What makes a good LinkedIn profile?

I started heavily editing my LinkedIn profile until the recruiter messages stopped, the opposite of what I wanted to happen.

Imagine. You’ve just been laid off and are working through your resume, case studies, a website redesign, and—of course—updating LinkedIn only to find that your edits coincide with a complete drop in inbound messages from recruiters.

I’ve never had a huge amount of inbound messages but usually about one or two a week asking for resumes or if I’m interested in specific roles.

The reason I was even editing my LinkedIn may be the reason I don’t finding a new job!

I did what anyone in the tech industry would do, I set Claude to work

I used a workflow and fan-out research to understand how recruiters search for candidates, what content they actually see, and used that to re-re-edit my profile. I now have inbound messages again.

GitHub - craigmdennis/job-search-skills: Installable agent skills for the job searchInstallable agent skills for the job search. Contribute to craigmdennis/job-search-skills development by creating an account on GitHub.

GitHub

What makes a good linkedIn profile?

Put the exact title you want next into the job-title field of your current role. The ‘Job titles’ filter reads that field literally and expands nothing.
Repeat the same three or four literal terms across the headline, the About section, and your current-role description. Exact matching decides most results.
Put your real title and top keyword in the first 60 characters of the headline. The field caps at 220 and roughly the first 60 to 70 show on mobile and in search results.
Set a real city or region as your location. A bare ‘Remote’ registers as empty to the geography filter.
Fill the Skills section and pin the right top three. LinkedIn’s own data says listing skills produces up to twice the profile views and up to four times the messages.
Rewrite your role bullets before you touch the About section. The first pass reads titles and bullets, and at least one recruiter says they only read the summary when nothing else tells them enough.
Keep your sharpest line inside the first 300 characters of the About, before the ‘see more’ fold.
Pin your portfolio and one case study to Featured.
Turn ‘Open to Work’ on in the private ‘Recruiters Only’ mode. The ‘visible only to you’ setting does nothing for discoverability.
Strip the buzzwords listed below, and any claimed skill with no number or artefact behind it.

Buzzwords to avoid

Every source I read landed on roughly the same words. These are the ones recruiters skip:

‘passionate’
‘results-driven’ and ‘results-oriented’
‘strategic’
‘innovative’
‘expert’
‘creative’
‘hard-working’
‘guru’, ‘ninja’, ‘rockstar’
‘open to opportunities’, and other status phrases

A buzzword study says to replace each one with a number or a concrete example, and none of them are searchable anyway, so in the headline they take up room a real keyword could have had.

‘Results-driven professional with a proven track record of leveraging innovative solutions’ is called out as the tell that a profile was written by AI, and multiple 2026 pieces report that generic, over-polished copy is now actively penalised rather than just ignored as a result.

While I was working through all this I built a way to visualise what a recruiter sees. Type a headline, an About blurb, and a few role bullets then see which parts show where. It is loaded with my own current profile; edit any field and the previews and the checks update as you type.

LinkedIn Validator

Type into any field. The search result and the phone show far less than the desktop view you edit against, which is why the first 60 characters have to carry the terms.Open standalone ↗

The full research details

What a recruiter is actually typing into the box

A recruiter does not read profiles to find you; they run queries. LinkedIn’s own Recruiter Help page describes a keyword search that highlights matches in a specific set of places: the profile card, the Summary (About) section, the Experience section (title, description or location), and the Skills section.

Recruiters build Boolean strings like ("Staff Designer" OR "Principal Designer") AND ("B2B" OR "SaaS") and match mostly on the title text. The tool matches text and cannot infer what you meant, so if your keywords do not match, you do not come back in the results.

A separate ‘Job titles’ filter behaves differently from the free-text keyword field. It searches the structured job-title field on your Experience entries, with literal matching and no synonym expansion. ‘Product Lead’ does not surface for a ‘Product Manager’ search unless that exact string is in that field.

My last title reads ‘Head of Product Design’, and one sourcing writeup claims LinkedIn’s seniority filter only recognises four words at all: Manager, Director, Vice President, Chief. If that is right, the leadership filter does not match ‘Head of’ and it is not an IC keyword either, which leaves the title matched by neither.

