Pipeline
How Recruiters Search, and How to Be in the Results
A recruiter searches a database and filters hard before reading anything, so four fields decide whether you appear at all: an accurate current job title, the plain-language name of your work, the location you'd actually work in, and how recently your profile was active. Keyword density is not one of them. Write the headline in the words a recruiter would type, and one plain sentence under it naming the kind of problem you work on. Inbound favors certain titles and seniorities, so it supplements applying rather than replacing it.
The takeProfile-optimization content is resume advice wearing a different word count, and it recommends stuffing skills because the tools selling it profit from a score going up, not from a recruiter actually finding you. The two goals aren't the same, and content written to raise a number a human reader never sees is optimizing for the wrong reader entirely.
Where Olive fits
Open a role and see what the work shows
Whichever channel led an employer to you, Olive itself is never a filter or a search: it's a role-grounded assignment done openly with an AI assistant, described afterward in six evidenced findings a person writes, and you're handed the same report the employer reads, free.
Rank your shortlistWhere Do You Actually Show Up in a Recruiter's Search?
You show up when four fields line up: your current title, the plain-language name of your work, your location, and how recently you were active. A recruiter searching a database is running the sourcing half of hiring, and it works on a different mechanism than the screening half most advice is actually written for.
What both halves have in common is that a person types words into a box and the software hands back whoever matched them. Greenhouse's filtering tool, which searches the full text of resumes and internal notes during application review, states that to appear in the results "the keyword from your search must exactly match the keyword in the application" 1. That is one vendor's wording for one product, and other systems stem words, expand synonyms or match on embeddings. It is still the plain mechanical reason to name your work the way the field names it.
None of that is a resume-score number, and no algorithm is deciding anything about you. A person types a query because they're trying to fill a specific role, and the software returns whoever's profile matches the words they chose.
In the Harvard Business School and Accenture "Hidden Workers" survey of 2,275 senior leaders in Germany, the UK and the US, fielded in early 2020, 63% said they run a recruitment management system, rising to 75% in the US, and more than 90% of those said they use it to filter or rank middle-skills and high-skills candidates before a person looks 2. Sourcing and screening often run on the same underlying software, which is part of why the two get confused.
But the query a recruiter types to find you is not the same event as the filter your resume runs through after you've applied. A search is initiated by a person looking for people, before anyone has decided to reject anyone; a screen runs against applicants who already opted in by applying. Writing for one well doesn't automatically write for the other, and advice that treats them as one problem tends to serve neither.
What each of the four fields has to carry:
- An accurate, current job title, in the field's own vocabulary
- Location, set to where you'd actually work
- Recency: when the profile was last updated or active
- The plain-language name of your work, not an internal title
Don't Stuff Your Profile, Match It
Put the terms where the work is described rather than in a band at the top. Workday's own documentation states its skills-suggestion feature reads the entire uploaded resume, not only a Skills section, when inferring what to recommend 3. A term inside a real bullet registers.
A wall of unearned keywords registers too, in the sense that the search still returns it, and that's exactly the problem: being returned by the query was never the hard part. Terms with no supporting sentence around any of them read as padding to the recruiter who opens the result a moment later, and that reader is the one deciding anything.
Employers running keyword-based screening are told the exact-match era of that particular tool is already ending, because writing to a scored vocabulary now costs anyone seconds with a model open. A recruiter's search survives that shift better than a screen does, since a person still reads what comes back rather than a rule deciding automatically, but a profile obviously written for the machine reads exactly as generic to the person as it does to the query. The match gets you into the room. It doesn't get you a callback once you're in it.
Write the One Line That Gets You Opened
Write your headline in the vocabulary a recruiter would type into the search box, not the internal title your employer gave the role. "Senior Backend Engineer, Payments" is a string a hiring team is actually typing. "Member of Technical Staff II" is one almost nobody types outside your own company, however senior it sounds internally.
The summary line under the headline does different work: it's what a recruiter reads before deciding whether to open the full profile. One sentence naming the kind of problem you work on, in plain language, does more there than a paragraph of adjectives, because it answers the question the recruiter is actually holding: is this the kind of person I'm looking for.
