Triplebyte and Hired were the two best-funded attempts to build a candidate-side hiring marketplace, and they died of the same three things. Both are worth reading as one case, because the shared cause of death is what makes the current generation — Paraform and its peers — structurally different rather than merely newer.
The two records
| Triplebyte | Hired | |
|---|---|---|
| Raised | ~$50M, incl. $35M in 2019 | $132.7M across 6 rounds, last $30M in 2018 |
| Peak mark | $135M (PitchBook estimate, 2019) | ~$500M (2020) |
| Revenue | Never disclosed | $11.5M (2021) across ~10,000 customers ≈ $1,100 ACV [WEAK] |
| End | Assessment products and content library sold to Karat, March 2023; the "Magnet" talent network wound down that month; terms undisclosed | Acquired by Vettery, Nov 2020 (undisclosed); merged under the Hired brand March 2021; now inside Adecco's LHH |
| Adjacent price | — | Vettery sold to Adecco for over $100M in Feb 2018 |
Sources: TechCrunch on Karat/Triplebyte; GetLatka on Hired [WEAK] — self-reported/estimated; Wikipedia; TechCrunch on Vettery/Hired; SIA.
A $500M valuation against $11.5M of revenue is roughly 43x — on a business with no recurring revenue, in a sector whose public comps trade at 0.2–1.3x EV/Revenue (Multiples.vc). That gap is the entire story of Hired, and $150M+ raised across the two companies made a $100M outcome a failure rather than a success. See What the public market pays for labour.
Triplebyte: growth was the solvent
The post-mortem comes from ex-employees in the Hacker News acquisition thread (HN 35184546):
- Candidate acquisition cost scaled superlinearly. They bought candidates with ads, and "the marginal cost of ad conversions grows as you try to scale." The high-touch process that made the product good was "extremely expensive in ways that scaled poorly."
- VC growth pressure forced pool expansion, which destroyed the value proposition. Broadening the funnel to hit targets meant "standards slid" — employers reported candidates who "could barely squeeze out a line of code" passing screening. Once the screen is not trustworthy, the premise — we pre-vet so you do not have to — is gone.
- It never became the primary channel. Employers treated it as a secondary source, and secondary sources get the leftover reqs and the leftover urgency.
- The 2021 public-profile privacy incident did not kill the company, but "was asked about during basically every one of the acquisition conversations," and the ad-spend cut that followed worsened supply.
- A former employee said they "were simply never able to make the marketplace version of Triplebyte work."
Hired and Vettery: a bet on a permanently hot market
The analysis is from Underdog.io:
- Hyper-curation did not survive blitzscaling. It worked at small scale; expanding into non-engineering roles and new geographies broke it. Identical to Triplebyte.
- One-time transactional revenue, no recurrence, rising CAC. Every placement required re-acquiring both sides. There is no retention flywheel in a marketplace where the buyer's need is episodic and the seller leaves the moment they succeed.
- The model assumed candidates would stay scarce. A reverse-auction where employers bid for candidates only makes sense in a hot market; "when the tech hiring market softened… their value proposition collapsed."
- Grow-or-die. The capital raised made every realistic outcome a disappointment.
The one disease
Every dead talent marketplace died of some combination of three things: acquisition cost on the scarce side rose faster than the take rate; growth targets forced quality dilution that destroyed the differentiator; and revenue was transactional and cyclical while the cost base was fixed. The first two are the same disease seen from opposite ends — the scarce side was candidates, and candidates are the one kind of supply that permanently exits your market when you succeed for them.
Why this matters for the current generation
Paraform inverts the scarce side. Its supply is recruiters, who are durable, repeat, and improve with use, so the acquisition cost amortises instead of resetting. And the platform does not own vetting quality — an outcome-paid recruiter does — so Triplebyte's specific death spiral is structurally unavailable. That is a genuinely better structure, and it deserves the credit.
What it does not fix is the third cause. Placement revenue is still transactional, still non-recurring, and still cyclical: the US staffing market as a whole fell from $243.9B in 2022 to $180.2B forecast for 2026 — perm placement is 11% of that total, not the whole of it — and the market is still below its 2019 nominal level of $185.5B (StaffingPulse). Hired's failure was, at bottom, a beta problem, and nothing in the recruiter-marketplace design changes the beta. See Contingency recruiting marketplaces and Which side you build first.
What the buyers paid
The exit prices are the most useful part of the record. Adecco paid $100M for Vettery in 2018. Vettery then bought Hired for an undisclosed sum after Hired had raised $132.7M and carried a $500M mark. Karat bought only Triplebyte's assessment technology and content library — the marketplace itself was not bought, it was switched off.
Staffing incumbents mop up failed talent marketplaces at scrap prices, and they buy the assessment IP rather than the network. That is the base rate a recruiting-marketplace exit should be modelled against, not the software comps the rounds were priced on.
Hired's revenue and valuation figures come from a third-party estimator whose data failed spot-checks elsewhere in this research; treat $11.5M and $500M as directionally right and precisely unreliable. Neither Karat/Triplebyte nor Vettery/Hired disclosed terms, so the actual recovery to investors in either case is unknown.