Boolean search for recruiters: a 2026 guide with examples

You wrote the boolean string. You ran it. LinkedIn returned six results. The same string on Google X-ray pulled 400, most of them wrong people.

That gap is not your fault. It is where boolean search actually starts to break.

Most guides skip this part. They list the operators, drop in a few clean examples, and call it done. This one does not. You get every operator, a platform-by-platform syntax reference, six worked strings by role, and an honest section on where boolean stops working and what to do about it.

According to LinkedIn, 67% of recruiters report struggling to find quality candidates consistently. Boolean is part of that picture. Not because boolean is bad. Because the rest of the industry is still pretending it is the whole answer.

What boolean search is, in 2026

Boolean search is a way of building precise queries by combining keywords with logical operators: AND, OR, NOT, quotation marks, and parentheses. You tell the search engine which terms must appear, which can substitute for each other, which to exclude, and which phrases to treat as a unit.

Boolean is a sourcing tool, not a recruiting tool. You use it when you need a specific kind of candidate: a particular job title, a required skill, a named tool, a location. It is fast, it is precise when written well, and it works on every major sourcing surface from LinkedIn Recruiter to Google to GitHub.

It is also under more pressure in 2026 than it has been in years. LinkedIn shipped an AI semantic search layer that interprets intent rather than matching keywords. New AI sourcing tools scan millions of profiles in parallel and return ranked matches. The recruiter who only knows boolean is doing one tenth of what is now possible. The recruiter who knows boolean and knows where it falls short is doing the other nine.

The five boolean operators, with recruiter examples

Five operators do almost all the work. Get fluent in these and you can write 90% of the strings you will ever need.

Three notes that trip up most recruiters. Operators must be uppercase on LinkedIn and most other platforms (lowercase "and" gets treated as a regular word). Most platforms do not support a wildcard * in boolean searches. And a space between two words is interpreted as AND on LinkedIn and Google but behaves differently on some ATS systems, so when in doubt, write AND explicitly.

Platform-specific syntax: where the same string behaves differently

The single most expensive boolean mistake is assuming that a string that works on Google works on LinkedIn. It does not. Each platform has its own dialect, its own field structure, and its own silent failure modes.

The "silent failure" is the bit that costs recruiters time. Your string returns zero results on LinkedIn and you assume the candidate pool is empty, when really the syntax was wrong. Test every new string on a known-good query first. If a string that should clearly return matches returns zero, the syntax is the problem, not the market.

Six boolean strings by role

These are working strings, not pseudocode. Copy them, adjust the locations and tools to your role, and run them. Each one is written for LinkedIn and lightly adapted for Google X-ray underneath.

Senior software engineer (Python, Bay Area)

("software engineer" OR "senior software engineer" OR "staff engineer") AND (Python OR Django OR FastAPI) AND (AWS OR GCP) AND ("San Francisco" OR "Bay Area" OR "Palo Alto" OR Oakland) NOT (intern OR junior OR student)

The OR-chain on titles catches the candidates who were promoted to Staff but still describe themselves as Senior. The Bay Area location group covers the standard self-descriptions. NOT clauses cut the early-career profiles that keyword-match Python but are not the role.

Data scientist (ML, healthcare, remote)

("data scientist" OR "ML engineer" OR "machine learning engineer") AND (healthcare OR clinical OR "medical imaging" OR biotech) AND (Python OR PyTorch OR TensorFlow) AND (remote OR "remote work" OR "remote-first") NOT (intern OR student OR bootcamp)

Healthcare ML talent is rare and lives across "healthcare," "clinical," "biotech," and the more specific subdomains. Loosen the industry OR-chain if you want to widen the net to general ML.

Product manager (B2B SaaS, growth)

("product manager" OR "senior product manager" OR "group product manager") AND ("B2B" OR SaaS OR "enterprise software") AND (growth OR activation OR retention OR "product-led" OR PLG) NOT (associate OR junior OR APM)

PM is the role where boolean strains hardest because titles and self-descriptions are highly variable. "Product Lead," "Director of Product," and "Product Owner" all overlap with PM but a tight OR-chain misses them. Build wider when sourcing PMs.

Enterprise account executive (fintech)

("account executive" OR "enterprise AE" OR "strategic account") AND (fintech OR "financial services" OR banking OR payments) AND ("$1M" OR "million quota" OR "enterprise sales" OR "$500K+") NOT (SDR OR BDR OR "inside sales")

Quota and segment language is what distinguishes a real enterprise AE from a mid-market or SMB rep. Sales reps tend to write quota numbers explicitly in their LinkedIn summary, which is one of the few places boolean keyword matching actually works well.

Senior product designer (Figma, portfolio)

("product designer" OR "senior product designer" OR "UX designer") AND (Figma OR Sketch) AND (portfolio OR behance.net OR dribbble.com) NOT (graphic OR print OR student)

The portfolio mention filters out designers who do not link work, which biases toward the more senior end of the market. NOT graphic and NOT print cuts the print-design candidates who keyword-match "designer."

DevOps and SRE (Kubernetes, multi-cloud)

("DevOps engineer" OR "site reliability engineer" OR SRE OR "platform engineer") AND (Kubernetes OR k8s) AND (AWS OR GCP OR Azure) AND (Terraform OR Ansible OR Pulumi) NOT (intern OR junior)

Platform engineering is splintering into half a dozen overlapping titles in 2026. Be ready to add "infrastructure engineer" and "cloud engineer" to the OR-chain as you go.

