Episode 33: 5 trends shaping value-based care in 2026

In our latest episode, we break down the biggest findings from Reveleer's  "The State of Technology in Value-Based Care 2026," a report based on a survey of 200 senior payer and provider executives. 

During this 26-minute discussion, Reveleer President and CEO Jay Ackerman weighs in on the growing gap between fast AI adoption and slow accountability. He shares what payers and providers can do now to improve governance, data confidence, and audit readiness before pressure peaks. 

About Jay Ackerman
Jay Ackerman is chief executive officer and president of Reveleer, where he directs the company's vision, strategy, and growth across risk adjustment, quality improvement, and clinical intelligence. He has spent more than 30  years in health care technology leadership, including prior roles as chief revenue officer at Guidance Software and senior positions at ServiceSource and WNS. He holds an MBA from NYU's Stern School of Business and a bachelor's degree in economics from Connecticut College.

About Reveleer

Reveleer, a health care software and services company, uses machine learning and intelligent automation technology to empower health plans control over their quality improvement, risk adjustment, and member management programs. With one transformative solution, Reveleer allows plans to independently execute and manage provider outreach and data retrieval, coding, abstraction, member management, and reporting. Reveleer leverages proprietary technology, robust data sets, and subject matter expertise, so health plans can execute programs that deliver value and improved outcomes.

About the report

The 2026 State of Technology in Value-Based Care report is based on a national online survey of 200 U.S.-based senior decision-makers at value-based care organizations.

Reveleer collaborated with Mathematica, a policy data and analytics company, to analyze the findings. The report identifies a widening accountability gap across value-based care operations. Organizations are deploying more technology than ever before, but many still rely on manual workflows, lack confidence in their data, and lack the processes needed to validate AI-driven outputs. The findings reveal how far adoption has outpaced the governance and operational discipline required to support it.

 Click here to download the report. For additional insights, listen to the Reveleer-RISE webinar, AI Governance, Audit Readiness, and the Defensibility Standard in Value-Based Care.

Transcript

Copy Transcript Copied

Welcome And Report Overview

Ilene MacDonald, host 0:11

Hello and welcome to the latest episode of RISE Radio. I'm your host, Ilene MacDonald. Today we'll be taking a deep dive into the findings of Reveleer's recent report, The State of Technology and Value-Based Care 2026, specifically the five trends shaping how organizations operate and prepare for what's next. My guest today is Jay Ackerman, CEO and President of Reveleer, a value-based care technology platform. Welcome, Jay.

Jay Ackerman, speaker 0:40

Hi, thanks, Ilene. It's good to be here.

Ilene MacDonald, host 0:44

This is the second year that Reveleer has conducted this report. The most recent findings, from what I understand, are based on the survey of 200 senior payer and provider executives. When you look across the full report compared with last year, was there anything that stood out to you, anything that you found surprising or concerning?

Adoption Outpaces Accountability

Jay Ackerman, speaker 1:05

Yeah, well, it's great to do it for a second consecutive year. And it's interesting to see how technology is moving and what are some of the gaps that are kind of showing up or the cracks that are starting to appear . I think what's stands out this year is that adoption has outpaced accountability. And every trend in the report is a variation on that same imbalance. Every core metric jumped this year, contracts jumped, AI adoption jumped, data investment jumped, but growth has moved faster than a governance that's built to manage it. And that's the story behind the trend in this report. And I see it in some of our own work, which we'll probably hit on as we go. Contract volume hit records across payers and providers, but mean payer value-based care revenue share fell about 12 points, signaling more contracts and capturing more value are two very different problems. And most organizations have only solved the first. And lastly, provider alignment on shared value-based care goals uh dropped from 100% to 88%, even as contract volume rose, signaling some of the friction that exists between payers and providers.

Ilene MacDonald, host 2:29

You mentioned every payer that was surveyed reported growth in the value-based contracts, yet execution capacity is slipping. In your conversations that you've had with health plans and provider leaders, why do you think value-based care contracting is accelerating so fast more than other organizations can operationalize it?

Why Contracting Scales Faster

Ilene MacDonald, host 2:54 Jay Ackerman, speaker 3:02

I think there's maybe a few things that sit behind it. First, contract acquisition and operational execution are two very different muscles, and they don't scale at the same rate. Every payer reported contract growth, and 99% expect even more contract growth in 26. But we're seeing the share drop and so it shows that scaling contracts and scaling share are really two different jobs. And I think with the work that we do and the interactions that I have with payers, it's still very clear that there is significant admin pressure. And so there are, you know, there are resource constraints. Signing a contract, you can measure that as you go and seek to increase contracts with providers, you can measure that in days. But around the operational work that takes place, like that gets measured in months, in months. So customers describe the same , like signing the contract, straightforward, but building the staffing, the workflows, and the governance behind it takes considerably longer. I think that admin pressure is real. You've heard every payer talk about it in their earnings falls. They're all focused on it, they're all driving it down. They must drive it down, which is why I think the adoption trends in AI, , you know, on one hand, it's really exciting because they have to find a technology lever to release some of the admin pressure.

