Building a Connected Claims Ecosystem: Why Integration Wins 

By James Benham, Co-Founder, CEO, JBK, and author of Be Your Own VC 

Picture a carrier that spent seven figures on a “modern, end-to-end” platform two years ago. It looked great in the demo. It still looks great in the demo. But the adjusters are toggling between four browser tabs, copy-pasting medical updates from a provider portal into the claim system, re-keying notes from a TPA dashboard, and downloading PDFs from an employer’s HR tool to attach back into the file. The “single system” promise lasted exactly as long as the implementation kickoff slide. 

Modernization in workers’ compensation has a truth nobody wants to flag in the demos. The platform on the org chart is not the system the work is actually happening in. Workers’ comp does not run on a single screen. It runs on handoffs between carriers, TPAs, medical providers, employers, vendors, regulators, and case managers. When those handoffs fail, the platform’s branding is irrelevant. The friction shows up in the adjuster’s queue and on the injured worker’s voicemail. 

Integration is the part of modernization that does not get the press release. It is also the part that decides whether your modernization actually changes anything, especially in our era, where integration and data access arguably matter more than the surface layer that is most visible. 

Internally, I’ve started to use the phrase “friction tax” to describe the cost every disconnected handoff in a claim pays. Every time an adjuster has to re-key a wage statement, hunt for the latest IME report, or chase a status update across two systems, the claim has paid a small toll. None of those individual tolls feel catastrophic. Stack them across a 250-claim caseload over twelve months and the friction tax is enormous. Most of it is invisible because nobody books a line item called “adjuster time wasted on copy-paste.”  

But the cost of manual do-overs is tremendous, and two costs in particular sit downstream. 

The first is duration. Studies of workers’ compensation systems, including longstanding work from the Workers’ Compensation Research Institute and a Monash University analysis of claim-processing times, consistently show that the longer the administrative phase of a claim takes, the longer the disability lasts and the worse the return-to-work outcome. The correlation holds even after adjusting for injury severity. The administrative tempo of a claim is not separate from the medical outcome; it shapes it. 

The second is litigation. Attorney involvement in workers’ comp is rarely a function of legal complexity alone. WCRI’s research on why injured workers retain attorneys points to perceived communication failures: workers feel ignored, confused about their care path, or unsure whether their claim has been accepted. When systems do not connect, the people running those systems cannot connect with the worker on the other end of the file either. They are too busy reconciling data. A WCRI study of more than 950,000 lost-time claims found that represented claims cost roughly $7,700 to $12,400 more than comparable non-represented claims. The failures here are paid for, in part, by every workflow that should have been automated and was not. 

Stan Smith, who founded Gradient AI and has spent the better part of a decade training models on workers’ comp claims data, made a point in a recent discussion that I keep coming back to: if an adjuster has to go hunting for information, they will not look for it. They will deal with what is in front of them and move on. It is not laziness. It is bandwidth. And every fractured system in your stack slowly raises the price of attention on every claim moving through it. 

Disconnected workflows are not an IT problem, but they are a claim outcome problem. 

When I talk to claims leaders about integration, the conversation usually collapses into a vendor bake-off. Three logos on a slide, a feature matrix, a final pick. The framing is the problem. Integration is not a feature you compare on a vendor matrix. It is the underlying strategy that makes any of those vendors worth buying. 

Over the years, I’ve outlined how bootstrapped operators have to scrutinize every dollar because nobody is going to bail them out. The same discipline applies here. A new tool that lives in its own silo, behind its own login, requiring its own export-import dance, does not save the adjuster time. It moves their time around. A bootstrapped claims leader insists on a different question before signing anything: where exactly will this tool live in the daily flow of work, and what does the adjuster have to stop doing to use it? 

It’s also why I am skeptical of “platform consolidation” pitches that promise to replace your ecosystem rather than connect it. Workers’ comp is too jurisdiction-specific, too regulated, and too dependent on niche specialty providers for any single platform to credibly own the entire chain. Stan Smith makes a related observation about AI tools. The horizontal generalists can do remarkable things at the surface, but they rarely understand a domain deeply enough to be useful for high-stakes, regulated work. Workers’ comp is exactly the kind of domain where the generalists fall apart on first contact. Real integration accepts as much. It treats the ecosystem as the unit of strategy, not any single application within it. 

