Most new hires decide whether to stay before their onboarding is even finished
Enboarder's 2025 survey of HR leaders puts a failed new hire at $25,000 to $50,000, and most of that group never had a documented first week to give them. Here's what a written onboarding plan is actually worth, and how to build one this week.
Losing an employee costs half to three-quarters of their annual salary to replace, by research cited from the Society for Human Resource Management (SHRM, via GMS, 2026). For a $50,000 hire, that is $25,000 to $37,500 gone before the chair is filled again. Most owners already know turnover is expensive. Fewer know the decision to leave, or to stay, is usually made in the first few weeks, long before anyone hands in notice.
That decision does not turn on pay or personality as often as owners assume. It turns on whether the new hire got a real plan or had to piece the job together by watching and guessing.
The failed hire has a price tag, and it is bigger than the ad you ran
Recruiting a replacement is the visible cost: the job post, the interviews, the background check. The invisible cost is bigger. HR leaders surveyed put the cost of a failed new hire at up to $25,000, and the people who own the budget, chief people officers, put it closer to $50,000 once lost productivity and the next search are counted in (Enboarder, via AIHR, 2026). Training itself is not free either: the average company spends 31.5 hours training one new hire, close to $1,900 in labor and materials for a business under 500 people (GMS, 2026). None of that is refunded if the person leaves in month two.
The window to keep them is shorter than it feels
Eighty-six percent of new hires decide how long they plan to stay within their first six months on the job (Enboarder, via AIHR, 2026). Twenty-two percent leave within the first ninety days, and of that group, 60% point to a lack of training or training that felt disorganized as the reason (InsightGlobal, via AIHR, 2026). That is not a hiring problem. Those are people who accepted the job and then watched the business fail to have a plan for them.
The same research puts the average new hire at six to seven months before they feel fully settled in the role (InsightGlobal, via AIHR, 2026). A small business runs exposed for that entire stretch: paying a training wage, covering their mistakes, and hoping the fit holds, with no system checking whether it will.
Documentation beats memory, every time
Devan Costa-Cargill of the Wyoming Small Business Development Center Network, which operates in part through a cooperative agreement with the SBA, put the real cause plainly: “What often gets mislabeled as employee unreliability is frequently a systems problem” (Costa-Cargill, Wyoming SBDC Network, 2026). When expectations are unclear or training gets rushed, people disengage. That is not a character flaw showing up on schedule. It is the predictable result of asking someone to learn a job with no map.
The data backs the fix. Where a formal, documented learning path exists, 82% of employees say they got one, and just over half say they knew everything they needed before they were ever asked to perform the role unsupervised (APQC, via AIHR, 2026). And the owner’s own presence compounds it: when a manager stays actively involved through onboarding instead of handing it off, new hires are 3.4 times more likely to call the experience exceptional (Gallup, via AIHR, 2026). The plan does not replace the owner. It gives the owner’s time somewhere specific to land instead of getting spent re-explaining the same three things to every new hire, one at a time, from memory.
Build the plan once, and it protects every hire after this one
This is the same argument we make for every system we build for a shop: the task is repetitive, the version in someone’s head is not reliable, and writing it down once means it never has to be reconstructed again under pressure, with a new hire standing there waiting. A documented first week, what they need to know by day one, who answers which question, what “done right” looks like for their first three tasks, costs an afternoon to write. It then runs behind every hire after this one, the same way an automated booking page runs behind every customer after the first, without the owner rebuilding it from scratch each time.
That plan does not have to live in a binder nobody opens. The same automated operations setup that handles a booking calendar or a review request can hold a new-hire checklist that fires on day one, sends the day-30 check-in without anyone remembering to schedule it, and gives a new hire somewhere to look besides the owner’s already-full inbox. Build it once, and the next new hire meets a business that had a plan for them instead of one improvising in real time.
This week
Write down the first week. Not a full manual, just the three things every new hire needs to know on day one, who owns the answer to their most common question, and what their first task should look like when it is done right. If that document does not exist yet, the next new hire is about to find out the hard way, and the data says there is a real chance they will not stay long enough to forgive it.
Sources
- The Cost of Employee Turnover for a Small Business, GMS, citing SHRM, 2026
- 27+ Employee Onboarding Statistics & Trends You Must Know in 2026, AIHR, citing Enboarder, APQC, Gallup, and InsightGlobal, 2026
- Rethinking Work in Small Businesses: Retention, Training, and the Real Cost of Turnover, Wyoming SBDC Network, funded in part by the U.S. Small Business Administration, March 2026