The following article was written in association with Applied Systems.
Some agencies did not just struggle with small-group benefits. They gave up on it entirely.
Tammi Shapiro (pictured), SVP and GM of benefits at Applied Systems, described one such agency at Applied Net 2026 last week. "We have an agency that told us: we focus on property and casualty. We were trying to sell benefits - small agency, serving smaller customers. And they said it was just people working off spreadsheets, working off manual systems. So they decided it was too much work. They put benefits on the back burner."
The agency subsequently adopted Epic's benefits experience, including Autofill for data entry. The outcome was not a process improvement. It was a line of business coming back from the dead.
"What they told us is: ‘now benefits is back on the books’,” she said. “‘We are actually selling it as a line of business’."
That agency is not alone. The manual data entry burden on small-group accounts - hours of spreadsheet work on accounts where the commission never justified the effort - has quietly pushed benefits off the table for agencies whose main business is P&C. The calculation has been brutal and largely unchanged for years. Automation is beginning to move it.
Moreton and Company, a full-service agency where Ariel Peterson serves as director of training and compliance in employee benefits, did not walk away from benefits. But Peterson knows exactly why other agencies have. She has navigated the same P&C-built platforms, the same terminology mismatches, the same workflows designed for a different line of business. Benefits teams, she said at the same event, have historically been treated as "the redheaded stepchildren of the insurance industry" by agency management platforms.
The gap is not just cosmetic. In a P&C platform, everything is a policy. Benefits teams work with plans and plan lines - terminology that does not match what their staff use internally or what their carriers use externally. The mismatch creates friction that compounds across every workflow a benefits team runs.
Shapiro frames the challenge from the platform side. "The data that you need to do your work is where it varies a lot. The nomenclature that you use is different. How do we bring benefits teams onto a standard and common platform but make it resonate for them - with the right data, with the right nomenclature?"
That question has driven Applied's multi-year investment in a dedicated benefits experience inside Epic. Moreton and Company went live in March and has been using it essentially from inception.
The time savings Shapiro cites from agencies using Autofill are specific: six to 20 minutes per plan on data entry. For enrollment guides, agencies using the automated document production tool are saving approximately one hour per guide. "Before Autofill, much of that benefits data was not entering the system at all," Shapiro noted. "So the gain is both time savings and data centralization."
Across a book of small-group accounts, those figures change the margin equation. They do not change the commission rate. They change the cost of earning it. For the agencies that abandoned the line of business because spreadsheet workflows made it uneconomical, that shift is what brings them back.
Peterson's description of Moreton and Company's pre-Epic workflow is recognizable to any agency that has managed benefits on spreadsheets. A master quote spreadsheet would be compiled. Once a client made their final decisions, that data would be manually copied into a separate spreadsheet for the marketing department to build the benefits guide, then into another for the online enrollment team. Three documents, all holding the same data, all diverging the moment someone copied a figure incorrectly.
"You just have to hope that someone's not feeling a little dyslexic that day and composing numbers differently," Peterson said.
Peterson had already begun solving this problem manually before Epic - reworking the master spreadsheet template so that all the data both downstream teams needed lived in one document, enabling a single send to all internal departments after client decisions were finalized. Epic replicates that logic at the system level: form fields that template directly into Word documents, benefit guides, and enrollment workflows without re-entry.
"I just have to enter it one time and it's there," she said. "Then I can push this data out into a benefit guide, off to my online enrollment team, wherever else it needs to go."
The efficiency gain is the visible part. The less visible part is the E&O risk that copy-paste workflows carry - not because errors are inevitable, but because the conditions for them are built into the process. Moreton and Company's defense has been hiring detail-oriented staff who check their work carefully. "We are probably losing efficiency as a result," Peterson said. "We can constantly see it as a risk." The single-source model removes the conditions rather than relying on the people to compensate for them.
The compliance workload in group benefits has grown considerably without becoming more profitable for most agencies. Shapiro notes that some agencies have begun charging separate compliance fees as a distinct service line, with Applied tracking some of that billing within the platform. For agencies not yet there, compliance remains an absorb-it cost of maintaining the client relationship.
Peterson is the more instructive voice on what AI can and cannot be trusted to do in a compliance context. The concern is not theoretical - a hallucination in a compliance communication, a filing date notice distributed from an incorrect list, a campaign triggered by bad data in the system. "If you are building campaigns and you're sending out information, do you have a human in the loop to kind of review content before it goes out?" Shapiro said. "Make sure that if you're building a list based on some date or something, that you have someone step in and review it and approve it."
The same model applies across Autofill and policy checking. Agencies can configure how much automation to allow and step in to evaluate outputs before anything is pushed into Epic or distributed externally. It is not full automation but directed automation - human judgment preserved at the points where the consequences of an error are highest. For a benefits team that has watched compliance complexity grow year on year, that distinction matters.
Peterson's approach to standardization at Moreton and Company is worth examining for any benefits broker trying to build repeatable processes across clients with very different needs.
The small group team - four people serving groups of 50 and under - operates on a highly standardized, predictable process. ACA standard rates, renewals beginning two to three months before the date, consistent deliverables. The process is essentially repeatable because the variables are contained. Once past the small group threshold, level funding and self-funding introduce complexity that requires more bespoke workflows - pharmacy carve-outs, custom plan design, longer renewal cycles. For self-funded groups at 20,000 lives, renewal planning begins roughly a year in advance.
What Moreton and Company standardizes is not the full workflow but the audit and compliance layer - five required presentation components, consistent deliverable types, legal compliance checkpoints. Within that framework, account managers retain autonomy to add what Peterson calls "local spice."
"Any other brokerage can get them good rates or negotiate them a deal," she said. "Really it's the way that we make them feel and the value that we help them feel like they are getting out of our agency that really keeps them with us."
Shapiro makes the same point from the platform side. Applied's benefits investment is not designed to standardize the relationship dimension - it is designed to automate the administrative layer so that the account manager's time goes toward the relationship rather than the paperwork.
Technology standardizes the infrastructure. The relationship is still the product.
Peterson's advice for agencies evaluating a platform change is the most practically useful content in either interview, and it comes from someone with nothing to sell.
The first recommendation is to build the evaluation team from people who work the process daily. "Not just that they understand what the process should be, but they know the process," she said. She includes herself in that requirement. Peterson keeps at least one client in her own book at all times so that every process change she implements for her staff is something she experiences herself first.
"I'm in the trenches with them," she said. "If we're making a change that's annoying or it's a lot of clicks, I'm also going, oh, this doesn't make sense. I don't like this. I understand why they're complaining about it."
The second recommendation is to attend a user conference before committing. Moreton and Company sent a group of eight to twelve people - account managers, middle management, operations leaders - to Applied Net before deciding on Epic. The distinction Peterson draws between that experience and a vendor sales presentation is direct.
"In these classes, you have expert panels full of people that are using it, and they're all using it slightly different ways. They all have slightly different solutions for their agency. So you can hear these things and see these ideas."
Peer users provided honest assessments of both what works and what does not - trade-offs a salesperson has no incentive to name. Some users got on calls after the conference, shared agency data, showed behind-the-curtain implementations.
"These are real people that have real problems that they found real solutions for," Peterson said. "They can tell us the pros and the cons so we can weigh those and decide: that is a con, but I'm willing to accept that in order to get this other huge list of pros."
By the time Moreton and Company's delegation left Applied Net, every attendee had reached the same conclusion. Peterson noted that at the close of the interview. She is now, as she acknowledged, the real person on the other side of that equation - the peer user in the room that someone else's evaluation team is watching.
