Why Employee Benefits Need a New Operating Model

American employers are spending more on health benefits than ever and getting less certainty in return. The KFF 2026 Employer Health Benefits Survey finds that average family premiums rose 5% in 2026 to $28,303, while workers' contributions to family coverage rose 9% to $7,471.

Workers are absorbing a larger share of that cost. Becker's Payer Issues reports that worker contributions are growing faster than overall family premiums: employees saw a 9% increase in their contributions, while average family premiums rose 5%. Healthcare Finance News notes that the premium increase outpaced both wage growth (3.7%) and general inflation (2.7%). The financial pressure is real, but it is not the only pressure employers face. They are also managing benefits across a workforce that expects instant answers, digital access, and year-round support. Most brokerage models were not built for that expectation. They were built for a world where information moved through phone calls, PDFs, and annual renewal meetings. That world is gone, but the operating model has not caught up.

The gap shows up in small frictions that compound. According to The Hartford's 2026 Future of Benefits Study, 73% of HR professionals report increased day-to-day responsibilities, and 64% find managing multiple benefits carriers challenging. An HR director needs to confirm whether a new hire is enrolled. The answer sits in a carrier portal, a spreadsheet, and an email thread that no one has updated. A CFO wants to know why the renewal came in higher than expected. The explanation requires a broker to pull data from three systems and schedule a call for next week. Each of these moments is minor on its own. Together, they consume hours that employers could spend on the work they were hired to do.

Cameron Feiner is one of three founders of Northbridge Benefits, an AI-native employee benefits brokerage built to replace that model rather than patch it. Feiner says the founding team is young, and he frames that as an advantage rather than a curiosity. They built the firm without the legacy infrastructure that makes change difficult for larger brokerages. "We're not trying to replace planks on the same ship," Feiner says, invoking the Ship of Theseus paradox. "We're trying to create a whole other model."

The problem with the traditional brokerage model is not that brokers are incompetent. It is that the structure creates friction that no amount of good intentions can eliminate. In many large brokerages, the structure involves multiple offices and layers of account management. When an employer needs a quick answer about a coverage question, or wants to add a new hire to a plan, the information has to travel through systems and people who are not always available. The employer waits. The broker catches up. The cycle repeats.

Feiner argues that technology investments at large brokerages often fail to solve this because they are bolted onto the existing structure rather than designed to replace it. "Implementing and shipping usable technology is very hard for them," he says. "They have so much infrastructure spread across their company." The result is a digital interface that looks modern but still requires the same manual workarounds underneath.

The insurance industry could have spent years digitizing forms and portals without changing the underlying model. It is a description of a pattern: the interface improves, but the dependency may have remained. The employer still relies on the broker to interpret the document, answer the question, and resolve the discrepancy. The technology has moved, but the structure has not.

Northbridge's approach is to combine human availability with agentic technology in a single system. Employers can access plan information, compare carrier options, and manage enrollment through a platform that consolidates what used to be scattered across emails and phone calls. When a question is routine, an AI assistant modeled on the firm's own expertise can answer it immediately. When a question is complex, a human advisor steps in. Feiner says the team is available around the clock, and he means it literally. "Anyone can call me at 3 a.m., and I will answer," he says.

This is not a claim that technology replaces brokers. It is a claim that technology should handle repetitive work so that brokers can focus on the decisions that require judgment. The distinction matters because it reframes what employers should expect from a benefits partner. They should not have to choose between a responsive human and a functional platform. They should have both.

Health benefits are among the largest line items in an employer's budget and one of the most visible components of employee satisfaction. When the system fails, employees feel it. When the system works, it fades into the background, which is exactly what a well-run infrastructure should do.

Feiner's message to employers is not that they need a different broker. It is that they need a different operating model, one built for the pace and expectations of the current workforce. The firms that understand this will be the ones that employers call when they need answers, not the ones they chase.