ChatGPT Prompts Engineering to Improve the UX of your Fintech App

AI has caused a radical and massive shift in the tech ecosystem, and UI/UX design is no exception. It has improved workflow efficiency and automation but has produced more superficial content with no depth.But what does all this truly mean for the design industry? To what extent will this insurgence of AI affect businesses and their offerings? Will Artificial Intelligence just be a complementing asset, or will it prove to be capable of mimicking human behaviour?

Introduction to Large Language Models (LLMs) and ChatGPT

Traditionally, there is software that takes input from humans, processes it and produces outputs that we can understand. These software follow certain rigid rules which are limiting. So, language models do what this software cant do; however, they need more data. A Language Model tries to understand and predict how words are used in human language.We want to understand the technology behind ChatGPT, Gemini, Claude, etc. If a language model tries to understand human language, imagine what a large language model will do. Lets compare pots and plates. A language model is a plate, while a large language model is a pot. They both serve as food containers but have different use cases.

Characteristics of LLM

• They train with a large data set of text, images, codes, etc., which helps the language understand complex and difficult human language.
• Because of its large capacity, it can perform related tasks it was not trained for
• It uses a deep learning system.

ChatGPT was built with an LLM by OpenAI to engage in conversational interactions.

Fintech and AI

The primary objective of fintech UI/UX app design is to ease the struggles associated with traditional banking and other financial institutions. Examples of these issues include:

1. Generic services
2. Lack of personal financial services
3. Slow processing of transactions
4. Low security etc

AI greatly benefits current UX design by automating repetitive operations, personalising experiences, generating insights, and enabling predictive design. AI can also automate common operations such as analyzing data, A/B testing, and structure changes.All of these features are what Fintech products need the most!AI tools deliver useful insights by identifying patterns and developments in user data. This allows them to better understand user demands and increase happiness and engagement.Artificial intelligence has brought a new design method more efficient and faster. Analysing large volumes of user data: UX designers no longer need to manually sort through voluminous user research to comprehend how customers behave today.

Metrics to consider when using ChatGPT For Fintech App

When designing, how do you measure your effort? If it aligns with brand goals or business requirements. Now that AI is in the picture, you need metrics that will assess the effectiveness of AI tools.This will help you understand how to train the AI tools to achieve the desired result.

Task completion rate

This measures the percentage of users who complete a specific task within the app (e.g., transferring funds, opening an account, making a payment). When using ChatGPT-powered features, track if these rates improve or decline. For example, if ChatGPT guides users through a complex form, a higher completion rate would indicate success.

Error rate

How frequently do your users encounter errors when using your product? If you have a high error rate, you should implement AI in your design system. LLM can also help with human oversight and provide clearer instructions or help with filling out forms.User satisfaction: This is another area you should prioritise when introducing ChatGPT or any AI tools into your design or user workflow. It measures users' overall satisfaction with the app and its features.This can be measured through surveys, ratings, and feedback. You can use chatGPT to analyse the UX survey result and even tell it to give strategies to see how it will improve user satisfaction.Support tickets: A high volume of support tickets means users need external help, maybe because the UX is not intuitive or does not solve their problem. So, you need to find a way to handle the support queries and analyse them.If you have not tried using AI chatbots, you are missing out on a deal. They can serve as first aid for users' issues and give you a breakdown of the support tickets based on the algorithm. This will help you know which area needs more focus, efficiency, and innovation.Once you have implemented the AI Chatbots, compare the difference with when your support ticket was manual. Use A/B testing to compare the performance of users who interact with ChatGPT features versus those who don't

What you can use instead of ChatGPT

Gemini: It is an LLM created by Google to handle different kinds of data, such as text, videos, images, and even codes. This can be very helpful if you want to analyse data, especially for your fintech app. However, there is no specific emphasis on fintech contexts. Gemini also integrates with Google Cloud services.Claude AI: Claude was created by anthropic, and its strength lies in it ability to maintain context in extended conversations. Itll honestly be the best for customer support. It is also designed with a more factual framework, so you can easily use it to generate complex analyses without the fear of generating incorrect or misleading financial advice.DALL-E: Google also developed this, focusing on generating images. It is limited in use but can still be useful. For example, you may need images for data visualisation or in-app pictures; you only need to describe what you want, and boom, your image!NotebookLM: Then, there is this, primarily for research. It works well with documents or preexisting research and synthesises information from multiple documents. However, as it was designed for research and analysis, it might fail as a conversational tool. If you have your user or market research, it can help you break it down into actionable steps to improve your product's UX.

