You open an email. Subject line: "Jobs that match your profile."
That little hit of hope. Maybe this time.
You click. You scan. And then the sigh. Three product manager roles. But you're a project manager. Close, but not close. Again.
You tell yourself the algorithm is broken. LinkedIn doesn't get you. The system is dumb.
But what if the system isn't broken? What if there's a gap between what you meant to say and what's coming through?
How Search Actually Works
Every search system (Google, LinkedIn, your company's ATS) does two things: discovery and ranking.
Discovery answers: are you in the pool at all? Do you exist for this search?
Ranking answers: now that you're in the pool, where do you show up in the results?
Two different problems. Two different dynamics. And they've changed.
The old way, keyword search, was simple. Discovery meant: does the exact term appear on your profile? If the recruiter searched "project manager" and you called yourself a "delivery lead," you didn't exist. You weren't in the pool. The system was matching strings, not meaning.
Ranking was equally blunt: how much of the keyword did you have? The profile with "project manager" in the headline, the about section, and six job titles outranked the one that mentioned it twice. Density won.
This created problems. Vocabulary mismatch kept good candidates hidden. And false positives ran rampant.
You recruited for lab roles, so your profile was saturated with GLP/GMP, PCR/qPCR and assay development. The system saw all that terminology and thought: Research Associate. You got discovered for roles you never did, just because you talked to people who did them.
The Semantic Shift
LinkedIn's 360Brew algorithm, and many modern search systems, have moved toward semantic search. Instead of matching strings, the system tries to understand meaning.
This is genuinely good news. "Project manager" and "delivery lead" can now be understood as related concepts. You don't have to guess the exact words a recruiter will use. The system can bridge the gap.
But here's what didn't change as much as you'd think: ranking.
You still have to compete against everyone else in the pool. And the system still needs to decide who shows up first.
In semantic search, your LinkedIn profile is viewed all at once. And the algorithm is attempting to infer how to categorize you from your headline, the about section, your experience and so on.
If your headline says recruiter, your about section talks about how you recruit, and your experience shows what you recruited, then the algorithm will categorize you correctly.
Now let's take the same recruiter, and add the role of Career Coach. Categorization becomes less clear and if the algorithm is ranking profiles, then it will order the list based on the profile that has the clearest message.
Even with semantic search there is a difference between discovery and ranking. Discovery asks: can the system find you at all? Ranking asks: once you're in the pool, where do you show up?
The recruiter who added a career coach role might still get discovered for recruiter searches. But they'll rank below the person whose profile only says recruiter.
Coherence isn't just about getting discovered in the right pool. It's about how strongly you register once you're there.
What This Means for You
If semantic search works as intended, the promise is real: better matches on both sides.
For you, the job seeker: more confidence that the roles surfacing actually align with your experience. Fewer of those deflating emails full of irrelevant suggestions. The system is trying to understand what you are, and if your profile clearly signals that, it can find you for the right things.
For the recruiter: a higher-quality candidate pool from the start. Less time sifting through profiles that keyword-matched, but don't actually fit. More time evaluating genuine candidates instead of sorting through noise.
But this only works if the signal is clear. Garbage in, garbage out still applies. It's just that the garbage is now incoherence instead of missing keywords.
The Recipe Card
Your LinkedIn profile isn't a form to fill out, section by section, checking boxes. It's a recipe you're making. And every ingredient either serves the dish or distracts from it.
A taco has a few essential elements. Protein, tortilla, salsa, maybe some cilantro and onion. Every family makes it differently. Carne asada or carnitas. Corn or flour. Red salsa or green. But it's still recognizably a taco. You taste it and you know what it is.
If you throw in marinara sauce, mozzarella, and basil because they happen to be in your fridge, you've made something. But it's not a taco. It's not a pizza either. It's just...confusing.
Your profile is the same. You can have your own version of the recipe. Your emphasis on stakeholder management versus someone else's emphasis on technical delivery. Those are family recipes. Both taste like project manager.
But if your project manager profile has product terminology from adjacent work, engineering jargon from roles you recruited for, and marketing language from a side gig, the dish doesn't taste like anything the system can identify.
The question isn't "what's the correct recipe?" It's "what dish am I making, and does every ingredient serve that dish?"
The Recipe Box
Each role you have held is like a recipe card. The knowledge, skills and abilities needed to perform the role get added like ingredients to your career. Over time with enough roles your recipe box gets full.
But you are not just a recipe or even a recipe box. You are the chef.
However the market seems to demand ingredients. This role requires an undergrad. To be successful in this role you need 10 plus years of experience in product lifecycle management. Or the ideal candidate has managed projects in a highly regulated Class III environment.
The more specific the role you are applying for, the more explicit your resume needs to be. This has built the habit of completeness. Even a LinkedIn profile partially inherits its nature from the resume. In some ways it has become a digital proxy. The profile like the resume list every job, every skill, every responsibility. Don't leave gaps. Show them everything.
Completeness is the keyword insurance policy that a jobseeker takes out. But semantic search doesn't reward completeness as much as it rewards coherence.
You are the chef of your own career, you get to decide what you want to serve to the market. It doesn’t have to be every dish you can make. It’s the dish you're choosing to make right now, for this audience.
For example, if a recruiter adds career coach to their profile, the signal becomes less clear. This could be a natural career pivot or it could be a temporary role. Regardless, these search systems don’t interpret intent.
However, knowing how you are discovered and ranked gives you agency. There are a lot of recipe cards in your recipe box. The point isn't what to choose. The point is choosing with awareness of what the system sees.
You're the chef. You decide what to serve.
For this moment. For this audience.