When integration is real, three things change. First Notice of Loss data flows cleanly into the claim system without an adjuster acting as a transcriptionist. Medical records, IME reports, and bill review summaries land where the work is happening rather than in a separate inbox. Return-to-work plans are visible to the employer, the provider, and the case manager in a form they can act on, instead of bouncing back and forth as PDF attachments. None of that is glamorous. All of it is what makes the rest of your modernization actually deliver. 

Most procurement processes ask the wrong questions of vendors. They focus on features, screens, and pricing. They rarely interrogate how a tool is going to coexist with the rest of your environment. If you want integration to be real rather than aspirational, the questions you ask before signing matter more than the demo. 

Here are five I would put on every short list. 

Does it integrate where the work actually happens? If the tool requires a separate login and a separate dashboard, you have not bought integration. You have bought a portal. Worse, if the “integration” is a hyperlink that punches the adjuster out of their claim system into someone else’s UI, you have bought a redirect. Real integration shows up inside the cockpit the adjuster already lives in: same screen, same claim, same workflow. Anything else is asking your team to context-switch on top of an already heavy day. 

Does it eliminate duplicate entry, or just relocate it? This is the test most “integrated” tools discretely fail. If your adjusters still have to re-key information from the new tool back into your system of record, or vice versa, the integration is cosmetic. Ask the vendor to walk you through the exact data path from their tool into yours, and what happens when a field does not match. The honest vendors will tell you. The rest will get vague. 

Is it built for the complexity of workers’ comp specifically? Generic horizontal tools, even excellent ones, tend to break against jurisdictional nuance. A platform that works beautifully for commercial auto in California can fall apart the first time it meets a Texas non-subscriber file or a Florida workers’ comp regulation. Ask the partner to describe a jurisdictional edge case they have already solved. If they cannot, you are about to become their first customer in it. 

Does it speed up time to first meaningful contact? This is the single most important operational metric in claim modernization, because it is the one most directly tied to the litigation and duration outcomes described above. Every integration should be evaluated against the question: does this give the adjuster bandwidth to call the injured worker sooner? If the answer is no, or “indirectly,” push harder on what the tool is actually delivering. 

Are we partnering, or just purchasing? This is the question that separates the vendors who will be helpful in eighteen months from the ones who will be a problem. Implementation in workers’ comp is not glamorous. It is workflow mapping, jurisdictional configuration, training, and feedback loops. The right partner shows up for that work. The wrong one hands over an API key and an account manager and disappears until renewal. You can usually tell which kind you are dealing with by how seriously they take the unsexy parts of the conversation. 

The Beneficiary at the End of the Chain 

It is easy, in any conversation about integration, to lose the thread of why this matters. Integration is not a technical achievement. It is not a procurement win. It is not a slide in a board deck. 

Integration is what removes the distance between an injured worker and the help they need. 

When the carrier, the TPA, the medical provider, and the employer can see the same claim at the same time, fewer things fall through the cracks. The adjuster has hours back to make the call that explains what happens next. The nurse case manager has the medical context to coordinate care. The employer can hold a job open with confidence because the recovery timeline is visible to them. The injured worker, the person whose Tuesday morning got disrupted by a torn rotator cuff or a back injury, gets a clearer, faster, less frightening experience of a system built for exactly them. 

The workers’ comp ecosystem is not going to consolidate into a single platform anytime soon, and frankly, it should not. Specialization in this industry is a feature, not a bug. What does have to change is how those specialized parts talk to each other. The carriers, TPAs, and risk managers who win the next decade will not be the ones who bought the most tools. They will be the ones who connected the right ones, in the right places, around the work that actually matters. 

Pick partners who treat your ecosystem as theirs to integrate into, not yours to be replaced. Pay attention to where the data hands off. And keep asking, on every tool, every vendor, every renewal: does this give my adjuster more time to be human with the worker on the other end of this claim? 

It’s the only integration question worth asking. 

James Benham 

James is the Co-Founder and CEO of JBK, a global technology firm powering major carriers, and the Co-Founder of Terra, a cloud-native platform reimagining Workers’ Compensation. He also hosts The InsurTech Geek podcast and is a frequent speaker on insurance innovation. 

Share This Article: