Instead of hiring a professional with five years of experience, a company brings in a recent graduate, gets them up to the required level of productivity within a few months, and spends the following year getting most of the work it needs at a lower cost. As the employee gains experience, their market value naturally rises, while the value assigned to the position itself may remain unchanged. The employee asks for greater responsibility and higher pay, is turned down, and leaves; the company hires another entry-level worker at the original salary, and the cycle begins again.
This is usually treated as a retention problem, but in some companies, constant employee turnover is not a failure of the system at all-it is a deliberate way of organizing the business. If the work can be learned quickly, tenure has little effect on performance, and replacements are easy to find, retaining an experienced employee may cost more than replacing one. In that case, the business does not need someone to stay for five years; it needs a system capable of turning each new hire into a sufficiently productive employee within a matter of months.
The most extreme version of this approach can be summed up as "hire, load up, extract as much as possible, and replace": the company continually recruits new people, relies heavily on their labor for a relatively short period, and accepts their eventual departure as a normal part of the operating model. High turnover alone, however, is not enough to prove that such a strategy exists. The more telling question is whether a role is designed to accumulate human capital over time or whether its processes are deliberately structured so that the value of any individual employee grows as slowly as possible-or does not grow at all.
When Turnover Actually Pays Off
Constant replacement is cheapest in work where knowledge resides primarily in the process itself: new hires do not need a long time to understand the context, standard tasks are clearly defined, and employees rarely need to deviate from established procedures. The less performance depends on individual experience, the easier it is for a company to absorb another departure.
MIT Sloan professor Zeynep Ton and Harvard researcher Richard Hackman studied 48 months of operations at stores belonging to a large U.S. retail chain and found that higher turnover was generally associated with lower profit margins and poorer service quality, but the relationship nearly disappeared in stores with strong process discipline. Where managers consistently enforced standardized procedures, employee turnover had a much smaller effect on performance; in less formalized environments, the damage was more pronounced.
A company can therefore invest not in retaining a particular employee, but in reducing its dependence on that person. Detailed procedures, short training periods, and consistent ways of performing the work reduce the time during which a new employee costs more than their productivity justifies, while the experience accumulated by their predecessor loses some of its economic value.
Differences in turnover also arise to a significant extent at the employer level. Economists Edward Lazear and Kristin McCue, analyzing U.S. labor-flow data, found that persistent differences between employers accounted for 36% of variation in turnover; even after worker characteristics were taken into account, some employers consistently lost large numbers of people while others maintained relatively low turnover. The gap between industries was substantial as well: turnover in leisure and hospitality was more than twice as high as in manufacturing.
Short tenure, then, cannot be explained solely by generational differences, labor-market conditions, or employees' reluctance to stay in one place. Some companies create conditions that make long-term employment less likely-and that does not necessarily mean the model is economically unsound.
A field experiment at a large retail chain illustrates why reducing turnover should not become a goal in itself. Managers at some stores were asked to do everything they could to reduce voluntary departures; the quit rate fell by roughly 20-25%, but sales did not increase because managers spent more time on staffing issues and less on customers. Retention improved as a workforce metric, but the management resources used to achieve it did not generate an additional business return.
What Replacement Really Costs
The salary difference between an experienced employee and a new hire is easy to see in a department's budget. The cost of repeatedly replacing employees is spread across different functions and across several months, which means that the decision to save on retention is rarely compared with the full cost of starting another hiring cycle.
Using customer service centers as an example, McKinsey estimated the cost of replacing one employee at roughly $10,000 to $21,000, depending on location and the amount of training required. The calculation includes recruiting and initial training, but the largest component of the upper estimate comes from lost productivity: in the illustrative model, a new employee operated at only about 50-60% of full productivity during the first three active months. McKinsey also noted that annual turnover at many customer care centers reached as high as 60%, while even top-performing companies still saw rates of around 20%.
That arithmetic changes what the company is actually comparing. The choice is not between an employee earning $60,000 and a new hire earning $45,000, but between the additional cost of retaining the former and another period of reduced productivity from the latter.
Some of the cost is never recorded as turnover expense at all. A manager explains a familiar process yet again, an experienced colleague answers questions instead of doing their own work, and the team temporarily absorbs the gap between the moment a new employee joins the payroll and the point at which others can genuinely rely on them. The faster the adjustment period, the smaller these losses become; when they recur several times a year in the same role, however, a low salary becomes a much less accurate measure of what the work actually costs.
This model can persist for years when one department captures the savings while the costs are distributed elsewhere. Payroll remains low, recruiters repeatedly fill the same vacancies, and productivity is lost during each adjustment period, yet no single metric captures the cost of the system as a whole.
When Experience Starts to Become Valuable
The first months of a new hire's tenure are spent bringing that person up to a level of performance the organization already had before they arrived. Beyond that point, experience can develop in two very different ways.
In a standardized role, productivity may quickly reach a plateau: after two years, an employee performs much the same work they were already doing after eight months, so an increase in their market value is not matched by a comparable increase in value to the employer. Retaining such an employee at any cost makes little economic sense.
In other types of work, tenure gradually becomes knowledge that cannot be fully captured in a manual. Employees recognize unusual situations faster, understand the consequences of decisions more clearly, and require less supervision; over time, they also learn how to operate with particular people and within the specific system around them. Replacing someone in this environment means losing more than a pair of hands.
A study of the supply chain of a major consumer-electronics manufacturer showed how expensive that loss can become even in formally standardized production. Turnover disrupted coordination among workers on assembly lines and weakened the transfer of accumulated knowledge; the researchers estimated direct additional costs from products that failed quality control alone at $206 million to $274 million.
