The Problem Isn't Finding People. It's the First Ninety Days.

Ask almost any employer what is wrong with hiring right now and the answer comes back in the same shape: not enough people. Not enough applicants, not enough qualified applicants, not enough who show up. The response follows the diagnosis — more postings, more sourcing, more screening, more spend at the front of the funnel.

The national numbers do not support that diagnosis. The U.S. Chamber of Commerce, which does the comparison carefully, puts the country at roughly 100 available workers for every 100 open jobs. The Bureau of Labor Statistics counted 7.3 million openings in July, and the quit rate sat at 1.9 percent — very few people are leaving one job for another. Its August jobs report found that 27 percent of unemployed Americans had been looking for longer than six months.

Put plainly: the jobs exist, the people exist, and they are not staying connected. If the problem were a shortage, more sourcing would have fixed it by now.

The part nobody measures

Hiring is measured at the front. Time-to-fill, cost-per-hire and applicant volume are the numbers that get reported upward, so they are the numbers that get managed. What happens after the offer — the gap before the start date, the first schedule conflict, the first week a child care arrangement falls through — mostly goes unmeasured, and what goes unmeasured goes unmanaged.

Employers are not alone in this. This month the International Labour Organization reported on a working session of public employment services and found that their digital tools concentrate on job search, skills and matching, with much less evidence on supporting people once they are actually in the job. The institutions built to connect people to work have, by their own account, the least evidence about the stretch where those connections break.

What breaks it has a name, and a price

When a new hire does not last, the explanation offered is usually about the person: reliability, commitment, fit. The data points somewhere more concrete.

The U.S. Chamber of Commerce puts child care at an average of $343 a week per child and reports that 58 percent of working parents have left a job because child care did not work out. In the Houston area the supports that make that math possible are getting thinner: Children at Risk reported in August that a 9.5 percent cut to the Texas Workforce Commission's child care scholarship targets will mean more than 1,000 fewer subsidized seats across Greater Houston, in a region where 169 ZIP codes already have at least three times more demand for subsidized care than supply.

That is not a hiring problem. A parent in that position has already been hired. It is a staying problem, and no job posting solves it.

The Bureau of Labor Statistics has now priced the other half of a job — the part that is not the paycheck. Benefits make up 31.6 percent of what an hour of work costs an employer: $15.61 of $49.46. In restaurants and hotels, where a great many people start, an hour of part-time labor costs an employer $17.21 against $27.95 for full-time, about 62 cents on the dollar. The cheaper hour is cheaper largely because it carries less of what keeps a person in the job.

The same release sets out who can afford that half, and the answer is uncomfortable. Within restaurants and hotels alone, an employer with fewer than 50 workers at the site spends $3.05 an hour on benefits; one with 500 or more spends $8.69. Same industry, same quarter, nearly three times the cushion. Most first jobs are at the small end. So the work of holding on to a new hire is least affordable in precisely the places where the most new hires start — which is a constraint on the recommendation that follows, not an excuse for skipping it.

A record year that stopped short of the first rung

On September 15 the Census Bureau reported that median household income reached $87,460 in 2025, the highest in a series that begins in 1967, up 2.6 percent after inflation. That is real, and it is good news.

It is also not evenly distributed. In the same report, household income at the 90th percentile rose 1.7 percent — a third consecutive annual gain — while income at the 10th percentile did not change by any amount the Bureau will call statistically significant. Among full-time, year-round workers, women's median earnings rose 3.2 percent and men's did not move measurably.

For an employer, that is a staffing fact rather than a political one. The households an entry-level workforce comes from had a flat year inside a record one. Whatever made this a good year for the American median did not reach the first rung of the ladder — which is exactly where turnover is most expensive and most preventable.

The cost of a broken connection is rising

The consequences are getting steeper for workers, too. The Bureau's Displaced Workers survey found that 3.3 million people lost jobs they had held for at least three years between 2023 and 2025. By January 2026 two-thirds were working again — but among those who had lost full-time jobs and found new ones, only about 49 percent were earning as much as before, down from about 62 percent in the previous survey. And an hour of pay bought 0.3 percent less in August than a year earlier.

A worker who loses their footing today falls further than one did two years ago. That raises the value of keeping them — for the worker, and for the employer who already paid to find them.

What follows

None of this argues against recruiting well. It argues that the return on the next dollar is no longer at the front. The employers who pull ahead will treat the first ninety days as part of the hire rather than as the new employee's problem. They will know which new hires depend on a bus line or a subsidized child care seat before the first missed shift, build the start-date conversation into onboarding, and report ninety-day retention alongside time-to-fill.

Blue Chip Works is built around that whole stretch — attract, hire, onboard, support, develop, retain. Finding people has never been the hard part. The first ninety days is.