How to show up for remote work

A third filter matters more to me, because I am after remote roles, and that’s location. LinkedIn’s geography filter reads a structured location, and a bare ‘Remote’ registers as empty, so recruiters sourcing remote roles search a timezone-qualified string instead. If my profile just says ‘Remote’, it never comes back for that search.

The keyword matching then feeds a ranking model, and LinkedIn’s engineering blog is unusually specific about it. Results are ordered by a “multi-pass machine-learning system using gradient-boosted decision trees with a pairwise objective” (whatever that means). It scores work-experience and skills similarity, job-posting location, and how likely you are to respond.

Semantic matching (surfacing ‘Software Developer’ for a ‘Software Engineer’ search) exists, but the blog describes it as “query expansion” that applies mainly when few profiles match exactly.

Exact-term matching decides most results, and the clever synonym-understanding you might be hoping for is a fallback. So repeating the same three or four real terms across the headline, About, and current-role description works better than writing for variety.

Sourcing guides go further and call the headline “the single most heavily weighted field on the profile”, recommending a keyword-dense ‘target role, domain, differentiator’ line over any personal-brand tagline. LinkedIn’s own docs never single the headline out as its own matched field though, so your mileage may vary.

A few smaller details emerge from the same primary sources. LinkedIn’s Talent Blog says listing skills produces up to twice the profile views and up to four times the messages, so populating that section and pinning the right top three is one of the few changes LinkedIn’s own data supports.

The public green ring is broadly destigmatised now and LinkedIn’s data says it increases outreach on average, but one analysis adds a senior-level exception: at leadership levels a public availability signal can clash with a ‘selective and in-demand’ read, and private, recruiter-only signalling is suggested instead.

LinkedIn’s help page states the private ‘Recruiters Only’ (obviously) helps surface you in recruiter searches. Being open to work then supplies a ‘Spotlight’ filter a recruiter has to actively apply, alongside an ‘Active talent’ spotlight that ranks recent profile activity and resume-sharing higher as a likely-to-respond signal.

Front load your headline

The headline field itself caps at 220 characters with roughly the first 60 to 70 shown in mobile and search previews, though I could only corroborate that number across secondary trackers rather than a live LinkedIn page.

Here is where I had to be careful. A large amount of ‘2026 LinkedIn SEO’ content states headline plus current position equals sixty per cent of ranking, verified skill badges rank thirty per cent higher, and a good SSI score produces recruiter messages. When Claude traced those figures they came from career-coaching and lead-generation blogs with no primary source behind them. So take them with a pinch of salt.

The first pass takes six seconds and skips your About section

Getting surfaced is only the start. A recruiter still has to read the profile.

Recruiters interviewed for one guide describe a six-to-ten second first pass that skips the About section entirely and reads job titles and bulleted scope instead. One said they do not read the summary unless there is not enough information elsewhere.

The Featured section turns out to be underused too. One piece reports a majority of surveyed recruiters calling it a ‘difference-maker’.

The fewer clicks it takes to find your work, the better. — One recruiter regarding the ‘featured’ section

Some filtering also happens before a human is involved at all. Recruiters can set a minimum-tenure filter that removes short-stint profiles from results, though the same guides note that a short stay reads as advancement rather than job-hopping if it came with a title bump or new scope. Again, not an exact science.

The language recruiters usually skip is also consistent across every source Claude found, which is where the list above comes from. The same goes for any skill claimed with no metric or artefact behind it. It all reads as filler or as a credibility risk.

One rubric I deliberately do not want to be measured against comes from the widely-cited piece on hiring design leaders at director level and above, which argues that ‘craft matters less than team-building, hiring practice, executive advocacy, and operations’. Worth knowing before optimising a profile towards it (or away from it, by accident).

Which title actually gets searched for

I also have to decide what to call myself, which is not a branding choice so much as a demand question.

Boutique retained firms exist whose entire product is placing Head of Design and VP Design roles, and cross-checked Glassdoor counts show Head of Design and Design Director as their own tracked categories in the low hundreds.