Infrastructure Versus Manual Work

Ilene MacDonald, host 4:41

Another finding was on the technology and workflows. And 98% of respondents say they have the right infrastructure, but 94% still rely on manual workflows. So you've emphasized the importance of defining what good looks like before building workflows. What does this disconnect really mean between infrastructure and day-to-day execution? Does it tell you anything about where organizations are getting ready for operation?

Jay Ackerman, speaker 5:14

I think it really comes down to that there's a governance and trust problem, not a technology one. 98% will say they have the right infrastructure, but 90% will also call it out as too complex to use , which is up significantly from the prior year. The technology is in place, the complexity of running it is getting worse. I think the gold rush of buying point solutions is over. And I think that's what you're seeing. You saw a rapid move to purchase solutions, and now there are technology stacks that haven't been integrated. 94% of providers and 91% of the payers are still running manual workflows, despite like that infrastructure imbalance. The gap between what's installed and how people actually do the work is showing an opening for improvement. And most are most organizations already have the pieces, but they have to do the work to connect the existing platforms. And I think the focus needs to be on that versus buying kind of net new solutions. I think the give the day of point solutions is over. And anecdotally I tell you that from my interactions with our customer set and prospects in the market, I see more national consulting firms present inside of our target customer set than I've ever seen before, which I think speaks to the admin pressure and the disconnected systems and the demand from executives to get it into alignment.

Ilene MacDonald, host 6:58

That's interesting.

Data Security And Data Quality Split

Ilene MacDonald, host 7:00

One other finding that I thought was sort of surprising to me, maybe it shouldn't be. Data security is their top investment priority, but the confidence in the data quality is low. As organizations now are trying to integrate clinical claims, provider, and third-party data across the value-based care operations, how do you think leaders should think about improving that data confidence while maintaining the security and interoperability required to use the data at scale?

Jay Ackerman, speaker 7:37

Yeah. Well, first, I think security and data quality are routinely conflated. But they are distinct disciplines funded differently and solving different problems. They have different leaders, different objectives, and different teams. And there's been a lot of attention put on data security. Uh, why? Because in 2025 we had 772 breaches exposing more than 138 million records, and that number's probably on the low end, which explains why security spend is so high. But I think with all that attention on security, it hasn't gone to gone to data compliance and data quality. And at the heart of these AI tools functioning properly is strong data governance. I mean, we see that. We have increased our own spend on data governance dramatically. And where do we are we spending that on how we bring data in from customers, how we cleanse it, how we normalize it, how we ingest it? Because if it's ingested incorrectly, it's not cleaned on the front end, then the AI outcomes are put in question. And so I think you're gonna see significant increase in data governance, data quality, and in tools and teams that can ensure strong data quality. It's at the core of AI.

AI Governance And Hallucination Risk

Ilene MacDonald, host 9:07

So you're mentioning the AI, and it looks like the report is showing 100% adoption across the board. But only 35% of respondents said they have a process to detect these incorrect AI outputs. We talked about how organizations have to catch up fast to this AI, What do you think is driving that urgency and where do you see the largest governance gaps today?

Jay Ackerman, speaker 9:36

Adoption certainly accelerated well ahead of the governance built to sustain I think it's a velocity problem, not a capability one. And I think, like, look, there was a bit of FOMO, and people felt like they had to move, they had to purchase tools. They're getting pressure from the top, signaling of something that they've seen, hear a story that they heard from a peer, and so there's this downward pressure to kind of act. Um, and it just put a lot of tools in organizations quicker than they were, you know, they could adapt and deploy it. And what I also don't see inside of these scaled organizations as much as I've seen in other companies. and like if I think about past experiences, are those groups and teams that kind of work across an enterprise to introduce large-scale change. And so these AI investments are being done within a function, they're being deployed by that function, they're being measured by that function, they're being refined, right? But there isn't an overarching kind of improvement team, quality team. And I say quality like in the biggest sense of the word, right? Like system quality to get the right outcomes. And so when you look at it, provider confidence and AI drop. They went from fully committed providers from 38% to 16. Hallucination concerns climbed. Now, hallucination is there when you when you run something through multiple times, you are going to see different answers. But if you're not going back to the source data and like tracing it back, then you can't refine it and and eliminate it. And so like the discipline around that is essential. You know, the usage is going up, but the trust is going down. So on our end, like we're piloting a new generative AI capability internally. Early results, evidence grounding failures drop roughly from 30% to five. We're measuring this every single day. We're tracing it back to the source data. And it's a small sample, but it's a real signal of like what we're trying to do. So when we interact and we run someone's data through our system, like we understand the strong outcomes and we understand where the outcomes are lining up. Look, I think it's the clearest signal that we have that adoption has outpaced governance. And so I think what we're gonna see on it is as we look in the back part of 26 into stronger governance posture from payers and providers to ensure these tools that they understand the outcomes that are being delivered from these tools.