ModelStrengthsWeaknessesGeminiMultimodal, Google Cloud integration, strong reasoning potentialRelatively new, limited public informationClaude AISafety focus, strong long-form conversation, less prone to hallucinationsLimited multimodal capabilities, fewer integrationsDALL EVisualising data, branding/marketing (limited)Not designed for conversation or financial reasoning, limited direct applicability to core UXNotebookLMWorking with documents, potential for regulatory complianceNot designed for user interaction, limited direct application to end-user UX

Why ChatGPT is relevant for Fintech UX?

Financial products and services are typically complex, involving technical jargon and governed by strict regulations. Staying ahead of these problems and guidelines can be time-consuming and daunting.Thats the price designers pay to provide the perfect UX for users. However, the introduction of AI into the ecosystem is really helpful, and you dont want to sleep on it as a UX Professional.Lets see how ChatGPT can help improve the quality of Fintech services

Ideation and Brainstorming

The UX design process might not be linear, but ideation comes with the package. ChatGPT can be your PA for ideation. Some Fintech design briefs just throw you off balance and would take a lot of back and forths to establish a concrete solution. It could even be that the solution is ubiquitous, and you need something unique.All you need to do is write your prompt like you are talking to a child, write out your expectations, and let ChatGPT do its magic! Sometimes, you don't need any prompt engineering; all you need is a simple idea, and you can develop it yourself. Whichever is it that suits you, you are sure to find it in ChatGPT.

Faster User Research

Amongst other tools, ChatGPT can significantly accelerate user research by automating tasks such as generating interview questions, analysing user feedback, and creating user personas. Instead of starting from scratch and wasting much time, you just need to refine the output from chatGPT and go ahead with your users.It could be pasting a collection of user feedback into ChatGPT and prompting it to "Summarize the key themes and sentiments in this user feedback and to provide a list of features that addresses the sentiments" can quickly provide valuable insights.ChatGPT will help with faster iteration cycles, quicker validation of design decisions, and a deeper understanding of user needs without extensive time and resource investment.

UX Content Enhacement

Once you open a new page for a project, you can use it until the end, and it keeps the context. Thus, you save yourself time for excessive explanation.You can just tell chatGPT to write user-friendly or edit your previous microcopy, error messages, tooltips, and any other copy that addresses issues in the user research or the solution youre going with.ChatGPT provides consistency in brand voice and tone in all user interactions. Human accuracy and speed might not be enough to help users and address their pain points.

It is Useful for Designing Personalised Experience

ChatGPT can be integrated into your design solution to create more personalised user experiences. You can input the user interactions and activities on the app, and then it gives personal recommendations to the user.For instance, a budgeting app could provide tips and advice based on a user's spending habits. It will excite the user to keep engaging with the app and even explore other services to see how they can benefit their financial health.

It Innovates solutions for your design.

There is a difference between innovative ideas and innovative solutions. ChatGPT can do both! An idea will help you get started on designing, while a solution will show you how to solve your user's needs with the idea.It could be that you already have an idea and need to infuse it with users' needs. Say no more! You just need to prompt ChatGPT to give you a unique perspective on differentiating your Fintech product.

UX Problems in Fintech That ChatGPT Can Assist With

Fintechs have industry-specific problems that users struggle with. There is a gap between technology and usability, which manifests as UX and usability issues such as confusing interfaces, complicated processes, and a lack of personalised experience.

Complex Financial Terms

We see a lot of technical terms flying around, and most users barely know what they mean.Terms like "APR," "compound interest," "derivatives," and "asset allocation" just leave users scratching their heads before they even think of the next step to take. They make the product look intimidating for an average user or like it was created specially for experienced financial professionals.ChatGPT can act as a translator for those technical terms. It can even be a unique value proposition for your product. Include it as a feature, a page or an icon beside those technical terms for users to click on and check for meaning. It alsoe makes the app look more approachable

Lack of accessibility

Many Fintech apps and websites are not designed with accessibility in mind, and this creates barriers for users with disabilities. Some of these issues are screen reader compatibility, insufficient colour contrast, left-hand mode and lack of alternative text for images.ChatGPT can help with providing alt text for images, providing text descriptions of complex charts and graphs, and ensuring compliance with accessibility guidelines instead of doing it manually. An award for adhering to accessibility and inclusivity guidelines do not only attracts more users to your product but also makes your brand socially responsible.

