting both is the strictest case. A “first German engineer” is a first that neither margin proves alone — the company may already have engineers elsewhere and Germans in other departments.
“First” means first among the people working there now. A company whose only German salesperson left years ago counts a new German hire as its first — which is the point: nobody from Germany works there today, and someone just joined.
Example: Specific criteria, country Germany, department Sales, 6-month window — your accounts putting their first salesperson on the ground in Germany.

Hiring in a new location

Detects: A tracked company with open roles in a country where it currently has no employees. Sources: Database, Job board This is a company entering a market it has no presence in — before any office announcement, and before the first hire lands.
For market entry in general, use Any country. Which market they’re entering is the answer, not the question — naming a list in advance means every market you didn’t think of goes unreported.
Example: Track 50 accounts with Any country and you get a card per gap: “3 open roles in Brazil, no employees there.”

Team growth or decline

Detects: A tracked company whose headcount in a country, a department, or overall grew or shrank past a threshold you set. Sources: Database It measures who actually works there, so it reports the outcome of hiring rather than the intent. Set at least one minimum. With only maximums, a change of one person satisfies the signal.
One signal measures one population. Naming several countries or departments combines them into a single number — Germany + France is one figure, not two. For per-country answers, create one signal per country.
Percentage and people thresholds catch different companies. A 30% rule finds fast-growing startups and misses enterprises; a 500-people rule does the reverse. Current team size — at least is what stops a team going from 1 to 2 reading as 100% growth.
Example: Growth, 6 months, department Sales, at least 30%, current team at least 10 — accounts scaling a real sales team.

Warm paths

Detects: Someone who currently works at a tracked company and shares real, dated history with a contact you already know. Sources: Database Shared history means the same employer or the same school, at overlapping times. Two people who were somewhere a decade apart never crossed paths, so they aren’t a match.
One signal is one person. To surface warm paths through three contacts, create three signals — one per contact.
Nothing else needs configuring. There’s no prompt and no keyword: whether two careers overlapped is a matter of record. Example: Add your VP of Sales’ profile. You get everyone at your tracked accounts who worked alongside them or studied with them — each one an introduction they can make personally.
Matches are limited to people who work at a tracked company today. Someone who has since left offers no live introduction path.

Detects: Posts matching keywords or a boolean query you write. Sources: Social Configure: Your keywords, or a full boolean query. Optionally turn on the AI filter — Clearcue builds a relevance filter from your description so only genuinely relevant posts are captured.

Boolean search

Operators, phrases, and grouping for every boolean field in Clearcue.
Example: Track your category’s vocabulary to find people discussing the problem you solve, before they’ve named a vendor.

Custom AI signal

Detects: Anything you can describe — a pattern none of the behaviours above covers. Sources: Custom Instead of choosing pre-built logic, you describe exactly what you want to find and AI evaluates every candidate against your criteria. Configure: Your prompt, plus who you want from the matching posts:

Write an effective prompt

Your prompt is a filter, not a search. It decides whether a piece of content counts.
  • State what qualifies in one sentence. “Qualify posts where the author says their team is struggling to hire developers.”
  • Name the hard disqualifiers. Most of the work is exclusion. Rule out self-promotion, recruiters, job seekers, listicles, and secondhand commentary.
  • Say who the subject is. “When qualified, the subject is the post author” — or the company they name.
  • Be concrete about edge cases. If a balanced review shouldn’t count, say so.
The behaviours above come with prompts we’ve already tuned. Pick the closest one first — a custom prompt is for what’s genuinely not in the library.

Content vs signal qualification

Two filters run at different moments:
  • Content qualification decides whether a post is relevant at all.
  • Signal qualification decides whether a specific person’s engagement with it counts.
Start with content qualification. Ruling a post out there means nobody who engaged with it reaches your records.
A prompt only ever sees the post text. It cannot filter on company size, industry, seniority, or location — that information isn’t in the post. Use filters and audiences for those.

Other ways to create a signal

Describe it to Cue AI

Type what you’re looking for in your own words. Cue AI picks the behaviour, drafts the configuration, and shows you every proposed signal before anything is created. This is also how you reach behaviours and options that have no tile of their own.

Lookalike signals

Give Cue AI a few example posts you’ve found valuable and ask for more like them. It analyses what they have in common and generates the boolean query and AI prompt for you, which you can review and adjust before saving.
Lookalike signals are created by asking Cue AI — there’s no tile for them in the library.

Outreach signals

If you run outreach through HeyReach or Lemlist, their activity becomes signals automatically. These are powered by webhooks — they activate as soon as the integration is connected, with no profile or query to configure.
Outreach signals need an active HeyReach or Lemlist integration.

Research

Add research to a signal so it looks into your tracked companies directly.

Manage signals

Pause, resume, edit sources, and understand signal limits.

Signal stacking

Combine behaviours so real intent separates itself from noise.

Core concepts

How signals, audiences, and lists fit together.