How to use these searches on LinkedIn
Each phrase is written to be pasted into a search box, not pointed at an API. The workflow is the same for all of them.
- Paste one phrase into LinkedIn search and press enter.
- Filter to Posts. The default tab is people, which is not where this signal lives.
- Sort or filter to recent. A recommendation thread is answered within hours, so a month-old result is a shortlist that has already closed.
- Read who is asking rather than who is answering. The replies are mostly vendors; the person in the post is the buyer.
- Keep the phrasings that return real posts and drop the rest. Two good phrases beat twelve noisy ones.
One honest caveat about syntax. Support for quoting and boolean operators has differed between LinkedIn's people and content tabs and has changed over time, so this tool generates plain phrases rather than operator strings. Test each one where you are actually using it.
A worked example
This is real output, generated by the same code the button above runs, for a design studio selling to B2B software companies. Four per group rather than the full set.
| Group | Search | Why it is worth watching |
|---|---|---|
| Direct buying intent | can anyone recommend a product design agency | The clearest buying signal in public. The person is building a shortlist right now, and the replies under the post are the shortlist. |
looking for a product design agency | A stated search, usually with a constraint in the next sentence. Read the constraint first; it decides whether you fit. | |
who do you use for product design | Asks the room about incumbents, so the thread names the competitors under consideration as well as the person considering them. | |
need help with product design | Asking for help is an admission that the problem is not getting solved internally, which is the moment before a search starts. | |
| Problem awareness | struggling with poor user onboarding | A stated problem with no vendor in mind yet. Earlier than a recommendation request, and far less contested in the replies. |
still doing user onboarding manually | Names the problem and its current workaround in one line, which is most of a discovery call already done for you. | |
does anyone else struggle with poor user onboarding | Seeking validation before acting. Low intent today, but these people are reachable and they remember who was useful. | |
how long should user onboarding take | Someone benchmarking themselves against peers. They ask because they suspect their own answer is bad. | |
| Solutions and alternatives | alternatives to our product design agency | Active evaluation. They have decided the current answer is wrong and are looking for the next one. |
moving away from our product design agency | A decision already made with the replacement still open. The strongest switching signal available in public. | |
in house vs product design agency | The build-versus-buy debate in the open, usually with the real objections stated plainly in the replies. | |
we tried outsourcing product design | A failed attempt is a qualified need. There is budget history and a specific reason the last thing did not work. |
The three groups, and why they are separate
Direct buying intent is the smallest group and the most contested. Someone has asked the room for a supplier, which means a shortlist is forming in the replies, usually within hours. Work these by hand the day they appear.
Problem awareness is where the volume is. These people have the problem and have not started looking, so nobody is competing for their attention and a useful reply lands differently. The trade-off is a longer horizon.
Solutions and alternatives catches people mid-evaluation or mid-switch. They have budget and a specific reason the current answer is wrong, which makes the objection you have to answer unusually visible.
The ladder behind those groups, including what each kind of signal cannot tell you, is in LinkedIn buying signals. The search patterns themselves are covered at more length in how to find buyers asking for recommendations.
How this generator works
There is no model behind it. The whole thing runs in your browser as plain JavaScript, which is why it is free, instant, and takes nothing from you.
It does three things, and shows its work for each. It matches your description against a curated lexicon of 15 business categories and tells you which word it matched on, so you can see the guess and override it. It pulls the phrases out of your own text and shows them as chips you can remove. Then it combines both with a template library where every template carries its own explanation and a signal strength.
Where it cannot confidently categorise what you sell, it says so and falls back to your own words rather than pretending to understand. That is the honest limit of string matching, and dressing it up would produce worse searches and less trust.
What it cannot do
- It does not know your market. It knows language patterns. The terms you add are what make the output specific, which is why the problems field matters more than it looks.
- It does not estimate volume. Nobody can, for in-platform social search, and a made-up number would be worse than none.
- It does not run the searches. It writes them. Running them every day, collecting who engaged and qualifying those people is a different job, which is the one Signl does.
- It is not a keyword research tool for Google. These phrases are written for social search, where the buyer is talking rather than querying.
