The Power of
Language

Linda Baumgartner: As a child, I enjoyed writing stories and letting my imagination run wild, but I was always interested in the diversity of languages. I studied German in Edinburgh and came to Germany after graduating to improve my language skills. I first worked as a translator then as a copywriter for corporate communications. In this way, I was able to keep expanding my skills and my areas of work. 

Dennis Hölzenbein: In grade school, my friend and I had a lot of fun writing stories together so that’s how we used to spend our afternoons. Many years later, I studied online editing in Cologne and freelanced for sportschau.de. This meant I learned more and more about writing for different formats and touchpoints. At some point mobile websites and apps appeared, requiring texts that were as short as possible – and the term UX writing was born.   

Dennis: Yes, and somehow, I really like working on short texts. I enjoy writing for apps and the various screens or situations in which users find themselves. For example, texts explaining how to use the app or motivate users. Or simply just notes and designations. These all have to work extremely well, be easy-to-understand and, depending on the context, also exude a certain liveliness or informality. Users should simply be able to easily use the app for its intended purpose. When you look at texts across an entire app or website, there’s something like a big picture. And creating consistency in this big picture via the medium of language is where the real work comes in. This applies to the overall perception of the product or the brand, but also to the labelling of individual elements in detail.

Linda: For me, UX Writing is the task that can be most difficult, but it’s the one I enjoy most. Sometimes you’re asked to convey a lot of information, but in as few words as possible and in a certain style.  

Linda: … and that must be unambiguous. I’ve sometimes found a good word and then a colleague manages to find a way in which it could be misunderstood. Then it’s back to the drawing board. 

Dennis: A good indicator of how well it works is the number of variants you already discarded. This is often how things develop in the design process. What also helps is UX testing with prototypes. I would like it if the topic of language was integrated more specifically into these user tests – and we discuss this topic regularly with our colleagues from the research team. But also more general questions like “How do you understand this screen?” or “How would you navigate here?” often help you see how people interact with the texts and whether they understand things. Users immediately notice ambiguities. Bad texts are more noticeable than good texts.  

On the other hand, if the language experts are already involved during the conception phase, the prototype texts are often already at a relatively good level. And then it’s often the case that users say: “Oh, everything was really clear”, or when it comes to a longer explanation you hear “Oh now I finally understand everything.” These are nice moments of appreciation for good texts. 

Linda: Clients will often ask why we’ve chosen a certain word or expression, but as a native speaker of English, they often just trust that I know what I’m talking about. However, Germans sometimes have a certain opinion of how the English should sound. In that case, the client will sometimes decide on a less precise way of expressing something so that German users are more comfortable with the English.

Dennis: I like being asked to justify my decisions. Our clients trust us, but they still find it interesting to go through each screen and have us explain why we decided on a certain text variant for a certain screen. This helps the client to understand and can also give them a new perspective. We like when they realize that an interface with good texts feels better.  

Dennis: This strongly depends on the company, its products and target groups. Take a customer service app with several million users, for example. Here, we have to address a wide range of people, and the character traits are not so specific. Another totally different case: We developed a concept for one client that allows the linguistic tonality within an app to be changed depending on the context and to be adapted as with a control dial. We look at each usage situation in detail: If something simply needs to be done quickly, we are matter-of-fact. But if the user has reached a milestone where something has been accomplished, then we speak in a more supportive and motivating way.

Linda: In Germany, there is this clear boundary between the formal “Sie” and the informal “Du”. The employees at my fitness studio call me by my first name, but at my bank I’m addressed formally. In both cases, there’s still an atmosphere of trust, but it’s established differently. On a personal level, I've learned that staying formal in Germany is something of a protective shield, because then conflicts are easier to handle. 

Dennis: Generally speaking, I’m pleased that we work in a culture where we are often able to use the informal style of address, also when writing texts. But even when using the formal style of address, it’s possible to create lively texts that are easy to read. 

Dennis: Ideally, yes. But to achieve this you need to remain consistent across various touchpoints and media and have character.  

Linda: I think it’s a value-add when organizations have a corporate tonality, a tone of voice in which they communicate. People also like a certain familiarity and to know what to expect, so that they can use the patterns they have learned in the past. It helps them connect with the company. 

Dennis: Of course, this can then vary within the organization, according to touchpoint and target group. But there should be some kind of overarching tonality that reflects the company’s values through language. That people can connect with and somehow recognize the interaction with the brand. And this also helps the company itself, in a very practical way: If lots of people are working with language, for example briefing agencies for advertising campaigns, designing internal communications, communicating via social media in various channels, formulating customer letters, writing texts in apps – then it’s good to have a defined tonality or even a guideline. This lays the foundations and gives specific examples for different touchpoints. In this way, the many people involved can quickly write texts that fit the brand and address the needs of the customers at the respective touchpoint through language. Most guidelines should still be sufficiently flexible. But if there are none at all, then everyone who writes for the brand at any touchpoint has to think about many things first – and they still all end up with different solutions. 

Dennis: There are various ways to approach this. It depends on what the project needs, what the next steps are for the company. We also look at what’s already being used: How strongly is the brand defined, is there already a tonality that works? Should we base our work on certain corporate values? Or are there aesthetic characteristics, a certain expression, that we should translate into language? Perhaps there are already user insights into their needs at individual touchpoints. We examine all this, develop an individual approach and then a tonality. This is initially a description of the language with various attributes: What should it be like, what expression should it convey? We then work out examples of how the language expresses itself at various touchpoints. It often becomes tangible when you also formulate the “Don’ts”, i.e. what types of wording should be avoided. All this then often becomes guidelines which can also include other formal rules for language use as well as the tonality. 

Linda: Yes, but not in a word-for-word translation – more as a feeling and aesthetic expression. I prefer when I have the freedom to translate a message and a feeling instead of just a formulation.

Dennis: You should not adapt or simulate. If a brand speaks in the tonality of a target group, there must be reasons for this: for example, because it’s doing something relevant for them or shares similar views. However, it’s also possible to communicate at the appropriate level without using the same linguistic tonality.   

Linda: For me, the language of organizations can, above all, be an expression of respect, especially when it comes to contract-relevant matters with a certain legal dimension. Comprehensibility often comes up short. Poor texts or unclear preparation of information, for example in cover letters or in software, steal time and good humor from millions of people. 

Dennis: Of course, texts often have to meet legal requirements. But if you actually talk to the legal departments, you end up with versions that are twice as good. This simplification should be what we strive to achieve as copywriters.

Dennis: That’s hard to say. AI tools are currently useful for obtaining suggestions and ideas. But you still have to do a lot of the work yourself, for example a good prompt engineering, e.g. how you formulate your request. This means several rounds of entries and corrections in ChatGPT before the output goes in the right direction. Even then, the solution is usually something completely different.   

Linda: As far as translations are concerned, AI is already at an acceptable level. If certain types of expression and tonality are not important, then I can use it. But it doesn’t help you achieve that certain quality. Especially for our purpose and our clients, for whom it’s all about experience, dialog and interaction – I can’t rely on AI for that. 

Dennis: At the moment, it doesn’t seem possible for the AI to independently write texts for a very specific context, e.g.  for a screen error message where users expect a precise explanation and help in a friendly and brand-compliant tone. But when I see how ChatGPT 4 is developing, I can imagine that in the future, even complete flows and screens for apps will first be automatically texted by AI. Either way, the assessment of the quality or the final optimization of the language will remain the responsibility of humans.