Why does underwriting get delayed? 6 places where underwriters lose time before making a decision

Too much work happens before the risk assessment

In underwriting, there will always be cases that require time, experience, document review, and conversations with the broker. Those are not things we can eliminate from the process entirely. At the same time, underwriters’ day-to-day work often shows clear signs that there is room to improve. 

To start, it helps to split underwriting into two very different areas, at least from a management perspective: expert risk assessment and the administrative and process work that comes before the decision. While it makes sense to invest time and expert attention in risk assessment, the second area can easily fall into a classic problem: over time, a series of seemingly small, harmless activities starts to create bottlenecks. 

And those bottlenecks are especially tricky because they rarely show up in reports. 

As a result, experts spend their time on repetitive, tedious tasks, and the workflow slows down. To begin changing that, let’s look at six places where underwriters most often lose time before the actual analysis and recommendation even begin. 

The case is “in progress,” but data is still missing

One of the most frustrating situations is when a case is formally active in the process but, in reality, is standing still. A document is missing, the broker has not responded, the data in the form does not match the attachment, or earlier arrangements are buried somewhere in the email thread.

A single loop like this may not look serious. But when similar situations happen every day, each one takes several minutes — sometimes much longer. After a few hours, the underwriter has done a lot of work, but is not necessarily any closer to a decision.

The first place to improve is often right here: the completeness of data and documents before a case reaches the actual underwriting analysis.

Status has to be pieced together from emails

Email is necessary for working with brokers, but it is a poor system of record for managing underwriting cases. The status is in one message, the document in another, the missing item in a third, and an important decision is in a reply from a week ago.

When a case comes back after a few days, moves from one person to another, or requires a quick response, someone first has to determine what is actually current. Are we waiting for the broker? Are the documents complete? Is the case already in analysis? Has someone made a partial decision?

If status has to be pieced together from correspondence, underwriting loses speed before the actual risk assessment even starts.

Documents are scattered across attachments, systems, and correspondence

The broker submission, application form, coverage history, technical attachments, financial data, prior terms, and additional explanations should form a single decision context. In practice, they are often scattered across emails, files, and systems.

As a result, the underwriter loses time checking which version is current, what is missing, which data does not match, and what still needs to be clarified. Instead of moving into risk assessment, the underwriter first has to prepare the working materials.

This is exactly where technology can take a significant amount of tedious work off the team’s plate: organizing documents, identifying gaps and inconsistencies, and then presenting the case to the underwriter in a form they can work with right away.

Findings and rationales stay outside the case history

An underwriter’s judgment is most valuable in risk assessment, interpreting exceptions, and making decisions that cannot be reduced to a simple form. It is a waste to spend that expertise reconstructing something that has already been established once — especially when some of the information exists only in emails, notes, or the memory of specific people.

Comments, rationales, partial decisions, similar cases, and prior arrangements should be easy to find within the case. Without that, each new analysis begins with rebuilding the context from scratch, and handing the case to someone else becomes needlessly difficult.

A well-designed workflow helps use the underwriter’s knowledge where it truly contributes to the decision: in risk assessment, not in constantly returning to the beginning of the case.

Priorities are driven by pressure, not by process rules

There will always be urgent cases, important brokers, difficult risks, and matters that require a quick response. But it is hard to maintain workflow continuity when, on a daily basis, priority is determined by the latest email, a phone call, or the person applying the most pressure.

In a well-structured process, the team should immediately see which cases are complete, which are waiting for data, which are approaching a deadline, and which require a quick response because of business value, risk type, or an agreed SLA.

When priorities are not visible in the workflow, work sequencing quickly moves outside the system. That increases tension, makes planning harder, and takes time away for manually determining what should be handled first.

SLA performance is hard to measure without manual reporting

An SLA in underwriting only makes sense when it is clear where the case actually stands: whether it is waiting for documents, sits with the broker, has already moved into analysis, requires an additional decision, or is stuck because of a missing response.