Lengthy Onboarding

Government policies and restrictions are really affecting many fintech products, so the demands extend to the users. Onboarding is the equivalent of having to deal with many paper works just to get through a deal.For fintech, It can be either opening a new bank account, creating a an account for a lending tech app, or setting up an investment account. The business needs the KYC details to meet those government requirements. This can be a major source of frustration for users, leading to high drop-off rates.How about including ChatGPT to help users fill out subsequent forms with existing details? ChatGPT could provide instant answers and even pre-fill certain fields based on previous user information till the account is fully set up and functional.

Security Compliance

Fintech apps handle sensitive financial data, which makes cybersecurity and compliance non-negotiable. Itll be quite difficult to implement security measures and requirements to users in simple terms to ensure compliance. An average user see cybersecurity as an extra chore.ChatGPT can be the cybersecurity expert, to give users tips on how to safeguard their accounts or remind them of the security actions they have not taken.

Lack of Proactive Support and Guidance

Users need help from time to time, especially if they are using complex financial processes or understanding specific features within a Fintech app. Traditional customer support channels (e.g., phone and email) might be slow and inefficient.Support chatbots can run on ChatGPT technology to provide instant and personalised customer support. They handle common inquiries and, if the queries escalate, connect the user to a customer care agent. This allows human agents to focus on more complex issues, and the users are attended to quickly. It is a win-win situation.

15 Ways to Use ChatGPT for Your Fintech Product

Fintech designs usually have a lot of back and forths, and thats totally fine! Thats the price you pay for your users. But do you have to pay for it all? I mean, when you now have ChatGPT. It is the real boost you need; it is not just for helping you write.

User Research

Research is tedious and time-consuming, it also requires accuracy. Even though UX research requires more human interaction and empathy, ChatGPT can still help with tasks that can be automated.Market Research: Instead of sitting alone and sifting through dense reports, you can use ChatGPT to quickly summarise the market reports you have. You can even go the extra mile to ask for important trends and generate demographics for the trends.Ask ChatGPT to provide a list of competitors your product or business would have. Additionally, ask for a short list of the offerings from that competitor in order to validate its response.Do you want a SWOT analysis of your competitors? Just attach the link to their website and ask ChatGPT to provide the strengths, weaknesses, and unique value propositions of each of the competitors you shortlisted from the answers received from the previous response.Develop Research Questions: Once youve identified your target users, competitors and other background information. Its about time to start hearing from users, right? So, how do you use ChatGPT on users?Research questions! The structure of your question determines the value and quality of the information youll get from users. If you already have an idea of the question, you can let ChatGPT rephrase your thought process to encapsulate it properly.Even if you do not know the ideal questions, it can still help you, provided you give it sufficient information.Instead of just telling your users, "We need to know how you feel about security,"You can prompt ChatGPT to Develop five open-ended research questions about users' concerns regarding mobile banking security." This might give us questions like:

1. What are your biggest fears when using mobile banking?
2. What measures would make you feel more secure using a mobile banking app?

Develop How We Might: How We Might is a popular ideation technique for getting solutions for the user's problems. HMW questions frame our design challenges as opportunities to be tapped into.If our challenge is improving financial literacy among young adults, we could prompt ChatGPT with "Generate 10 'How Might We' questions to improve financial literacy among young adults using a mobile app."The result will give you questions you can easily answer depending on your knowledge of the product. E.g.,

1. How might we gamify learning about personal finance?
2. How might we provide personalised financial advice in a simple and accessible way?

Develop Usability Scripts: Usability scripts help maintain consistency and cohesion to accurately assess how users interact with a product, website, or application. ChatGPT can assist by generating initial drafts of tasks and questions.

