Great talent is often hidden in plain sight.
Kristina Golovko
MindDesign
The visibility chain
Visibility
Where companies actually look
Signal
What indicators are used to filter
Sourcing quality
Depth of search logic
Hiring outcome
Who actually gets found
Strong candidates don't hide. Companies search too narrowly to see them.
Why this matters
Most companies search for talent using the same channels, the same keywords, and the same profile assumptions as everyone else. Then they conclude the market is shallow. The market isn't shallow — the search is.
Exceptional candidates are often not on the first page of LinkedIn results. They're not updating their CVs. They're not applying to job boards. They are building things, contributing to communities, and solving hard problems — in places where most recruiters never look.
Invisible talent is not a market problem. It's a search logic problem. Companies that find strong candidates consistently do so because they've built sourcing approaches that reach beyond the obvious — not because they got lucky.
Founder reality
Before assuming the market is thin, interrogate the search:
Are we searching the same channels as every other company in our space?
What signals are we using to filter — and are those signals actually predictive?
Have we defined what a strong candidate looks like beyond their job title and company?
Are we filtering for comfort — familiar logos, recognisable paths — or for capability?
When did we last find a strong candidate who surprised us?
If the last answer is 'I can't remember' — the search logic needs redesign, not more volume.
The framework
Each one has a sourcing fix. The fix requires changing how you search, not just where.
Narrow channel logic — searching where everyone else searches
LinkedIn, job boards, and referrals from the same network produce candidates shaped by those channels. Deep-tech talent, research engineers, and rare specialists are often found in technical forums, open-source contributions, conference talks, and academic adjacent communities.
CV-first filtering — screening on presentation, not capability
CV screening filters for people who are good at CVs. Strong candidates — especially technical ones — often have sparse CVs and rich output. Inverting the filter to look for output first surfaces profiles that CV-first searches miss.
Logo bias — overweighting brand names in candidate evaluation
Filtering by previous employer is a signal proxy — and a weak one. The strongest engineer on a team at a less-known company beats the weakest one at a famous one. Logo bias narrows the effective market artificially.
Passive candidate blindness — assuming active candidates are better
The best candidates are rarely actively looking. If sourcing only reaches people who applied or who are signalling availability, it's missing the majority of qualified talent. Passive sourcing requires different outreach logic.
Common mistakes
01
Concluding the market is shallow after a single sourcing pass
One LinkedIn search pass is not a market assessment. Strong sourcing often requires three to five iterations of search logic before the right profile emerges.
02
Using job title as the primary search field
Job titles vary wildly across companies and geographies. Searching by title misses everyone who does the work under a different name.
03
Prioritising responsiveness over fit
The candidates who respond fastest to outreach are not always the strongest. Response rate is a measure of availability, not quality.
04
Ignoring output signals
Papers, repositories, conference talks, community contributions, and technical writing all signal capability before a CV ever arrives. Treating these as secondary is a sourcing error.
Example scenario
A compiler infrastructure startup, 20 people. Hiring a senior LLVM engineer. Three months of LinkedIn sourcing produced 40 profiles. Zero strong fits.
The problem
Every profile found had a conventional software background. LLVM specialists rarely use 'LLVM engineer' as a title. Most were found only via CVs. No GitHub, paper, or conference signal was incorporated.
The sourcing reset
Search shifted from job title to output: GitHub contributors on LLVM-adjacent projects.
LLVM conference (EuroLLVM, LLVM Developers' Meeting) speaker lists reviewed.
Academic adjacent: PhD candidates and postdocs working on compiler research.
Community: LLVM mailing list contributors, technical blog authors on compiler topics.
The outcome
Six qualified profiles surfaced in two weeks. Two entered the process. One accepted an offer. Total elapsed time from search reset to offer: 7 weeks.
Strong candidates are not rare — they're just not where you're looking.
Changing search logic is faster and cheaper than concluding the market doesn't have who you need. Start with the assumption that the talent exists — then design the search to find it.
Related thinking
Precision Sourcing
Signals Before CVs
Great candidates reveal themselves earlier than most companies realise.
Precision Sourcing
Precision Sourcing Explained
Better hiring starts with better search logic.
Precision Sourcing
Passive Candidate Psychology
Strong people are rarely actively looking.
Continue reading
Precision Sourcing Explained →