Insurance Team Doubles Email Productivity with AI

Every insurance contact centre has a version of the same story. Agents who care. Customers who need answers. And somewhere in between, a communication gap that neither side can fully explain.

For one leading insurance firm managing multiple emails a day, that gap had a very specific shape. It was not about knowledge. It was about words. The agents knew the answers. They just could not always find the right way to say them.

Penpal changed that. Not over a quarter. Not after a lengthy rollout. In just a few days.

The Situation

The team handled email and chat support for insurance customers, processing close to 45,000 emails every month. On paper, the operation was functional. In practice, it was stretched thin.

Agents spent close to ten minutes drafting a single email response. Not because they were slow, but because writing professionally under pressure is genuinely difficult, especially for agents whose first language is not English or who are still finding their feet in a production environment.

The consequences were predictable. Responses leaned heavily on pre-written templates, which made communication feel generic. Soft-skill errors crept in. Escalations climbed. Supervisors spent a significant portion of their time coaching on communication rather than on the work itself.

The quality scores reflected these challenges. They were not a crisis, but they were a clear signal that something fundamental needed to change.

What Changed

The firm deployed Penpal across its email and chat workflow, giving every agent access to four core capabilities: Drafting Assistance, Rephrasing, Tone Correction and Grammar Support.

The ramp-up was intentionally fast. Every target agent was live on the tool, in just few days. Importantly, this included agents who were still in on-the-job training and had not yet built the daily writing practice that more experienced colleagues rely on. For those agents in particular, Penpal effectively compressed weeks of learning into a handful of shifts.

The tool did not replace the agent. It sat alongside them. When an agent needed to convert a set of key points into a properly structured email, Penpal built the draft. When a message felt too formal or too casual for the context, Penpal recalibrated the tone. When spelling or sentence construction was off, it quietly corrected without drawing attention to the error.

The result was that agents could focus almost entirely on the accuracy of the information they were communicating, rather than the mechanics of how they communicated it.

The Numbers

The results across every key metric moved in the same direction and they moved quickly.

Metric	Before Penpal	After Penpal
QA Score	65%–70%	80%+
Email Drafting Time	~10 minutes per email	3 to 4 minutes per email
Emails per Agent per Day	25 approx	60+
Ramp-Up Time	Weeks of onboarding	Deployed in just over a week

The most striking shift was in daily throughput. Agent productivity saw rapid growth from the first month, resulting in more than double the email-processing capacity. That is more than double the output, achieved without adding headcount and without compromising quality. In fact, quality improved.

The standout achievement was a 60%+ reduction in escalations, reflecting stronger first-contact resolution and enhanced operational efficiency. Escalations in an insurance context carry real weight. They consume supervisor time, delay resolution for customers and erode the kind of trust that takes considerable time to rebuild. Fewer escalations meant fewer compounding problems downstream.

What the People Said

The numbers are useful. But the more telling signal is how the agents and their managers described the change in their own words.

“Penpal is a solution that helped me in a lot of ways. I could take a complicated email and turn it into something the customer could actually understand. If I just put in the key points, it would give me back a properly formed message I could send straight away.”
‍- Agent Email Support Team

“It saved a lot of time when drafting. I used to worry about spelling when I was in a hurry. That pressure went away. The sentences came out right the first time, and I could move on to the next customer faster.”
- Agent Email Support Team

The view from management reinforced the same picture. The Deputy Manager noted that overall volume had increased, team dependency on supervisors for communication support had reduced and the consistency of written communication had reached a perfect score internally.

“It was genuinely supportive for the team. Their speed went up. My quality numbers went up. The escalations came down. The sentence formation alone made a visible difference, because agents were no longer second-guessing themselves every time, they had to write something.”
‍- Deputy Manager

What this Means

This was not a transformation that required months of change management or a complex technical rollout. Solution was deployed in just over a week. The impact was visible almost immediately.

What Penpal addressed at its core was a gap that most contact centres carry quietly: the assumption that if an agent knows the answer, they can communicate it well. The two things are not the same. Communication is a skill in its own right and under the volume and pressure of a live production environment, it is the first thing to slip.

By giving agents a writing partner that worked in real time, inside their existing workflow, the firm effectively raised the floor for every interaction, not just the ones handled by the strongest communicators on the team.

The CSAT improvement of approximately 57% reflects something more than faster responses. It reflects customers receiving communication that felt considered, clear and delivered by a brand that knew how to speak to them.

That is the difference between a contact centre that processes queries and one that actually builds trust.

How a Leading Insurance Firm Doubled Agent Output and Rebuilt its Customer Experience in Just Over a Week

The Situation

What Changed

The Numbers

What the People Said

What this Means