Why Financial Companies Should Treat Language as Part of the Customer Experience

Financial companies spend heavily on digital platforms, product development and brand campaigns, yet clients often form their strongest impressions from something much less sophisticated: a letter they cannot understand, a claims decision written in legal language or an employee who responds to an anxious customer by asking immediately for a policy number. As financial products become easier to compare and digital services remove many practical differences between providers, the language companies use with existing clients increasingly influences whether a relationship feels straightforward or unnecessarily difficult.

Zurich Insurance has spent several years working on precisely that problem. In an interview with Finews, Group Chief Customer Officer Conny Kalcher described how the insurer concluded that its communication had often remained technically accurate while sounding too formal and distant. Zurich responded with a company-wide Tone of Voice programme, empathy training for customer-facing employees and a systematic rewrite of customer documents. At the time of the interview, the insurer expected to have simplified 60 percent of its written communications by the end of 2025.

Zurich did not treat the exercise as a copywriting project. It trained employees through scenarios with actors, changed how customer-service staff opened conversations and linked the broader customer programme to commercial measures such as retention. Kalcher also reported that Zurich’s brand value had increased by 35 percent after the insurer changed its visual and verbal presentation, while its Instagram audience had grown 185 percent since 2021 and its TikTok following eightfold since 2023. Those numbers cover a much wider transformation than language alone, but they illustrate why large financial companies increasingly connect communication style with customer behaviour rather than leaving tone of voice to the brand department.

Customers Read Financial Language When Something Has Already Happened

Banks and insurers face a particular communications problem because clients often pay greatest attention to their language at moments when money, uncertainty or risk has already raised the stakes. A mortgage customer reads conditions while committing to years of payments, an investor examines a portfolio report after markets fall and an insurance customer studies a claims decision after an accident, illness or loss. The company may regard the document as routine correspondence, while the recipient reads every sentence looking for an answer to a personal question.

Technical accuracy therefore solves only part of the communications problem. Legal, compliance and product teams need precise language, but clients also need to understand what happened, what the company has decided and what they can do next. When those objectives become separated, organisations can produce documents that satisfy internal review while forcing customers to decipher terminology that employees use every day but clients encounter perhaps once in a decade.

Recent Swiss insurance figures add another dimension. The Ombudsman of Private Insurance and of Suva recorded 4,771 queries and complaints within its jurisdiction in 2025, up 55 percent from 3,078 a year earlier. Daily sickness allowance cases almost doubled from 426 to 850, while legal-protection cases rose 76 percent to 660. The Ombudsman attributed part of the overall increase to policyholders using artificial-intelligence tools to interpret correspondence and formulate complaints, noting that AI sometimes produced incorrect legal conclusions but also helped customers overcome language barriers and express their concerns more clearly.

Companies now communicate with customers who can paste a rejection letter into an AI assistant within seconds and ask what it means, whether the reasoning looks correct and how they should respond. Insurers cannot control the answer another system produces, but they can reduce unnecessary ambiguity in the original communication. A clear explanation of the decision, the contractual basis, the next steps and the available appeal route gives both the customer and any digital tool a better starting point.

Plain Language Does Not Require Companies to Remove Precision

Financial institutions sometimes resist simpler language because regulated products genuinely contain complexity. A derivatives disclosure cannot become a lifestyle article, while an insurer cannot remove qualifications from a claims decision simply because clients prefer shorter correspondence. The more useful objective separates complexity that belongs to the subject from complexity that the organisation has added through habit.

A bank can retain a legally required term while explaining it immediately afterwards. An insurer can state the contractual reason for a decision before reproducing the detailed clause. An investment firm can show the number a client needs first and place the methodology below it. Employees can use precise terminology when the terminology adds information, while replacing internal expressions that merely demonstrate familiarity with the company’s own processes.

Zurich’s programme provides a practical example because the insurer combined document simplification with employee training rather than expecting a new style guide to change thousands of interactions by itself. Kalcher told Finews that staff learned to begin service conversations by asking how they could help instead of moving immediately to the policy number. The difference appears small on paper, yet it changes the sequence of the interaction: the employee first acknowledges the customer’s purpose and then moves into the administrative process required to solve it.

