The Letters Nobody Reads: AI, Plain Language, and Insurance Communications

An insurer sends a customer a letter about their policy. It's accurate, compliant, legally reviewed — and completely incomprehensible to the person receiving it. They don't understand it, so they call. The call centre explains what the letter said. Multiply that by millions of letters a year and you have a hidden, self-inflicted cost that no one owns: communications so bad they generate the very contacts they were meant to prevent. Insurance has a plain-language problem, and it's more expensive than it looks.

Why insurance writing is the way it is

It's not incompetence — it's incentives. Every customer-facing communication passes through legal and compliance, whose job is to make it defensible, not readable. Precise, hedged, jargon-heavy language is safer for the insurer even when it's useless for the customer. Add decades of template accretion — nobody rewrites the renewal letter, they just amend it — and you get communications optimised for auditability and pessimised for understanding. The customer, who needed to know one thing, gets three paragraphs of protective throat-clearing around it.

The cost is real and measurable

  • Contact deflection in reverse. A confusing letter is a call generator. Each incomprehensible communication drives avoidable contacts, and the contact centre spends its day translating documents the company itself produced.
  • Bad outcomes, not just bad UX. A customer who doesn't understand a coverage change, an excess, or an exclusion makes worse decisions — and later, a worse complaint. Misunderstanding is a claims dispute waiting to happen.
  • Regulatory pressure. Outcomes-based regimes increasingly expect communications customers can actually understand. "Legally accurate but incomprehensible" is becoming a compliance risk in its own right, not a safe harbour.

The tension AI can finally hold

The reason plain-language rewrites never scaled is that doing it by hand across thousands of templates and personalised documents was impossible, and any rewrite reopened the legal review. Language models change the economics: they can transform accurate-but-dense source content into clear, plain-language versions at scale — while a governance layer keeps the legally-required substance intact. The goal isn't to let an AI freewheel about someone's coverage; it's to hold two things at once — say it clearly and say it correctly — which was the exact tension that made this intractable manually.

Doing it without creating new risk

  1. Separate substance from expression. The facts and required disclosures are the fixed payload; plain language is how you deliver them. Structure content so the mandatory parts are explicit and preserved.
  2. Generate, then guard. Use models to produce the clear version, but gate it — verify the required elements are present and the meaning is unchanged before anything reaches a customer. Clarity with a correctness check, not clarity instead of correctness.
  3. Measure the right thing. Track calls-per-letter and complaint rates by communication. The confusing templates announce themselves in your contact data — fix the worst offenders first.
  4. Keep humans on the consequential edges. High-stakes communications get review; routine ones get automated clarity. Match the oversight to the risk.

The enabling work is, once again, foundational: structuring content so substance and expression are separable, and building the governance that lets you generate clarity safely at scale. That's the kind of data-and-content foundation we build with insurers at IntelliBooks.

Your customers aren't calling because they're confused people. They're calling because you sent them a letter engineered to be unreadable — and the call centre has been quietly subsidising that choice for years.

Comments

Post a Comment

Popular posts from this blog

Why Your Insurance Data Warehouse Didn't Fix Anything

You built the data warehouse. It took eighteen months and cost more than planned. Policy data, claims data, billing — it's all in there. The project was declared a success and everyone moved on. And yet: underwriting still can't see the whole customer. The AI initiative is still stalled. Someone still asks for a report and waits three weeks. Nobody says it out loud, but the question hangs there — what exactly did we buy? You're not alone, and it isn't because you built it badly. It's because a warehouse solves a problem you didn't have. The problem a warehouse solves Warehouses solve location . Before: your data was in eight systems. After: it's in one place. That's genuinely valuable — you can run a query across policy and claims without begging two teams for extracts. But notice what that fixes: it fixes where the data is . It doesn't touch what the data means , whether it's trustworthy, or whether the records refer to who you think t...
Read more

Straight-Through Processing: From 10% to 90%

Straight-through processing is the number every insurance executive quotes and almost nobody achieves. The industry benchmark has moved from 10–15% to 70–90% at the carriers who've cracked it. That's not a marginal efficiency gain — it's a different cost structure. So why is your STP rate still stuck in the teens? Because STP gets treated as a workflow project. It isn't. It's a data problem wearing a workflow costume. What STP actually requires Straight-through means a quote, policy, or claim goes from submission to completion with zero human touches . For that to happen, every decision point in the chain has to be answerable automatically. Every single one. A chain of nine automated steps and one manual check isn't 90% straight-through — it's 0%. One human touch and the whole thing is a manual process with extra steps. That's the bar. And it means the constraint isn't the workflow engine — it's whether the data needed at each decisio...
Read more

Embedded Insurance: Why the API Is the Easy Part

Embedded insurance is the idea that coverage should be offered at the moment of need, inside someone else's product — travel insurance at checkout, device protection when you buy the phone, cover for the rental car in the booking flow. It's one of the fastest-growing distribution models in the industry, and every insurer has a strategy slide about it. Most of those strategies underestimate the same thing, and it isn't the partnerships or the product. It's the data plumbing. Everyone assumes the API is the hard part. The API is the easy part. What embedded insurance really demands The promise of embedded is speed and invisibility: a customer buys a flight and, in the same breath, is offered exactly the right travel cover at exactly the right price, and accepts with one tap. No forms, no friction. Deliver that and you've won. But "exactly the right cover at exactly the right price in one tap" makes brutal demands on your back end: You must price i...
Read more