Human Review For High Stakes Decisions

Ilene MacDonald, host 12:35

And are there any other suggestions that you may have on how these organizations should balance like this such innovation, the speed of the innovation with the controls that they you need for validation, oversight, and transparency and accountability?

Jay Ackerman, speaker 12:55

I think there's a few things that I might say on it. I mean, there's old adage, right? Like go slow to go fast. I think some of it is you've got to slow down a little bit here. And so I think like consequential questions are sequencing and what controls and warrants priority and why the high-stakes decisions merit human review before anything else does. Every organization surveyed is running AI. I mean, that's amazing, right? To think , you know, kind of three and a half years ago when all of a sudden the generative AI is kind of the talk of the town. And now every single organization is deploying it in some way, shape, or form. I've no nowhere have we seen an adoption curve that rapid. But the open question still remains whether they can document how clinical, financial, encoding decisions were reached. And so I think we're gonna see things slow h to get underneath those outcomes to make sure that we understand what's driving it and that there's traceability to it. So the sequence, the controls, that put human review the center of the highest stakes decisions first. Like when you talk high-stake decisions around clinical and financial outcomes, like you got to have a human in the loop there before automating everything else. I think the last piece I would say, particularly with clinicians, is giving clinicians visibility into how an AI recommendation was generated alongside the recommendation itself. Like the two of those going hand in hand process and output together are what earned trust. So when we talk clinicians, like it is very easy to lose trust.

Ilene MacDonald, host 14:51

Well, for sure.

CMS Audits And Traceability Readiness

Ilene MacDonald, host 14:54

I wonder if we can turn our attention to regulatory pressure now, because the report indicated that a small percentage of payers, maybe a higher percentage of providers say they feel prepared for CMS requirements. And I know RADV audits are expanding and CMS expectations are rising. Where do you see organizations most exposed? And do you have any advice on how they should prepare for this increasing need for traceability, accountability?

Jay Ackerman, speaker 15:29

Well audits in all kind of shapes and sizes are more prominent than ever. And on one hand, I think they put a significant kind of tax and burden on payers, but I think they're essential to ensuring that there's integrity in value-based care programs. So I'm hopeful that like this heavy focus towards the back part of 25 and into the first half of 27, like all this catch-up kind of audit activity will then settle down into kind of a new normal. That's just part of what we all have to do to ensure that you know the billions and billions of dollars that are flowing into payers and providers for value-based care programs are going there for the right outcomes. So like I think it's here to stay. I think we need to embrace it. I'll say that regulators adjudicate on documentation, not on intent and that's the core of the regulatory argument. So when you think about what they're doing and the small sample size they're looking at, how they extrapolate, like the data better be sound, it better be verifiable, it better be traceable. Our data shows that only 13% of payers and 29% of providers feel that they're ready for these requirements. I hear it over and over, and we focus on RADV, but then when you start to talk to people in the MA teams or the compliance teams, they'll talk about RADV and OIG and then their state level audits. So, I mean the audit activity is at an all-time high and I think it's going to continue, like we'll see it at this kind of high level before it settles down. And I think look, the advice I have you need to take accuracy and traceability to heart. And traceability is probably the one word. It's like there's a word someone takes out of this podcast it would be traceability. When you think about the systems you use, highly automated or highly people driven, d o you have traceability? Can you tell what happened every step of the way? Are you confident in the decision that someone made? And can you trace it back to sound clinical evidence? If you can't do that, you've got a problem on your hand, and you're playing like a really high-stakes game of poker that has billions of dollars at stake.

Vendor Overpromises And Performance Proof

Ilene MacDonald, host 18:12

One of the things that came up too was on vendor accountability. That 93% are saying that the tools that they got, these AI equipment, that they overpromise. Maybe they've been too reliant on the AI. And so as organizations are more dependent, you told we found it, they're all embracing it. What should leaders be asking vendors to make sure that you can follow that traceability and measure the performance expectations and just operate more efficiently?

