TL;DR
- A Boolean search combines keywords with the operators AND, OR and NOT, plus quotation marks for exact phrases and brackets for grouping, so that a database, job board or search engine returns only the records that match the logic. The name comes from Boolean logic, the true-or-false algebra that search engines apply to every term.
- Recruiters use it to find candidates whose profiles or CVs contain the right combination of skills, titles and locations. LinkedIn documents the five operators and the order in which it applies them: quotes first, then brackets, then NOT, then AND, then OR, all in capital letters. Google uses quotes and a minus sign to exclude, and has no AND operator because a space already means it.
- Build a string in four steps: list the must-have skill, list every way people describe it, group the synonyms with OR inside brackets, then join the groups with AND and remove the noise with NOT. Widen before you narrow, test the string against a person you know fits, and remember that a string describes how candidates write about themselves, not the job.
Type “project manager” into a CV database and you get everyone who has ever managed a project, plus everyone who wants to. Type a well-built Boolean search and you get the fifty people who match the role. The difference is five operators and a method for combining them, and it is the most useful technical skill in sourcing.
This article explains what a Boolean search is, sets out the operators as LinkedIn and Google themselves document them, walks through building a search string in four steps with worked examples, and lists the mistakes that quietly hide good candidates.
The search finds the candidates; the hiring process decides. See how New Dynamics workflows let you shape the stages, forms and approvals of your own hiring process rather than adopting someone else's template.
What is a Boolean search?
A Boolean search is a query that combines keywords with logical operators so that a search tool returns only the records matching the logic. The name comes from Boolean algebra, the branch of mathematics named after George Boole in which every statement is either true or false. A search engine tests each term against each record in exactly that way: does this profile contain “Python”, true or false? The operators tell it how to combine the answers.
Most search boxes accept Boolean logic whether or not they say so: LinkedIn, job boards, applicant tracking systems, CV databases, library catalogues and, with its own syntax, Google. Recruiters use it to search for people; the same logic works for anything.
The five operators
LinkedIn's help centre documents the five operators its search supports. They are the same five almost everywhere.
| Operator | What it does | LinkedIn's wording |
|---|---|---|
| Quotes | Match an exact phrase | “For an exact phrase, enclose the phrase in quotation marks” |
| AND | Require every term | “Type AND (all caps) to see results that include all items in a list” |
| OR | Accept any one of several terms | “Type the word OR (all caps) to see results that include one or more items” |
| NOT | Exclude a term | “Type the word NOT (capital letters) immediately before a search term to exclude it” |
| Brackets | Group terms so the logic is applied in order | “Combine terms with parentheses to build complex logic” |
Two rules matter as much as the operators. First, capitals: LinkedIn says “you must type them in uppercase letters for the search to work correctly”, and most databases are the same, so write AND, OR and NOT, never and, or and not. Second, the order of precedence. LinkedIn applies quotes first, then brackets, then NOT, then AND, and OR last. Brackets exist so that you can override that order, in the same way they do in arithmetic.
LinkedIn also says what it does not support: braces, square brackets, angle brackets and wildcards such as the asterisk, and it warns that the plus and minus signs “may appear to work in some cases” but “are not officially supported”. Other databases differ, so check each platform's own help page before relying on a wildcard.
Google's version
Google speaks a dialect. Its search operators page documents quotation marks for an exact phrase, a minus sign placed directly before a word to exclude it, and prefixes such as site: to search one domain and filetype: to find a document type, with the instruction not to put a space between the operator and the term. There is no AND, because a space between two words already means both must appear. Recruiters use Google for what LinkedIn cannot show: portfolios, conference talks, published papers and personal sites.
How to build a search string in four steps
A search string is a hypothesis about how the right candidate describes themselves. Build it in this order and widen before you narrow.
1. Name the must-have
Start with the one skill, qualification or title without which the person cannot do the job. Not the whole job description: one thing. For a payroll role it might be a payroll qualification; for a developer, a language.
2. List every way people describe it
People do not write their profiles from your job description. Collect the synonyms, abbreviations, US and UK spellings and older titles. “Payroll manager” is also “payroll lead”, “payroll supervisor” and “head of payroll”; “CIPP” is also “Chartered Institute of Payroll Professionals”.
