We’ve learnt a little about what doing research with bilingual users means. We’ve also tested techniques to make content bilingual and user-centred. We therefore have some shared knowledge about bilingual user research and content design.
But what about service design? What is ‘bilingual’ about the wider design of the service?
This is something that has come up in conversations about designing service patterns, or mapping services. Is there anything ‘bilingual’ about these things? What should we consider before choosing our design approaches and methodologies?
This post is exploratory. It’s aim is to get other people thinking, too. My passion for this topic may mean I’m seeing things that aren’t there. Sometimes, I may be stating the obvious. I might just be plain wrong. Please let me know what you think!
It’s more than words
Here’s a little background which shows the context that I’ve been thinking about this.
I’ve long considered bilingual design to be more than just content design in 2 languages (or even – as is often the case – content design in one language and hope for the best in the second). However, there is a wider perception – particularly amongst non-Welsh speakers – that language and words are just a vehicles for communication – somewhere between symbolic outputs and reading materials for users. They miss that they are reflections and expressions of a whole culture and way of thinking (mental models, in design-speak).
By focusing simply on the creation of words (and translations) as an output or an element of the user interface, we’re devaluing the importance of content to successful service design. This isn’t unique to any language in particular, it’s nothing new and it’s a common gripe of the content design world.
People have tried to bridge this gap between user needs, (content) design principles and business objectives by doing research with users in Welsh and English as well as adopting techniques such as pair and trio writing.
But this is merely scratching the surface of what needs to be done. We still have – despite the best will of most organisations – a reductionist interpretation of the Welsh language’s place in corporate comms and user-centred design. It’s also a view that I’m afraid Welsh speakers have somewhat accepted and bought into. This might be comfortable for many public servants but this is often not helpful for a service nor its users.
Being stuck in this bilingual content loop, I’ve struggled to articulate in a meaningful way what else we should be considering. Here’s an attempt at righting this wrong.
Learnings from user research
Based on some research done by CDPS and others, we’ve learnt that:
- bilingual users may choose to alternate which language they use throughout a service, depending on the circumstance
- users want confidence that they can complete a task in their chosen language when they start it
- bilingual users need to know that they can alternate languages if that’s what they want, without losing their progress or data.
Why users choose Welsh
We know that users choose to use Welsh for many reasons including:
- it’s natural to do so
- they feel it’s a duty to use service in Welsh
- they want to improve their Welsh
How users interact with language choice
Some of the circumstantial reasons to alternate languages when using a service can be:
- pausing to check unfamiliar Welsh terms in English, to proceed in Welsh
- needing to involve, ask or show something to another professional, colleague or family member
- a lack of confidence or assurance in the service
- a lack of confidence in their own understanding of the subject matter in their chosen language
Some research I’ve heard about also shows that some users see ‘toggling’ as a failure that the user internalises; if there’s something they don’t undertand or are unseur aboout, it’s their lack of knowledge, not a lack of clairty in the language used by the service. That’s probably another blog post…
Interpreting the research
Reading into this little research that’s been done, I’d be confident in saying that, broadly:
- user journeys won’t be different in different languages
- the service patterns (used and steps they involve) won’t be different according to language
Yet, I might be wrong. This is something that needs more testing by trying to design services bilingually with users, from the start, at scale. This means focusing on more than content.
Considerations for designing a bilingual service
What else, therefore, do we need to consider when it comes to designing bilingual services?
We have a clear set of service design principles by Lou Downe, to guide us on what makes a good service. There are other frameworks or models I could choose. I chose this as it feels like the right balance between strategic a tactical considerations that might be helpful here.
I haven’t included all of Lou Downe’s principles. I’ve omitted the ones that feel mostly relevant to service content as it’s not what I’m after.
Be easy to find
- Consider if the service is named well in English and Welsh.
- This is more than language, it’s about understanding how your users in both languages would understand and describe their task.
- This needs user and desk research in both languages.
- Be aware that verbs for common patterns in English (‘Book’, for example) don’t always have one direct translation in Welsh contexts. This means thinking at the service pattern level about differences
- Is there consistency between what something was called in campaign materials and what the service is called? What if a Welsh speaker only saw the campaign material in English on the side of a bus but would use the transactional service in Welsh – how do they know it’s the same one?
- Consider whether people in certain dialects understand some nouns or verbs differently. Test it
Set the expectations a user has of it
- Make it clear that using a service in Welsh won’t cause delays (by law, it shouldn’t)
- Let users choose their language proactively
- Explain what happens if they wish to choose their preferred language at any point
Be agnostic of organisational structures
- Don’t redirect Welsh speakers to other departments or teams (this happens). They’re just trying to complete their task regardless of your own internal structures
- Make sure all customer-facing steps can be handled directly in both languages
Require the minimum possible steps to complete
- Don’t add unnecessary steps for Welsh language users, on any channel.
- Minimise any steps for those who may want to change their preferred language.
Be consistent throughout
- If users alternate between Welsh and English interface, the service should remain familiar visually and logically
- Make sure that designers and translators have worked closely to develop and apply clear language and consistent terminology.
Have no dead ends
- Make sure it works end-to-end in each language, across channels, regardless of their language choice.
- A Welsh-language user should not unexpectedly land on an English page or end up speaking English to someone on the phone.
Be usable by everyone, equally
- You must make it equally accessible and usable in all languages.
- Support Welsh language users wherever they are on their journey with Welsh by using clear, everyday language
- Design interfaces and interactions for both languages to avoid visual issues when displaying content
- Support people who need to check what something means in the other language
Respond to change quickly:
- Have a record of a user’s language preference
- Make it possible for the user to update their language preference and respond accordingly
Work in a way that is familiar
- Bilingual (Welsh and English) speakers will be accustomed to alternating between languages, services should accommodate this
- Build common design patterns and systems that document any variations or considerations for Welsh language versions
Encourage the right behaviours from users and staff:
- Design services front-to-back, and identify what it takes to run a live bilingual service
- Have clear internal capacity, processes and expectations for running a bilingual services (such as, do you have staff to deal with queries and responses, how can technology help?)
- Make it easy to get human assistance: if a Welsh speaker needs to speak to a human, make it possible for them to continue in Welsh
Many of these Welsh language functions are already considerations covered by the Welsh Language Standards. However, they are often treated as compliance matter. From my experience, this means that the focus is ultimately on the making sure the organisation doesn’t fall foul of the rules; anything else is a bonus.
Service design goes a step further – it makes sure that organisations provide services in a way that meets the users’ needs and expectations as well as meeting business and compliance goals.
Organisations often fail when they offer a service in Welsh, such as offering online information or bilingual letters, but users find that there are no advisors who speak Welsh, or find a team has little capacity to deal with requests. These can lead to bottlenecks and service failures.
Your ideas
I hope this is something to build on. Your feedback and input would be welcome.