Jay Ackerman, speaker 18:54

Yeah , I think vendor accountability is a good topic. Um, 93% of the survey would say vendors have overpromised, which is up meaningfully from the prior year, which I think is not surprising considered the kind of the high volume of like tools and applications purchased in that window. I guess I might say a couple things. I'll use an analogy, although the analogy I'm gonna use, I'm not I'm not actually a Tesla driver, but you know, you'll hear uh people talk about their Teslas. And you know, if the standard, you know, for certain models 350 miles between charges, and you go survey 10 people, you might find one who gets 350 to the charge. Right. Now, do we think Teslao overstated that? I don't think so. But that that was done in ideal conditions. So you drive your car, you gotta load it up with a full family, you got the air conditioning going, someone's got the window down, you get the trunk full of family goods as you head out on a vacation, you're not gonna get 350. That's your first drive. Are you gonna think Tesla overstated its claim? You might. And I think the same is kind of true when you look at how these AI tools function. People are reporting often with kind of ideal conditions around the set of data that they're processing. And so maybe sit within the domain that we work. All records came from one single EMR. All records came in a digital format, not in PDF. None of the PDF records were handwritten. And you start to look at those variables that impact performance. And so I think when you hear those claims, you have to understand the variables and the activity that kind of took in allowing them to make those claims. We're trying to do something on our end, which is to come out in the market and plant our flag around our performance. And we're gonna, which we have done, and then we will continue to do that so that people understand the volume of the data that goes behind the performance claims that we're making, and talking about performance at a number of different levels that our customer in the market kind of more confidence in what we're doing. And so we've recently shifted from reporting how our AI performs at an HCC level to the ICD level. And so you move from HCC to ICD, like you have to have a higher level of precision. And so that's what we're doing. We recently upgraded our AI and now reporting claims our performance at an ICD level. And we've seen a seventh going from a 74.2% success rate to 88% as compared to our prior model. And that's an example of what we're doing. I understand, I understand the concern in the market. And I think I, you know, I get why it's happening. You know, there's one example of what we're trying to do to give to better informed providers and also equip providers to ask more qualified and intelligent questions to get at how does it perform and where does it not perform well?

Ilene MacDonald, host 22:47

When you look at all the findings, if you were to your clients or a payer or provider today, d o you have like a top three recommendations on how they should prioritize their performance, governance, regulatory readiness , and what should they do to prepare for the upcoming year?

Three Priorities For 2026

Jay Ackerman, speaker 23:10

Uh sure. I think I have three things that I would highlight or encourage people listening to focus on. I'd say number one, put a AI governance process in place, including review and hallucination before a RADV finding forces you to go there. And I say that recognizing that at some level, putting that in place may actually slow down activity with a potential partner like us. Number two, I'd consolidate around data and vendor accuracy over feature count. It's easy to get excited about features. Fewer well-integrated solutions outperform a larger stack. Number three, my last one, build documentation into the workflow itself from retrieval through submissions. So records are ready while supporting detail before an auditor is asked for you to deliver that. So those would be my three recommendations. I think focus and energy put up around them would be well served.

Ilene MacDonald, host 24:25

And is there anything that I haven't asked that we haven't touched on yet, that you think is critical for leaders as they navigate value-based care in 2026?

Trust Between Payers And Providers

Jay Ackerman, speaker 24:36

No, I think the conversation's been great. I think we've covered a lot of ground. I might, if I can, just share a couple of things. I think every trend in this report points to the same shift. Technology, technology and adoption move faster than accountability. In 26 is the year that's gonna get tested. And I'm hopeful that those who listen to this podcast, like these words will resonate. when they're hearing it, they'll say, Yeah, that's what we've been feeling . Second, I'd say trust between payers and providers in each other's data intentions matters as much as the technology connecting them and ensuring that that trust is there is is super important. I think more and more we're gonna see that neither party can be successful without the other. And data flowing smooth, fast, with confidence is super important. Lastly, customers have gone from skeptical of NLP to feeling like they couldn't afford to be left behind. So, but buying the technology turned out to be the easy part. Workflows have to adapt, governance has to exist, and organizations have to define what

Download The Report And Wrap

Jay Ackerman, speaker 25:56

good, great actually look like.

Ilene MacDonald, host 25:58

Thank you, Jay. This has been really terrific. And if people have any questions, they can turn to Reveleer right? That and if there's we'll put in the show notes too, how they can download the report.

Jay Ackerman, speaker 26:11

Yeah.

Ilene MacDonald, host 26:12

and, any other kind of material that might be helpful for them.

Jay Ackerman, speaker 26:16

Yeah, we'd love anyone who hasn't read our report would love you to go download it from our website. Like I said, it's our second year publishing this report in partnership with Math Mathematica. We're already thinking about what do we see unfolding as the critical questions for us to probe on in 2027.

Ilene MacDonald, host 26:35

So we'll stay tuned for that.

Jay Ackerman, speaker 26:37

Yeah.

Ilene MacDonald, host 26:37

Thank you, Jay

Jay Ackerman, speaker 26:39

Ilene, i t's been a pleasure. Appreciate you spending the time with me.