Companies that want a different tone therefore need to change behaviour as well as vocabulary. A brand can tell employees to sound empathetic, concise and human, but those adjectives offer limited guidance when somebody has to explain a rejected claim, a delayed payment or a falling portfolio. Scenario training, approved examples and rewrites of real customer documents give employees a clearer model of what the organisation expects.

AI Raises the Value of a Recognisable Human Voice

Generative AI will make financial communication easier to produce, translate and personalise, which should reduce the cost of rewriting large volumes of customer material. It also creates a new risk: companies can generate more competent-looking language without deciding how they actually want to speak.

That problem becomes visible when every email follows the same polished structure, customer-service answers rely on interchangeable phrases and executives publish commentary that could belong to almost any institution. Financial companies have traditionally protected consistency through templates, legal clauses and approval processes, but AI can extend standardisation so far that the company loses the verbal characteristics that make communication recognisable.

Marketing and communications teams therefore need to define tone through choices rather than adjectives. They should decide whether executives normally write in the first person, how directly employees explain bad news, which technical terms the company retains, how much context customers receive before an answer and when a digital interaction should move to a person. Those decisions give writers and AI systems practical boundaries that terms such as “professional”, “trusted” and “client-centric” cannot provide.

The same work also requires cooperation across departments. Marketing may control campaigns, legal teams approve regulated statements, operations produce customer letters and relationship managers handle conversations that never appear in public channels. Clients, however, experience one organisation. A company that sounds informal and helpful on Instagram but bureaucratic in its customer correspondence exposes the internal boundaries that its branding tries to hide.

Customer Communication Can Be Measured Against Revenue

Zurich also offers an interesting approach to measurement because it moved beyond counting customer acquisition. Kalcher told Finews that the insurer had gained nine million retail customers during its previous three-year cycle, but argued that acquisition alone did not explain whether the company was strengthening relationships. Zurich introduced Net Revenue Retention to measure how much revenue it retained or expanded by serving existing customers more effectively.

The company’s 2025 reporting develops that approach further. Zurich served more than 80 million retail customers worldwide and more than 11 million customers used its digital platforms. During the year, the insurer surveyed 1.6 million customers and reported a 3.6-point improvement in its global transactional Net Promoter Score. It also incorporated Net Revenue Retention into local strategy planning across 23 markets.

Communications teams cannot claim that clearer letters alone produced those outcomes, and attempts to assign revenue mechanically to individual wording changes would create false precision. They can still measure whether customers understand documents more quickly, whether service centres receive fewer clarification calls, whether complaint patterns change, whether digital tasks require fewer follow-ups and whether relationship teams see recurring confusion around particular products or processes.

Those measures move tone of voice away from aesthetic preference. A sentence that reduces misunderstanding can save an employee a phone call, prevent a complaint or help a client complete a transaction, while thousands of such interactions can influence service costs and customer retention.

Financial Brands Need to Work on the Communications Clients Cannot Avoid

Brand teams naturally devote attention to the visible parts of a company: advertising, websites, social media, campaigns and major executive appearances. Clients can choose whether to engage with most of those channels. They cannot as easily ignore a premium notice, portfolio statement, mortgage letter, claims decision or message asking them to provide documentation.

Those compulsory interactions deserve the same editorial discipline as public marketing because they often reach customers when the relationship faces a genuine test. Zurich’s work on customer language suggests a broader principle for financial companies: modernising a brand requires more than changing what prospective clients see before they buy. Companies also have to examine the words customers encounter after the relationship has already begun.

As AI gives customers more ways to interrogate every document and compare every answer, financial institutions will find it increasingly difficult to hide unnecessarily complicated communication behind technical authority. The companies that write clearly still need lawyers, compliance controls and detailed contracts, but they make the institution easier to understand without asking clients to become experts in its internal language. For banks and insurers that depend on long relationships, that ability belongs much closer to the commercial product than conventional definitions of corporate communications have allowed.