3. Group the synonyms with OR, inside brackets
Each group of synonyms goes inside its own brackets, joined by OR, with multi-word phrases in quotes:
("payroll manager" OR "payroll lead" OR "payroll supervisor" OR "head of payroll")
4. Join the groups with AND, then remove the noise with NOT
Combine the must-have group with a second group for the next requirement, and exclude what keeps polluting the results:
("payroll manager" OR "payroll lead" OR "head of payroll") AND (CIPP OR "Chartered Institute of Payroll Professionals") AND (Bristol OR Bath) NOT recruiter
Run it, look at the first page, and adjust. If a person you know is right for the job does not appear, the string is wrong, not the person.
Worked examples
These strings are examples. Adapt the terms to your platform and your role, and test each one before you trust it.
A software engineer with a specific stack, excluding agencies
(developer OR engineer OR programmer) AND (Python OR Django) AND (AWS OR "Amazon Web Services") NOT (recruiter OR "talent acquisition" OR agency)
A registered nurse for a specialism, either spelling
("registered nurse" OR RN OR RGN) AND (paediatric OR pediatric OR children's) AND (Manchester OR Salford)
A finance lead with a qualification, excluding trainees
("finance manager" OR "financial controller" OR "head of finance") AND (ACA OR ACCA OR CIMA) NOT (trainee OR "part qualified" OR student)
A Google search for portfolios rather than profiles
"product designer" (portfolio OR "case study") Figma -jobs -vacancy site:*.co.uk
Notice the pattern: every group of synonyms is an OR inside brackets, every requirement is joined by AND, and NOT is used sparingly for the terms that keep appearing in the wrong results.
Mistakes that hide good candidates
- Narrowing too early. A string with six AND groups returns nobody. Start with two and add one at a time.
- Searching the job description, not the person. Candidates write “led a team of 12”, not “people management skills”. Search for the words they use.
- Forgetting synonyms and spellings. Paediatric and pediatric, organisation and organization, CV and résumé. Every missed variant is a missed person.
- Over-using NOT. Excluding “manager” to avoid managers also excludes every profile that mentions reporting to one.
- Lower-case operators. On LinkedIn,
python and djangosearches for the words “and” and “python”; onlyPython AND Djangoapplies the logic. - Encoding bias. Strings that require a particular university, a continuous work history or a native-sounding phrase screen people out for reasons unrelated to the job. Search for skills and evidence, keep the string reviewable, and treat it like any other selection criterion.
Boolean search and AI sourcing tools
Many sourcing tools now accept a plain-English description and build the Boolean string, or something like it, behind the scenes. That is convenient, and it is worth knowing what the tool did: the string is the selection criterion, and a criterion you cannot see is one you cannot check for fairness or fix when it misses. Our AI recruitment guide covers the questions to ask before trusting a tool with that step.
Frequently asked questions
What is a Boolean search in recruiting?
A search for candidates that combines keywords with the operators AND, OR and NOT, plus quotation marks for exact phrases and brackets for grouping, so that a CV database, job board or LinkedIn returns only profiles matching the combination of skills, titles and locations you specify.
What are the Boolean operators?
AND requires every term, OR accepts any one of several terms, and NOT excludes a term. Quotation marks match an exact phrase and brackets group terms so the logic is applied in the order you intend. LinkedIn applies them in the order quotes, brackets, NOT, AND, then OR, and requires the operators in capital letters.
Does Google support Boolean search?
Partly, with its own syntax. Google documents quotation marks for exact phrases, a minus sign directly before a word to exclude it, and prefixes such as site: and filetype:. A space between words already means both must appear, so there is no AND operator.
Why does my Boolean search return no results?
Usually because it is too narrow: too many AND groups, too few synonyms in each, or a NOT that removes more than intended. Remove one AND group at a time until results appear, then add synonyms before adding requirements back.
Do I still need Boolean search if my tool uses AI?
It helps. Tools that build the search for you are still applying a selection criterion; knowing Boolean logic lets you see what the tool searched for, check it for fairness and fix it when the right people are missing.
Your next step: write one string and test it
- Take your hardest current vacancy and name its single must-have.
- List ten ways candidates describe that must-have, including abbreviations and spellings.
- Build the string in the four steps above, with two AND groups at most to start.
- Test it against someone you know fits, then widen or narrow one term at a time.
For the fair process that follows the search, read and download our recruitment and selection guide. The guide is free to read, and the PDF uses our short download form.
Want the hiring stages, forms and approvals after the search to run on your own process? Book a New Dynamics demo and bring your current workflow. You can also email contact@new-dynamics.com.
