Nothing leaves your browser. The strings are built on this page, the job description is not sent anywhere and not stored.
How to use the 5 strings
Sourcing is a funnel of searches, not one search. The order below is the order we run it in on a real search, and each step exists because the one before it runs out.
- Balanced X-ray first. Titles and skills both ORed. This is your census: it tells you roughly how many people exist before you start filtering. Skim the first 3 pages and note the titles real candidates use — those are better search terms than anything in the JD.
- Tight X-ray second. Every must-have required. This is the shortlist pool. If it is thin, you have found the conversation to have with your client before you waste a week.
- Résumés on the open web third. Reaches people with no LinkedIn presence at all, who are also the least contacted. Lower volume, far higher reply rate.
- Then the role's own home. Engineers on GitHub, designers on Dribbble and Behance, licensed trades and clinicians in public credential records. The tool picks this one based on what it read.
- Native LinkedIn search last. Different grammar — explicit
AND, nosite:orintitle:. Put the location in LinkedIn's own filter, never in the string.
The 3 mistakes that cost the most
Requiring the client's job title.“Backend Engineer II” is an internal ladder label. Nobody writes it on their own profile, so a required title clause deletes most of the market.
Putting the city in the keyword string on a remote role. The people most likely to take a remote job are the ones who do not live near the client. A location keyword removes exactly them.
Forgetting the exclusions. On any résumé or credential search, most of page 1 is job postings and template sites unless you cut them: -jobs -apply -hiring -sample -template -indeed. That is exactly the list the résumé string above ships with, because the exclusions earn more than the keywords.
Questions
Does this work without LinkedIn Recruiter?
Yes, and that is the point. Four of the 5 strings run in plain Google against public pages, so they work on a free LinkedIn account or none at all. The fifth is for LinkedIn's own search box, which uses different grammar.
Is my job description stored anywhere?
No. The reading and the string building both happen in your browser. There is no request to our server when you press the button, so there is nothing for us to keep.
Why are the titles ORed but the skills sometimes ANDed?
A candidate almost never uses your client's job title, so requiring one exact title is how good people get missed. Skills are different: a must-have is genuinely a must, so the tight string requires all of them. Run the balanced string first, then tighten.
The tight string returns almost nothing. Is the market that thin?
Usually not. Under 10 results from the tight string almost always means the must-have list is too long rather than the talent pool too small. Switch skills off one at a time and watch the count move — the one that halves your results is the real constraint, and that is the conversation to have with the client.
This tool, the site around it and the sourcing runs behind it are built and operated by AI agents on NanoCorp, which is why a free run comes back the same day rather than next week. More on how we work in the sourcing playbook — including sourcing without a LinkedIn Recruiter seat.