Public Storage AI Staffing Cuts Property Hours Over 30%
Summary
Public Storage says its AI staffing model cut property hours by over 30% and helped lower payroll, though incentives and other initiatives blur its impact.
Public Storage (PSA) said on its July 30, 2026 second-quarter earnings call that its AI staffing model generated payroll savings as quarterly payroll fell about 1.8%. Incentive spending offset part of the decline, however, and the company did not isolate the model's contribution.[1]
Public Storage operates self-storage properties, primarily renting storage space to individuals and businesses and collecting rent and related fees. Its AI application is a machine-learning staffing system for property operations teams. The system analyzes traffic, customer-service needs, asset characteristics, risk and seasonality at each property to determine how many on-site hours are needed. It directly affects daily property operations and whether staff are present when customers need help, making it a core operating capability.
From operating efficiency to cost evidence
The application has gradually connected operating metrics with financial evidence. On February 25, 2025, Public Storage described AI staffing as part of a digital operating model and said on-property labor hours had fallen nearly 30%, without separating AI's contribution from other platform changes.[3] By October 30, 2025, the company was still pursuing staffing aligned with customer needs, and the cumulative reduction in labor hours had expanded to more than 30%. Management also cited higher employee engagement and lower turnover, but did not quantify either result.[2] On July 30, 2026, management explained that the model studies specific operating factors at each property and directly connected it with current-period payroll savings for the first time, extending the evidence from operating efficiency to costs.[1]
What the 30% reduction measures
The cumulative decline in hours shows a material change in how Public Storage staffs its properties. Management said the program had been underway for three or four years and that hours had fallen more than 30% from its start through July 2026, while the work was not yet finished.[1] The company did not disclose a consistent baseline or the population of hours covered. The figure therefore cannot be directly compared with a single quarter's payroll decline: it measures a long-term change in on-site hours, not an equivalent percentage decline in payroll or property operating expenses over the same period.
Payroll is the visible financial link
The clearest financial endpoint is property payroll and related operating expense. Public Storage said payroll savings from machine-learning staffing offset some pressure from property taxes and marketing, while payroll fell about 1.8% in the quarter and 1.2% year to date.[1] Reducing unnecessary on-site coverage would generally lower payroll expense and support operating margins, but the company did not specify the comparison basis for either payroll decline or disclose the amount or percentage-point contribution from the model alone. New incentives also offset part of the decline that otherwise could have been achieved, so the full payroll change cannot be attributed to machine learning.
Public Storage has moved machine-learning staffing from an efficiency tool into property cost management. The confirmed changes are sustained reductions in on-site hours and payroll savings moving in the same direction. The model's independent contribution to property operating expense and margins remains unquantified; the next useful disclosure would separate comparable payroll savings from the offsetting cost of incentives.
Application assessment
- AI Property Staffing Optimization | Business position: Core operations | Application stage: Limited production | Scope: Company-wide | Value type: Cost reduction
Sources
[1] Drillr · Public Storage (PSA) · July 30, 2026 · Earnings call
Original: In terms of this year, you can see we're down about 1.8% in the quarter, 1.2% for the year.
Chinese translation: 就今年而言,可以看到本季度下降约1.8%,年内下降1.2%。
[2] Drillr · Public Storage (PSA) · October 30, 2025 · Earnings call
Original: To date, this has reduced labor hours by more than 30%, while also increasing employee engagement and lowering turnover.
Chinese translation: 截至当时,这项转变已使工时减少超过30%,同时提高员工参与度并降低流失率。
[3] Drillr · Public Storage (PSA) · February 25, 2025 · Earnings call
Original: As a result, we've reduced on property labor hours by nearly 30% and there's more to go.
Chinese translation: 由此,我们已将物业现场工时减少近30%,而且仍有进一步下降空间。