Kevin Wright

Co-Founder @ | International Staffing | Testing Platform for Finding the Best Applicants

[Kevin Wright @kevinwright_vrt]

@Kevihaiceth @mariannehere That self-audit before a hiring push is underrated. You walk in knowing exactly what you're selling candidates.

[Kevin Wright @kevinwright_vrt]

Every RTO mandate gets sold as a culture fix. The retention numbers say it's usually a culture cost.

8 in 10 companies with strict return-to-office policies report losing employees over it. Annualized turnover at those companies runs about 13 points higher than at flexible peers, 169% versus 149%.

If the mandate is really about output, measure output. If it's about control, it would save everyone time to just call it that.

[Kevin Wright @kevinwright_vrt]

@SherryYanJiang @ChatGPT @gabrielchua @unprofeshme Reviewing in a structured layout instead of scrolling through raw chat output is a real productivity difference. The visual organization alone cuts review time by a third.

[Kevin Wright @kevinwright_vrt]

You don't need to delegate everything. You need to delegate the gap between what a task pays and what your hour is actually worth.

[Kevin Wright @kevinwright_vrt]

Try this for a week: log every task you do, then guess what an hour of that specific task is worth on the open market.

[Kevin Wright @kevinwright_vrt]

Most founders find at least a third of their week is spent on work worth a fraction of their own time. Not because they're bad at delegating. Because nobody ever made them do the math.

[Kevin Wright @kevinwright_vrt]

It's not just us seeing this. Industry-wide, 22% of accepted offers never show up for day one, and restarting that hire costs about $4,700 on average, per CNBC.

That's the bill hiding behind our 6.3%.

[Kevin Wright @kevinwright_vrt]

A stat that surprised even us: 6.3% of everyone we've ever hired never worked a single day. They accepted the offer, confirmed the start date, then never showed up.

Most companies never see this number because those people get coded as normal churn. We track "never started" as its own category now.

Someone who quits in month 3 and someone who ghosts on day one are completely different problems, and mixing them hides both.

[Kevin Wright @kevinwright_vrt]

Software engineer placements out of LatAm grew roughly 250% year over year, the fastest-climbing role we track.

Cost gets all the attention, but speed compounds just as hard. Nearshore hires close about 2x faster than US recruitment, averaging 2-6 weeks per role. Companies aren't only saving money. They're saving whole quarters of an open req sitting empty.

[Kevin Wright @kevinwright_vrt]

Medical practices are some of the last holdouts on remote admin support, usually over compliance concerns.

Most of what eats a front-desk day, scheduling, insurance verification, follow-up calls, doesn't require anyone to be in the building. It requires someone trained on the process.

[Kevin Wright @kevinwright_vrt]

This isn't just us. Talview's 2026 AI Threat Index found 88% of assessments now face active AI cheating risk. TestPartnership found 83% of candidates would use AI if they believed detection was unlikely.

The test format matters more than the proctoring software at this point.

[Kevin Wright @kevinwright_vrt]

AI quietly broke online hiring tests. Applicants can now fully cheat the verbal and numerical sections of a cognitive assessment, and in an unproctored online test there's no real fix.

So we rebuilt ours around what AI can't do fast: image-based spatial puzzles with a time limit measured in seconds per question, not minutes. Too fast to screenshot into ChatGPT.

Unexpected side effect: cutting the test from half an hour to a few minutes nearly doubled completion.

[Kevin Wright @kevinwright_vrt]

The stereotype is that nearshore hiring means entry-level work. The placement data disagrees.

84% of recent LatAm placements were mid-level or senior. A third were senior, including VP and director titles. Software engineers hit a 98% placement rate at the experienced tier.

Companies aren't nearshoring the easy roles anymore. They're nearshoring the ones that sat open for six months domestically.

[Kevin Wright @kevinwright_vrt]

Fewer meetings doesn't fix a communication problem. It just delays it.

The teams that actually run lean on meetings are the ones that got obsessive about writing decisions down, not the ones that got better at scheduling.

[Kevin Wright @kevinwright_vrt]

@chamath Anthropic's scaling laws paper gave everyone the blueprint to catch up. Slowing capability gains while keeping research closed doesn't erase the knowledge, it just controls who profits from applying it first. That's not safety, that's market positioning.

[Kevin Wright @kevinwright_vrt]

@Jason robotaxi economics only work when the driver cost hits zero. that's the whole model. Alphard fits Black because those riders already pay for the experience, not just the destination. which city actually proves this out first is the real question worth watching.

[Kevin Wright @kevinwright_vrt]

@garyvee comments are a feedback loop, not a vanity metric. watch time and CTR tell you what happened. comments tell you why. the silent majority rarely speaks, but when they do, those sentences are gold. question is: how do you separate real signal from the noise?

[Kevin Wright @kevinwright_vrt]

For years our data showed a strange pattern: hires scoring between roughly the 70th and 85th percentile on our cognitive test, sitting right around one standard deviation above the average applicant, quit more than any other band. I had a whole theory about above average people getting bored in mundane roles.

We finally tested it properly. Hold the client constant and the effect disappears entirely. That band wasn't quitting more. They were disproportionately being sent to the clients everyone quits on.

I believed a fake pattern for years because I never controlled for the right variable.

[Kevin Wright @kevinwright_vrt]

@Codie_Sanchez most owners are two archetypes stacked, not one. the "good guy running a one stop shop" combo is basically a discount machine with no positioning and no identity. that pairing predicts burnout faster than any single type ever could.

[Kevin Wright @kevinwright_vrt]

@APompliano values first only works if you have a real system to test for it. most companies don't. they run a 30 minute vibe check and call it culture fit. the part worth stealing from Koch's approach is whatever behavioral framework actually sits behind that principle.