The same rule holds for any AI-fluency claim you add to that headline or summary. Employers checking a line like "uses AI daily" are told to verify it in about ten minutes, off an artifact rather than the claim itself, so a vague line adds a keyword without adding anything a ten-minute check would confirm. Name the tool, the task, and something you could point to if asked, since that specificity is what survives the ten minutes, and the vagueness is what a stuffed profile can't.
When Inbound Won't Carry Your Search
Inbound sourcing favors certain titles, seniorities, and locations by design, and for some roles and career stages it will never carry a search on its own. A recruiter's query is built out of titles, years and specifics, and an early-career profile has less of each, so a thin work history gives a search little to match against however the profile is worded. Treat inbound as a complement to applying, not a replacement.
Which employer you're trying to reach also matters more than which software sits behind their hiring. In a correspondence study that sent more than 83,000 fictitious applications to entry-level jobs at 108 of the largest US employers, which third-party application-system vendor a firm used explained only 0.1% of the variation in racial contact gaps between firms, and the authors conclude that screening algorithms are unlikely to drive the differences they measured between employers 4. The software a company bought is rarely the thing standing between a strong candidate and a reply. Where you're looking, and whether that employer is looking at all, more often is.
Common questions
Does a recruiter's search work the same way as a resume screen?
No. A search runs against a database and narrows on fields like title, location and recency before a person opens anything, while a screen usually runs after you've applied to a specific posting. They can share the same underlying software, but the query that finds you and the filter that screens you are different events.
Will keyword-stuffing my profile help recruiters find me?
Not reliably, and it can work against you. Matching stopped being the barrier once writing to a scored vocabulary got cheap, and a profile that reads as stuffed to a person reads as generic the moment it's opened, which is the step that actually decides anything.
Should I list my exact internal job title or a more standard one?
Lead with the standard, plain-language version. An internal title like a leveling code returns almost nothing outside your own company, while a title in the vocabulary a recruiter would actually type is what makes you appear in the result set at all.
Can inbound sourcing replace applying to jobs?
For some people and roles, mostly no. Inbound favors certain titles, seniorities and locations, and a recruiter's query is built out of titles, years and specifics that an early-career profile has less of. Treat inbound as a second channel worth setting up correctly, not a substitute for applying.
Should I put 'proficient in AI tools' on my profile to get found?
A vague line like that adds a keyword without adding anything checkable, and employers verifying such claims are told to do it fast, off a specific artifact. Naming the tool, the task, and something you'd point to survives that check; a generic claim doesn't.
References
- 1. Talent Filtering (Greenhouse Support) support.greenhouse.io Supports that one named vendor's recruiter-facing keyword filter runs on an exact match against the full text of an application, the mechanical basis for naming your work the way the field names it. Cited as Greenhouse's own wording for its own application-review tool, not generalised to all search.
- 2. Hidden Workers: Untapped Talent hbs.edu Supports how widely recruitment management systems were reported in use and used to filter or rank candidates, the software layer sourcing and screening share. Employer self-report from 2,275 senior leaders in three countries, fielded early 2020; cited with that cohort and date attached.
- 3. Setup Considerations: Skills in Recruiting (Workday Administrator Guide) doc.workday.com Supports that skills inference reads a resume's full text rather than a Skills section alone, correcting the idea that only a keyword block gets read.
- 4. Systemic Discrimination Among Large U.S. Employers (NBER Working Paper No. 29053, revised May 2022) nber.org Supports that which third-party application-system vendor a firm uses explains only 0.1% of the variation in racial contact gaps between large US employers, so the employer matters more than the software. Cited with the study's own outcome measure and cohort, not widened to response rates generally.
4 sources, numbered by first appearance. How Olive sources claims
General guidance for hiring teams. What works at one company and one volume may not transfer to yours.
Olive assesses how a person works with AI. It does not detect AI-written documents, and it never produces a score, a ranking, or a match percentage for a person. Candidates read the same report the employer reads.