Independent agents spend upwards of two hours a day on manual data entry. Not prospecting. Not advising clients. Not closing business. Rekeying information they already have into systems that cannot talk to each other, filling out carrier submission forms that share 80% of the same fields, and managing a technology stack assembled from tools that each solve one problem while creating new ones at every seam.
That figure comes from EZLynx's annual survey of its own customer base - a dataset large enough, the company says, to be a reliable read on the independent agency market. Rob Bourne (pictured), general manager of EZLynx, does not describe it as a technology problem. He describes it as a consolidation failure.
"They think about the sexier part of selling insurance," Bourne said, speaking to Insurance Business at Applied Net 2026. "It's landing the customer and being a strategic advisor to that customer. They spend a ton of time doing administrative work or answering what they might deem to be mundane questions."
The gap between what agents expect and what their working day actually looks like is not just a morale issue, it is a capacity issue - and one the insurance technology market has played a significant role in creating.
Before the productivity conversation goes any further, there is a more uncomfortable dimension that the industry is not handling well.
A 2026 Big I report Bourne cited found that 56% of independent agencies have no written policy or guidance around technology use. For more than half of the market, adoption of new tools - including tools that routinely handle policyholder data - is happening without any governance framework in place.
The practical consequences of that gap have surfaced in conversations Bourne has had directly with agency principals.
"I've run into a few agencies who have told me, 'Oh yeah, I just take my customer policy documents, dec pages, and I throw them into [a consumer AI tool], and I have them do X, Y, Z.' This is their personal account. And my recommendation always is: please stop doing that immediately,” Bourne said.
The risk is specific. Consumer-grade AI tools are not behind enterprise-grade security walls. Their terms of service typically reserve the right to use inputs to improve models and to serve outputs to other users. An agent uploading a client's declaration page or personal information into a personal account is placing policyholder PII into a system with no data processing agreement, no security audit, and no contractual protection of any kind.
"You are putting PII out there in the space," Bourne said, "as well as a lot of information about your company and your business that I just don't think is good to put in the public sphere."
His recommendation to agencies is to establish a standard operating procedure for how new tools are evaluated before adoption, mandate that any system handling client data is vetted for data privacy and security compliance, and build that framework before scaling - not after a problem forces the issue.
The governance gap is, in part, a symptom of how the agency technology market has evolved. Independent agents, particularly smaller ones, are not arriving at fragmented tech stacks by choice. They are arriving there because the market has offered them fragmentation as the only available option.
First Connect's 2026 State of the Industry Report found that 53% of independent agents say they are optimistic about technology, but only 35% are actively using it in day-to-day operations. That 18-point gap is not explained by resistance or skepticism. It is explained by the way solutions have been built and sold.
"What we've seen is a proliferation of point solutions in the marketplace," Bourne said. "This means that an independent agent, especially a smaller one, is often adopting seven tools. The excitement often exceeds their ability to adopt the tool."
Each new tool requires a login. Each requires training. Each solves one part of the agency's workflow while creating new friction at the join between that tool and everything else. An agent who adopts a standalone quoting tool, a separate policy download system, and a third platform for client communications is not building a tech stack - they are accumulating complexity.
The result is that agents default back to manual workarounds rather than adopt tools that sit outside the system where they do most of their work every day. The technology exists. The will to use it erodes the moment it demands yet another context switch.
Rating and quoting is where the fragmentation problem is most visible in day-to-day agency operations.
Personal lines have largely been addressed. Platforms, including EZLynx, have built carrier integrations that pull real-time rates into a comparative rater, collapsing what once took hours of portal-hopping into a single submission. For standard personal lines business, that problem is substantially solved.
Commercial lines is a different experience entirely.
"Historically, that has been a very painful process where you go to multiple carrier portals, and you probably repeat 80% of the information within those carrier portals for a particular carrier submission," Bourne said. "Repetitive work over and over again just to get one quote from each of those carriers for the same customer and line of business."
An agent quoting a commercial account across five carriers is entering largely the same information five times, across five separate portals, each with its own interface and its own supplemental questions. EZLynx's response has been to consolidate that into a single input - the agent enters the information once, carrier-specific questions are appended within the same flow, and quotes return from multiple carriers without the agent leaving the platform.
Carriers have a role in making this work and Bourne is straightforward about it. "It takes two to tango," he said. Technology partnerships only function when carriers actively invest in maintaining the integrations that allow data to move cleanly. The agent in the middle bears the cost when they do not.
When EZLynx surveys its customers on what they want from their management system, the answer consistently comes back the same way - and it is not a feature list.
"The number one thing that we hear, it's maybe less about features and it's more about: as an entrepreneur, you wear a lot of hats, especially as a small agency," Bourne said. "They are looking for time back more than anything."
Certificates of insurance, auto ID cards, routine servicing requests - these are the tasks that consume the hours agents did not expect to spend on administration when they went into the business. They are also exactly the kind of high-volume, rules-based work that lends itself to automation. Bourne describes agentic solutions capable of receiving a servicing request and automatically generating and sending a COI without agent intervention - the agent sets the parameters, the system executes, and the agent reviews rather than does.
The obstacle is less technical than behavioral.
"Most say they want to be the strategic advisor to their customers, but they are really comfortable in a world where they're answering mundane questions via email," Bourne said. "Are they willing to cede those things to a well-trained, secure model? That's a question."
Bourne's estimate is that around 80% of inbound servicing requests could be handled by an agentic solution with agent oversight. The agent always owns the final answer. But the volume of work that demands their direct attention rather than their approval shrinks considerably - and with it, the two hours a day that currently goes nowhere near a client conversation.
EZLynx was founded in 2003 on a single argument: that a comparative rater and an agency management system belong on the same platform. More than two decades later, Bourne's view is that the argument has not changed - the urgency of making it has.
An agency running its rater and its management system on separate platforms is re-entering the same customer data at every stage of the policy lifecycle. The prospect record created at quote does not follow the customer to bind. The policy download does not feed back into the system where the client relationship lives. Every transition in the policy lifecycle is also a data entry event.
"If those systems don't talk, you are effectively again re-entering information and managing a source of truth across two different systems," Bourne said. "There's a lot of pain in that."
The EZLynx model keeps the customer record continuous from first contact forward - the applicant profile created at quote becomes the client record at bind, and every subsequent servicing interaction is logged against it without the agent leaving the platform or re-entering data they already have.
Two hours a day on data entry is not an inevitable cost of running an independent agency. It is what happens when the technology market optimizes for selling tools rather than solving workflows - and what is recoverable when it stops.
The following article was written in association with Applied Systems.
