Boolean search strings for recruiters: syntax, 20 strings and how to fix a bad one

A Boolean search string is a query that combines keywords with AND, OR, NOT, quotation marks and parentheses so a search engine returns only the profiles that match your requirements. For recruiters the working pattern is short: synonyms for the title joined with OR inside brackets, AND one or two must-have skills, NOT the titles you do not want. Everything else is getting the details right for the tool you are searching in.

Below are the operators as LinkedIn and Google document them, 20 strings you can copy and adapt, and a method for fixing a string that returns nothing or far too much. Turning results into conversations is covered in how to source candidates on LinkedIn; this page is about the string itself.

The operators, and what each tool actually supports

Most Boolean guides were written for an older web and list operators that no longer work. These are the rules as published by the vendors themselves, as of September 2026.

LinkedIn's Boolean search help page gives the order it evaluates a query in: quoted phrases first, then parentheses, then NOT, then AND, then OR. It also says common stop words such as "by", "in" and "with" are ignored, and that wildcards are not supported. The Recruiter and Recruiter Lite page adds that the + and - operators are not officially supported, that the job title, company and keyword fields accept Boolean, and that overly long queries can cause result problems. Google's own search operators page documents quotes, site:, the minus sign, filetype:, and before:/after:.

Why the precedence order matters

Because AND binds tighter than OR, this string does not do what it looks like:

java OR kotlin AND android

It is read as java OR (kotlin AND android), so every Java developer on the platform comes back, Android or not. Brackets make the intent explicit, and on LinkedIn you should use them every time you mix OR with anything else:

(java OR kotlin) AND android

Build a string from the intake, in four blocks

A string is only as good as the requirements behind it. Before writing one, you need the answers from the intake meeting: the titles people doing this work actually hold, the two skills the hiring manager will not compromise on, and the backgrounds they have already ruled out. Then build it in four blocks.

1. Title block. Every title the role goes by, joined with OR, in brackets. Ask the hiring manager what their current team's titles were before they joined.
2. Must-have block. One or two skills that separate a fit from a near-miss, each with its own synonyms in brackets.
3. Context block (optional). An industry, a system or a certification that narrows the pool, only if the client said it is required.
4. Exclusion block. NOT and the titles that keep polluting results: usually more senior, more junior, or a neighboring function.

(title OR title OR "two word title")
AND (must-have OR synonym OR acronym)
AND (context OR synonym)
NOT (excluded OR excluded)

On LinkedIn, put the title block in the title filter where you can, rather than the keyword box. A keyword search matches anywhere on the profile, so "controller" in the keyword box finds people who once reported to a controller.

20 Boolean strings by role

Adapt these; do not paste them unchanged. Every requirement below is an example, so swap in the systems and certifications from your own intake.

Technology

1. Backend engineer
("backend engineer" OR "back end engineer" OR "software engineer") AND (golang OR "go developer") AND (postgres OR postgresql) NOT (intern OR student)

2. Data engineer
("data engineer" OR "analytics engineer") AND (airflow OR dagster) AND (snowflake OR bigquery OR redshift) NOT (analyst OR scientist)

3. QA automation
("qa engineer" OR "sdet" OR "test automation engineer") AND (playwright OR cypress OR selenium) NOT manual

4. Salesforce administrator
("salesforce administrator" OR "salesforce admin" OR "crm administrator") AND ("sales cloud" OR "service cloud") NOT (developer OR architect)

5. Network engineer
("network engineer" OR "network administrator") AND (ccnp OR ccna) AND (bgp OR ospf) NOT (sales OR "account manager")

6. Help desk
("help desk" OR "service desk" OR "desktop support") AND ("active directory" OR intune OR "microsoft 365") NOT manager

Finance, HR and operations

7. Controller
(controller OR comptroller OR "assistant controller") AND (cpa OR "certified public accountant") AND ("month-end close" OR "month end close") NOT (intern OR "financial controller trainee")

8. FP&A analyst
("financial analyst" OR "fp&a analyst" OR "fpa analyst") AND (forecasting OR budgeting) AND (anaplan OR adaptive OR excel)

9. Payroll specialist
(payroll) AND (adp OR paylocity OR ukg OR workday) AND ("multi-state" OR "multistate") NOT (sales OR "account executive")

