EMeetESign in

Why job matching needed rebuilding from the ground-up, not just another filter.

Seek, TradeMe, and LinkedIn all match jobs to people using keywords, matching words in a CV to words in a job ad. EMeetE reads and digests both in full, checks the requirements that actually matter, gives the match a score out of 100 and tells you why a job is or isn't a fit. That's the entire difference, and it changes everything.

How the approaches actually differ

ApproachSeek / TradeMe / LinkedInEMeetE
Matching methodKeyword and title overlapFull CV and job description read and digested, scored 0–100 with clear stated reasoning
Hard requirementsNot distinguished from preferencesLocation, experience, certification, and visa requirements checked before scoring
OverqualificationNot considered will blindly show junior roles to senior candidatesFlagged explicitly, and can be adjusted per user
Duplicate listingsSame role often shown multiple times across sourcesOne canonical listing per real job, regardless of source
Why a job matchedNot shownPlain-language reasoning on every result
Search precisionFixed keyword searchNo search bar. You've already done the hard work in your CV. A single sliding scale shows your exact-match to open-to-anything
Cost to job seekersFree to search, but time cost is high (see below)Free
Cost to employersExpensive job listings that results in hundreds of unvetted CV's (see below)Receive a focussed list of actual potential applicants

We tested 150 'recommended' jobs. One was real.

Over a couple of weeks, we collected every job Seek, TradeMe and LinkedIn actively recommended to a real, detailed 20+ year graphic designer and project management CV. Not jobs we searched for, jobs their own algorithms chose to push and recommend. Then we scored all 150 properly; hard requirements checked first, domain and seniority fit assessed honestly, no keyword shortcuts.

150

Jobs recommended by Seek, TradeMe and LinkedIn's own algorithms.

1

Genuine strong match. One other result scored over 75. It was the same job, posted twice.

The 'recommendations' included: a drainlaying role (2/100, zero trade skills or experience), a mining processing supervisor (2/100, no mining industry background whatsoever), a mechanical maintenance engineer (2/100), an architectural builder/carpenter (2/100), and a commercial drone pilot requiring a CAA licence the candidate has never held (4/100).

One role, a delivery lead position at a healthcare provider, was recommended three separate times: twice via LinkedIn, once via TradeMe. Same job, same employer, counted as three different opportunities.

New Zealand's unemployment rate hit 5.6% in the June 2026 quarter. This is the highest in 11 years, with 166,500 people out of work and 440,000 underutilised. Job seekers in this market are routinely submitting well over a hundred applications per listing. The platforms charging employers to post into that volume have no incentive to make the matching any more precise than a keyword search; the fee is the same either way.

It costs employers too. Every one of those hundred-plus applications lands on someone's desk, and most of it is exactly this kind of noise; candidates who were never a genuine fit, sorted by a keyword match rather than an actual read of the role. Hiring managers lose real hours to it, on top of everything else the job already asks of them. Nobody currently paying for a listing is being served well either.

How a score is actually produced

  1. 1

    Your CV is read in full and digested, not keyword-indexed

    Overlapping roles, freelance work alongside permanent jobs, and genuine timeline gaps are handled explicitly, not summed up incorrectly.

  2. 2

    Hard requirements are checked first

    A missing degree, certification, or work-visa requirement weighs heavily on the score before anything else is considered. Strong skill history can't distract from a genuine disqualifier.

  3. 3

    Seniority mismatch is named, not ignored

    Being significantly overqualified for a junior role is treated as a real mismatch, with the option to turn this down explicitly if you're genuinely open to it.

  4. 4

    You get the reasoning, not just the number

    Every match includes the specific detailed summary behind the score. What fits, what's a genuine gap and what would need to change.