If that information has to be pulled from emails, spreadsheets, or individual notes, reporting quickly becomes yet another administrative task. The team feels that there is a lot of work, but it is harder to show exactly where delays are occurring and what is actually blocking the flow of cases.

An SLA should not mean manually tracking deadlines and then explaining delays after the fact. When it is clearly visible in the process, it helps align expectations, priorities, and accountability for the next step — not only for managers, but for the entire team.

“You can’t automate this” — where resistance to underwriting automation comes from

These six examples make one thing clear: a lot of time gets lost before the underwriter can even move into risk assessment. And this is exactly where a meaningful conversation about automation begins.

To be clear: our goal is not to automate the underwriting decision itself. In practice, that is not the right direction anyway, because every decision requires experience, accountability, portfolio knowledge, and the ability to assess exceptions that cannot be reduced to a simple mechanism.

We have heard many times about resistance to underwriting automation — and we understand where it comes from. That is why we decided to approach the issue differently.

What should workflow automation and AI cover?

In conversations with insurers, brokers, and advisors, a similar scenario often comes up: the underwriting process has already been improved, sometimes more than once, and the results still fall short of expectations. Quotes still are not going out fast enough, the team still talks about capacity constraints, brokers continue to ask for status updates, and the end customer waits longer than they should.

Often, the first instinct is to look for improvements where underwriting seems most “specialized”: in the rating manual, calculation rules, pricing approach, or AI tools that support assessment. But changing the pricing approach alone may not shorten the path from submission to quote. If the case has already become stuck because of missing information, unclear status, or manual checks of process requirements, faster calculation will not solve the whole problem.

As we have already shown, underwriting is not necessarily delayed because of complex cases and difficult decisions. It is delayed because before a decision can even be made, the case has already had time to get stuck.

Automating the pre-decision stage

This is where workflow and AI have a role to play. Not as a label added to the process because “everyone is talking about AI,” but as support in the places where underwriters should not be losing time.

In our approach, an AI agent helps prepare the case for work: it analyzes the submission, breaks it down into components, and identifies gaps and inconsistencies in the documentation provided — for example, in the broker submission, application form, or email content. It can also analyze client-specific procedures and process rules to check whether a given case should go to an underwriter at all, and if so, to whom specifically.

This matters because if a given group does not qualify for a quote because it fails to meet basic process requirements, there is no reason to involve an underwriter only for them to discover that manually. The system can recognize this type of case, indicate the reason for declining to quote, and mark the case status accordingly.

The same applies to customer verification and credibility checks. With the right integrations and business rules, the underwriter can receive the relevant context from the start, instead of searching through additional sources and assembling the information from scratch.

Workflow ties all of this together in a process: it manages statuses, ownership, deadlines, and next steps. AI helps prepare the materials. Together, they do not make the decision for the underwriter — they bring the case to the point where a person can finally focus on risk assessment, not on organizing everything around it.

What can underwriters stop doing manually?

As this example shows, a well-designed AI-enabled workflow can take over many of the time-consuming tasks that need to be completed before the actual underwriting analysis begins.

Underwriters no longer have to:

• check whether all required documentation has been submitted,
• search broker submissions for missing information,
• compare different versions of documents,
• verify that a submission meets process requirements,
• determine who the submission should be routed to,
• check basic eligibility criteria,
• gather customer data from multiple systems and sources.

Where does the underwriting process really slow down?

To close, it is worth returning briefly to the starting point: underwriting rarely slows down at the decision-making stage. The process loses momentum much earlier—when an underwriter opens a submission and cannot quickly answer three basic questions:

• What do I already know?
• What is still missing?
• What should happen next?

When those answers are buried across emails, documents, and multiple systems, it is difficult to keep the process moving efficiently. Before the underwriter can analyze the risk, they first have to piece together a complete view of the submission.

A well-designed workflow streamlines this stage. It organizes the available information, flags what is missing, and moves the submission to the point where the actual risk analysis can begin—with the full context and without another round of searching, cross-checking, and reconstructing what has already been established.