10. HR generalist
("hr generalist" OR "human resources generalist" OR "hr business partner") AND ("employee relations" OR investigations) AND (shrm OR phr OR sphr)

11. Supply chain analyst
("supply chain analyst" OR "demand planner" OR "inventory analyst") AND (sap OR oracle) AND (forecasting OR "s&op")

12. Operations manager
("operations manager" OR "plant manager" OR "site manager") AND (lean OR "six sigma") NOT (intern OR assistant)

Healthcare, trades and sales

13. Medical coder
("medical coder" OR "coding specialist" OR "coding analyst") AND (cpc OR ccs) AND ("icd-10" OR "icd 10")

14. Radiologic technologist
("radiologic technologist" OR "x-ray technologist" OR "rad tech") AND (arrt) AND (ct OR "computed tomography")

15. HVAC technician
(hvac OR "hvac technician" OR "service technician") AND ("epa 608" OR "universal certification") AND (commercial OR rooftop)

16. Construction superintendent
(superintendent OR "site superintendent") AND ("osha 30" OR "ssho") AND (commercial OR "ground up" OR "tenant improvement")

17. Enterprise account executive
("account executive" OR "enterprise sales" OR "strategic account manager") AND (saas OR software) AND ("president's club" OR quota) NOT (sdr OR bdr)

18. Recruiter
(recruiter OR "talent acquisition partner" OR "technical recruiter") AND (bullhorn OR greenhouse OR lever) NOT ("sourcing coordinator" OR intern)

Google X-ray strings

X-ray searching uses Google to search inside a site. It reaches public pages only, and it is useful when you do not have a Recruiter seat or want a second view of a market.

19. Public LinkedIn profiles in a city
site:linkedin.com/in ("site reliability engineer" OR "sre") "greater chicago" -jobs -recruiter

20. Code-hosting profiles by language and location
site:github.com "rust" "berlin" -topics -trending

Treat X-ray results as leads, not profiles. Public pages can be months out of date, and what Google shows depends on what each member has chosen to make public.

Fixing a string that returns zero or 10,000 results

Too few results

• Check the syntax first. Lowercase or, curly quotes pasted from a document, or an unclosed bracket break a string silently. Retype the quotes in the search box.
• Remove one AND block at a time. Whichever removal brings results back is the requirement the market does not describe the way your client does.
• Add synonyms, not new blocks. People write "P&L" and "profit and loss", "AWS" and "Amazon Web Services". Add both.
• Drop certifications from the string. Many qualified people never list them. Ask about the license on the screen instead.

Too many results

• Move the title into the title filter. This alone often removes most of the noise.
• Add a context term that only practitioners write. "Month-end close" is written by accountants; "accounting" is written by everyone who ever worked near one.
• Exclude what you keep scrolling past. Look at the first 25 results and add a NOT for each pattern that is wrong, such as vendors, trainers, students or sales roles selling to your target function.
• Use quotation marks on two-word titles. Without quotes, project manager matches anyone with both words anywhere on the page.

A worked example: from intake note to final string

An invented search, shown in three passes.

The last step is the one strings cannot do. A profile that says "multi-state payroll" may mean running it or supporting the person who did. That is a screening question, and the payroll specialist screening questions page has the wording.

Keep a string library, not a folder of one-offs

The strings that work are the most reusable thing a desk produces. Save them where the team can find them, with enough context to reuse:

Record what the string misses, not only what it finds. The next recruiter to use it will otherwise assume it covers the market.

Mistakes that quietly cost you candidates

• Searching for the job description. Stacking eight AND terms finds people who wrote their profile like a job ad, not the people who do the work best.
• Excluding too broadly. NOT manager removes every "engineering manager" you did not want and every engineer whose profile mentions working with a product manager.
• Treating the string as a filter for protected characteristics. Strings that proxy for age, such as graduation years, or for gender or ethnicity, such as names or affinity groups, invite discrimination claims. Search on the work.
• Never searching the other way. Run the string against your own ATS or CRM before LinkedIn. People who already know your agency answer more often.
• Automating the search. LinkedIn's User Agreement restricts scraping and automated access. Browser extensions that bulk-export profiles put the account you depend on at risk.