The Big Three pharmacy benefits managers' (PBM) share of the employer market fell from 63.4% in 2025 to 54.3% in 2026 - and 55.7% of employers still with Optum, CVS Caremark, or Express Scripts say they are considering switching within the next one to three years.
Those numbers from the 2026 Pulse of the Purchaser survey are both a warning and an opportunity. With 92% of the survey's 408 employer respondents saying healthcare costs are actively hurting their competitiveness, plan sponsors are not going to wait. The question is whether their broker will lead this conversation or be left out of it.
Shawn Gremminger, pictured, president and chief executive officer of the National Alliance of Healthcare Purchaser Coalitions, in Washington, DC, whose members represent more than 90 million Americans and spend over $850 billion on healthcare annually, told Insurance Business America what the data demands of brokers right now.
The employer exodus from the Big Three PBMs is not a future trend; it is already happening, and largely without brokers leading it.
Every independent PBM Gremminger speaks with has reported consecutive record growth years over the past four to five. The movement has been driven by smaller employers; now the data shows larger ones are watching.
"If a small PBM captures a 3,000-person company, that came from somewhere and most of the time it came from one of the Big Three," Gremminger said. "I actually think this is one where it's been smaller companies driving the change."
As independent PBMs scale up and prove they can handle large claim volumes, the hesitation larger plan sponsors have historically felt is eroding. The PBM market shift that started with small employers is now building toward mid-market and large group clients.
Some plan sponsors are already excluding the Big Three from RFP processes entirely or requiring at least one independent finalist on any shortlist.
"I think the Big Three should be scared. If our data is correct and this is actually a leading indicator, you could see that sort of 80% market penetration decline rapidly in the next couple of years," Gremminger said.
Beneath the headline numbers the survey reveals that 23.4% of employers with a Big Three PBM say they do not know what is in their contract.
Nearly one in four clients is operating blind and it’s a direct opening for any broker willing to shine a light. The structural challenge is that the Big Three continue to pay commissions and volume-based overrides that independent PBMs cannot match, and Gremminger is direct about the pressure that creates.
"You have to be abundantly clear with your clients that if you place this money with Optum, you yourself are going to make $2 million," he said. "The small PBMs cannot afford to pay those types of commissions and they're not going to."
Full disclosure is also a legal requirement. The Employee Retirement Income Security Act (ERISA) and the Consolidated Appropriations Act (CAA) of 2021 require brokers working with ERISA-covered health plans to disclose all direct and indirect compensation, including PBM-related overrides.
The Department of Labor's guidance on broker compensation disclosure under the CAA is unambiguous: failure to disclose is a prohibited transaction.
Gremminger described a pattern in which broker-controlled analysis identifies a clinical cost driver, leads to a vendor recommendation, and the broker collects a placement commission on that vendor; a conflict he witnessed directly at a National Alliance high-cost claims workshop.
"Most of the employers in the room were using the same point solution. And then our expert asked, who's your broker? Every single one of the people using that vendor had the same broker. The people who were not working with that broker had not hired that vendor. This isn't necessarily about driving the best solutions. This is about driving the most money for this particular broker."
The brokers who stand apart, Gremminger said, are those who separate the analysis from the recommendation; bringing in independent data partners such as Inovu or Springbuck, or the National Alliance's new partnership with Peterson Health Analytics, a public benefit corporation created by Peterson Philanthropies whose revenue comes from analysis, not placement.
The survey found that employers with full claim-level access deploy 11.9 high-value purchasing strategies on average, versus 7.9 for those with limited or no access.
With 87.6% of respondents rating PBM reform as helpful and 75% citing high-cost claims as a significant threat, the appetite for substantive guidance is there. The broker who delivers it earns the relationship.
Gremminger identified three baseline analyses every broker should bring to any self-funded client.
First: a high-cost claims breakdown. According to National Alliance data, 0.5% to 1% of covered lives accounts for 40% to 50% of total plan costs in any year. Identifying that cohort and what is driving their spend underpins every meaningful cost management decision, from centers of excellence to site of service changes.
Second: a provider network cost analysis. Price variation between hospitals in the same metro area can be dramatic. Gremminger cited New York Presbyterian as routinely running two to three times more expensive than competing systems in an already costly market; exactly the kind of intelligence a broker should surface before a client renews.
Third: a cross-carrier pricing model: taking the employer's actual claims history and modeling what those same claims would have cost under a competitor's network contract. Carriers push back on this, Gremminger said, but it is entirely achievable.
"Those three levels of analysis, high-cost claims, how much you're paying per provider within your network, and how much you would be paying versus other networks, those just by themselves, if done right, could save employers a ton of money," he said.
The survey found that 83% of employers say healthcare cost increases are trading off against wages, and 93% expect to shift costs to employees if nothing changes. Plan sponsors need someone to show them what is possible, an opportunity perhaps for brokers?