Google X-ray search for recruiters

X-ray search is boolean plus Google operators. You use it when LinkedIn search is too narrow, when you do not have Recruiter access, or when you want to see public profile pages outside the LinkedIn search index.

The five operators that matter:

site:linkedin.com/in/ restricts results to public LinkedIn profiles. Combine it with your boolean string and you get a Google-indexed view of candidates.

inurl: matches a string in the URL. inurl:resume filetype:pdf finds resumes posted as PDFs.

filetype: returns one file type. filetype:pdf python engineer finds PDF resumes that mention Python and engineer.

intitle: matches words in the page title. Useful for portfolio and personal sites.

- excludes a term, equivalent to NOT. Boolean operators on Google use the dash, not the word NOT.

A working X-ray string for senior product designers:

site:linkedin.com/in/ ("product designer" OR "UX designer") ("San Francisco" OR "Bay Area") ("portfolio" OR "behance" OR "dribbble") -junior -student

X-ray gets you reach. It does not get you precision. The same string on LinkedIn Recruiter will return better matches because the platform knows the field structure (current title, location, years of experience). X-ray is the right tool when the platform-specific tool is not available, or when you suspect the platform is hiding candidates from you. If you find yourself running X-ray often because LinkedIn returns too few results, dedicated resume databases may give you better coverage without the Google middleman: we compared the major ones in free resume search tools for recruiters.

Where boolean breaks

This is the part most guides skip. If you have been running boolean strings for more than a year, you already know it. Here it is named.

The synonym blind spot. A candidate calls themselves "Product Designer" instead of "UX Designer" and your string misses them. You can extend the OR-chain, but you do not know what you do not know. Senior designers in particular use "Design Lead," "UX Architect," "Senior IC," and a dozen other variants that no static OR-chain captures cleanly. The string returns 12 candidates when there are 80.

The platform tax. A string that pulls 200 candidates on Google X-ray returns 6 on LinkedIn and 0 on Indeed. Different stemming, different operators, different field coverage. You cannot transfer knowledge cleanly across platforms. Every new platform is, in practice, a new query language to learn.

The maintenance problem. Job titles drift. "DevOps" splits into Site Reliability Engineer, Platform Engineer, Cloud Engineer over two or three years. Tool names shift. Your string for "ML engineer with PyTorch" needs to be rewritten when half the market moves to JAX. Strings rot. Recruiters rebuild them every quarter, often from scratch, and the institutional knowledge of what works lives in someone's saved search rather than in a system.

The scale ceiling. Boolean is sequential. You read one profile at a time. AI sourcing tools scan tens of millions of profiles in parallel and return ranked matches in seconds. A human running boolean is doing serial computation against a parallel problem. You can be very good at boolean and still hit a hard ceiling around how many roles you can source per week. This is also where passive candidate sourcing breaks down: passive talent does not optimize their profile for keyword matching, so the candidates you most want are exactly the ones boolean is least likely to surface.

The intent gap. This is the one boolean cannot fix from inside. Boolean matches words. It cannot match "I need someone who has scaled an early-stage data team from one to ten in a healthcare context." That is intent, not keywords. No OR-chain captures it. You can keyword-match "data team" and "healthcare" and "scaled," but you will return 200 candidates of whom three actually fit, and you will not know which three until you read all 200.

The honest summary: boolean is necessary, sufficient for some searches, and a ceiling on others. The skill is knowing which is which.

When boolean works, when intent-based search works

Most senior recruiters in 2026 use both. The split is not random. It tracks the kind of search.

The hybrid workflow most senior recruiters land on: intent-based search for breadth and ranking, boolean for narrowing and verification. You start with a description of the role in plain English, get a ranked candidate set in seconds, and then apply boolean filters to enforce the non-negotiables. Different intent-based tools work for different roles, which is why we broke down the category by what each tool actually does rather than ranking them.

This is where Glozo was built. The Skill Graph weighs candidates against a job description rather than matching keywords, and the Open-to-Offers signal tells you which candidates are likely receptive before you spend a credit. It is the answer to the synonym blind spot and the intent gap. Boolean still has its place: you use Glozo's intent-based search for the breadth pass, then narrow with boolean filters when the role demands precision.

A practical sourcing checklist

Run a new role through this before saving the search.

Start broad with two or three OR-chains for titles. Layer in must-have skills with AND. Add NOT only for known sources of false positives, never preemptively. Quote-mark every multi-word title. Test the string on two platforms before committing, and compare result counts. Save the working string with the date and the platform it was tuned for. Strings rot quickly, and a string that worked in March can return half the candidates in September. Re-check every quarter, especially in fast-moving domains like AI and data.

Flag the gaps as you go. If a search feels narrow on synonyms, mark the role for an intent-based pass before assuming the candidate pool is small. The gap is usually in the string, not the market. If you want market data to confirm before you commit to the role, Glozo's job market intelligence shows live supply and demand by role and location.

Better sourcing starts with knowing where boolean stops

The recruiters who get the best results in 2026 are not the ones with the longest boolean strings. They are the ones who know exactly which roles benefit from a tighter string, which roles benefit from a wider semantic search, and which need both. If you have hit the synonym blind spot or the intent gap one too many times, Glozo was built for the searches boolean cannot describe.