‘Staff Product Designer’ is a real and growing title at AI-native B2B SaaS companies, with live postings and a Figma study reporting far higher demand for senior-plus hires than junior ones. ‘Principal Product Designer’ is thinner and skews towards large, scaled companies, and career-track analysis says it is mostly filled by internal promotion, which limits how often an external recruiter search returns it.

‘Founding Designer’ is the thinnest of all on recruiter-search volume because that role is usually sourced through founder networks and startup boards rather than recruiters.

Two signals really matter for the individual-contributor track. A 2026 analysis argues the Senior-to-Staff promotion path broke, because the Staff job now demands judgement about whether to build a thing at all, and Senior designers were never given a reason to develop it.

This means Staff roles increasingly get filled by external and lateral hires. That points to external search being more active at Staff level than at Principal, though I have no search-volume figure for either.

‘AI fluency’ now reads as a requirement across every one of these framings, with Figma reporting designers who have AI skills earning notably more than peers who do not. It does not favour one title over another, but it strengthens whichever title you put first.

Generic copy is now penalised

The best-practice writing advice suggest a recurring headline formula is:

{{target role}} ・ {{scope or scale}} ・ {{outcome}} ・ {{the exact keywords a recruiter would type}}

An example would be:

Head of Product Design • Built the design function twice • AI Voice UX • Design Systems • B2B SaaS • AI Transparency = User Trust • Remote preferred (EU/US overlap)

All while trying to get title and scale before the mobile truncation point. Try yours out above, and keep the buzzwords out of it.

The About section follows a Hook, Proof, Close structure, with the first roughly 300 characters carrying the hook before the ‘see more’ fold and three to five quantified proof bullets rather than adjectives. Role bullets follow a Challenge, Action, Result pattern where the specificity is the whole point: not ‘reduced costs by 15%’ but the exact system and mechanism that did it.

Specificity matters for a second reason, and it is the one behind the buzzword list at the top. Generic, over-polished copy is now actively penalised rather than just ignored, so every stripped adjective needs a number or a concrete example in its place.

For anyone framing towards a senior individual-contributor scope without implying people-management, Intercom’s public writeup of its IC design ladder is worth borrowing from for vocabulary. It describes Staff and Principal influence as scope across teams and a quality bar raised through mentoring and strategy, rather than headcount.

What I have, and what I still do not know

I now have a description of the mechanics drawn from primary sources: how the keyword fields match, how the ranking model weights exact terms over synonyms, which title has recruiter demand behind it, and which words recruiters skip.

So I made the changes. I rewrote the headline, current-role title, About opening, and top skills around the same three or four literal terms, pinned the portfolio and a case study to Featured, and flipped the private ‘Recruiters Only’ signal on.

Within two weeks I had three recruiter messages about potential roles.

I cannot tell you the profile did that. Three is a small number. It could be people coming back after the summer break. It could be confirmation bias, because I went looking for a result I wanted.

For anyone else in the middle of a search, it can’t hurt to try.

I turned the research into a skill

Reading a list of rules and applying them to your own profile are two different jobs, so I packaged all of this as an agent skill. It takes your existing profile, or interviews you if you are starting from nothing, pulls the literal search terms out of two or three job ads you would actually apply for, and writes every field ready to paste in. It works for any job, not just design ones.

In Claude Code:

/plugin marketplace add craigmdennis/job-search-skills
/plugin install linkedin@job-search-skills

Then run /linkedin:profile.

For any other agent:

npx skills add craigmdennis/job-search-skills

The code is at craigmdennis/job-search-skills. Every rule it follows carries the source it came from, and the claims I threw out are listed in there too, so you can check the reasoning rather than take my word for it.

Sources

What LinkedIn documents itself:

Everything else:

I used Claude Code for the research (multi-agent fan out plus a red-team pass) and to build the tool.

What makes a good linkedIn profile?

Buzzwords to avoid

The full research details

What a recruiter is actually typing into the box

How to show up for remote work

Which ‘Open to Work’ banner to use

Front load your headline

The first pass takes six seconds and skips your About section

Which title actually gets searched for

Generic copy is now penalised

What I have, and what I still do not know

I turned the research into a skill

Sources