UX Design

UX design basically translates all the research into tangible design direction for the UI and visual designers. Lets see how ChatGPT can help with refining the ideas to properly fit into the user needs.It gives you a variety of user scenarios: You already have your ideal audience in mind, and you know their pain points, but how do you understand how these pain points fit into their everyday lives?User scenarios! It puts you into perspective of what their everyday lives look like.Let ChatGPT help flesh out realistic scenes of their daily interaction so that you can focus on refining the ideas and making them as accurate and empathetic as possible. It gives you options for human interactions since no human experience is the same.Provides User Stories: User stories add more depth and focus to user needs and help you prioritise features that will meet those needs. ChatGPT can assist in generating user stories based on user scenarios or desired functionalities.You can have a scenario based on Sarah's notification about exceeding her spending limit. Your ChatGPT prompt can be something around this:Create a user story for Sarah receiving a spending limit notification."The user story will then look like"As Sarah, I want to receive a notification when I exceed my spending limit so that I can be aware of my spending habits and adjust accordingly."Accessibility Guidelines: This has become very important in UX Design. ChatGPT can be your partner in ensuring that all accessibility guidelines are strictly adhered to in your product. It will give you quick access to accessibility guidelines and best practices, ensuring our designs are inclusive.Arranging your Content: You can have perfect UX Copies, animations, videos, etc, but arranging them in a frame will do the work of presenting to your users. If it is arranged, like the fish in a tuna fish can, it has no chance with users.ChatGPT helps you make sense of all the content you have for your product. It will assist in structuring content and creating intuitive navigation by generating sitemaps user flows and suggesting optimal content hierarchy.Ask ChatGPT to create a sitemap for your product. Give it the context of all the pages you have planned and features you havent decided where to place yet.

UI Design

This is the part of designs users are most likely to interact with. It requires a well-done UX Design and research work for the users to have the best digital experience. White chatGPT won't connect the pixels to form a full-fledged design; it can bring pieces for you to bring together to achieve your desired result.Helps with colour Palette: As designers, we all know the basic psychology of colours, e.g blue and white, are mainly associated with trust. Most Fintech products use blue, but how do you find the perfect shade and colour combination to make your design stand out?Sometimes, the client or brand doesn't have a colour, and you have to start from scratch. Trial and error with colour pickers would take time, so you can use ChatGPT to explore different palettes based on specific moods or target demographics. Ask ChatGPT to provide some colour palette suggestions for your product.Typography selection: Typefaces are an important element of brand identity. In the the quest to be unique, you might have to do a lot of research before settling on a typeface. Save yourself and your team time and stress; you can just describe the look you want to achieve and let ChatGPT do the work.It could also be that you want a font combination that communicates stability and professionalism for a corporate banking app. To this end,you could prompt ChatGPT with "Suggest font pairings for a corporate banking app that convey stability and professionalism, considering both headings and body text."It would suggest pairing and explain why these choices work well together. You can then go on to visualise it.Draft design and brand guidelines: ChatGPT has limitations, like creating designs, but it can help you with the important elements you should include in your brand design guidelines. These include defining brand voice and tone, outlining typography usage rules, and establishing guidelines for microcopy.This will give you a solid textual foundation we can translate into visual design elements, ensuring consistency between our brand messaging and UI.

UX Writing

This is where ChatGPT has more strength if used properly. UX Writing carries the efforts and solutions of the designers in words, and it is the first stage of interaction with users.It helps with brand adherence: One of the things that builds trust in FinTech products is their consistency across all touchpoints. Consistency in subtle elements like CTAs, form entries, confirmation copies, etc.Human errors are bound to happen. You can remove any iota of doubt in the minds of your users if you let your copies through the eagle eyes of ChatGPT.It can refine your UX Copies: Error messages, tooltips, and onboarding text are critical for a smooth user experience. ChatGPT can help us refine these elements to be clear, concise, and helpful.For writing actionable CTAs: Are you struggling to get your design to convert? Or maybe your go-to CTA is BUY NOW. CTAs are supposed to be persuasive and relevant to the content of the design. ChatGPT will help you turn your Buy Now into. Get yours for 10% off only today.It gives you various options to choose from and perform A/B Testing. It also explains why each option would be a good fit for the design.It understands basic SEO principles: The main function of ChatGPT is not SEO but can help with basic SEO activities that can help your design gain more visibility. You can tell ChatGPT to help you with a list of primary and secondary keywords that would rank well in the fintech industry.