Replaceability therefore depends less on the complexity of the job description than on where the knowledge required to produce results actually resides. If that knowledge is embedded in technology and process, an employee's departure can be absorbed relatively easily. When part of it lives in the accumulated experience of the team, putting a new name on the org chart does not amount to a full replacement.
This is where the paradox of low-cost hiring emerges: the company pays for the most expensive stage of an employee's development, then loses that employee once the investment begins generating additional value. The first employer finances entry into the profession and the period of low independence; the next employer may receive someone who has already completed that stage.
The problem is not that a company refuses to increase every employee's salary every year. It begins when the business requires accumulated expertise while the role continues to be priced as though it were filled by someone who has not yet developed it.
A Model Designed Around Employee Departures
Some industries have deliberately operated with high turnover for decades. Professional services firms have traditionally recruited large cohorts of graduates at the base of a staffing pyramid: junior employees performed much of the labor-intensive work, a smaller share advanced to the next level, and only a few ultimately made partner. The departure of most employees did not undermine the model; it was one of the mechanisms that allowed the pyramid to maintain its shape. ()
Such a system works as long as employees stay long enough for the company to recoup its investment in their training and maintain a pipeline for the next level of talent. In 2026, the Financial Times reported that this mechanism was coming under pressure in the U.K. accounting industry: young professionals were leaving the Big Four and other firms because of slow advancement and the conditions of their early years, raising concerns about the future pipeline for management and partnership roles. A traditional pyramid can accommodate substantial attrition, but if people leave too early, fewer remain to populate its upper levels.
The distinction is critical: planned turnover does not require retaining everyone; it requires employees to remain within the system for roughly as long as the economics of that system assume they will. If a company expects three productive years after training but employees begin leaving after one, the old model stops working even if the cost of hiring has not changed.
AI is now creating a similar challenge. Professional services firms were able to hire large numbers of entry-level employees for decades because there was a substantial volume of junior analytical work; automation is reducing that work while simultaneously changing the way future managers acquire experience. PwC in the U.K. has already reduced graduate hiring, while the broader industry is reconsidering what its traditional pyramid structure should look like once some entry-level tasks have been automated.
A turnover-based business model can therefore break down in two ways: entry-level employees may leave before the company has recouped its investment in them, or the organization may lose the mechanism through which a constantly replenished junior workforce eventually produces the experienced professionals it will need several years later.
What Changes When a Company Decides to Retain Experience
The opposite model does not begin with simply paying people more. If the work remains unchanged and a more experienced employee creates no more value than before, a higher salary genuinely makes the economics of the position worse.
Zeynep Ton, who has studied operating models in retail companies, describes a different approach in which investment in employees is combined with redesigning the work itself so that accumulated experience can be used more productively. In her "good jobs" model, standardized routine work is paired with greater employee autonomy, while the broader operating system allows more expensive employees to generate more value rather than simply doing the same amount of work at a higher cost. Companies Ton initially studied through this framework included Costco, Trader Joe's, QuikTrip, and Spain's Mercadona.
Costco provides a large-scale example of this approach in an industry where high turnover is hardly unusual. In 2025, the company reported a 94% retention rate among U.S. and Canadian employees who had been with Costco for at least a year; the average tenure of a U.S. employee exceeded nine years. Costco links long employee tenure to compensation, development, and promotion from within, and most of the company's warehouse managers began in hourly positions.
Those figures do not, of course, prove that long employee tenure is responsible for Costco's financial performance: the company reports its own workforce metrics, and differences between businesses make direct comparisons between staffing models difficult. The example demonstrates something narrower but important: high-volume, low-margin retail does not have to be built around constantly replacing low-cost employees if a company can make experienced workers more valuable to the operating system itself.
Ton frames the alternative as a choice between two economic models rather than between a "good" and a "bad" employer. In one, labor remains primarily a cost to be minimized, so high turnover is accepted much like a recurring operating expense. In the other, more expensive labor must improve execution enough for the investment in people to be repaid through better business performance.
Simply raising salaries without changing the other side of the equation is not enough. A company that wants employees to stay longer has to determine what it will actually be able to do better because of that additional year of experience.
What to Measure Instead of the Turnover Rate
The same turnover rate can describe very different companies. In one, employees leave after spending several years creating more value than it cost to retain them; in another, they leave before reaching full productivity. One employer primarily loses workers who are easy to replace, while another repeatedly sends competitors people it has already trained to handle the most demanding parts of the job. An average turnover rate captures none of these differences.
The payback period on a hire is far more revealing. If an employee reaches full productivity within three months and spends the next year and a half in a role where tenure changes very little, a short employment cycle may be economically sustainable. If genuine independence emerges only toward the end of the first year and departures begin soon afterward, the company is repeatedly paying to prepare employees while receiving only a brief period of full return.
The productivity curve matters as well. It may flatten quickly or continue rising with experience; in the latter case, the cost of retention cannot be compared only with the salary of the next new hire because the company is choosing between employees with different economic value.
Turnover costs also need to be measured across the company rather than only within the HR function. If one department saves on salaries while recruiters continually search for replacements, managers repeatedly spend time on onboarding, and quality falls after every new hire, the apparent low cost of staffing may exist only within a single budget.
Four or five departures from the same position over several years do not, by themselves, show whether a company is deliberately operating this kind of model or whether that model is sustainable. A business can consciously buy short employee tenures and earn an adequate return when its processes are designed for rapid replacement. The contradiction appears when the company depends on knowledge that develops with tenure while its compensation system continues to treat employees as though that tenure has added no value. That mismatch only becomes a problem the company needs to solve if its priorities lie in long-term efficiency and developing future leaders from within.
