Designing Use-Case Prompt Suggestions - NN/G

Find the training option that fits your learning style.
June 27, 2025 2025-06-27
Use-case prompt suggestions are intended to help users understand what they can use the AI tool for and how to interact with it. When designed well, they set accurate expectations for the system and guide users toward effective prompting.
Use-case prompt suggestions are examples of good AI prompts. They are displayed within most AI tools, usually (but not only) to novice users.
Unlike other types of prompt suggestions, which focus on continuing a conversation (followup suggestions) or helping users complete the prompt (prompt autocomplete), use-case prompt suggestions aim to support learnability — helping users quickly understand what the genAI tool can do for them and how.
Use-case prompt suggestions vary widely in complexity — from short phrases or single-sentence examples to rich, realistic formats like entire conversations, images, or videos that show full user–system interactions.
Choosing the right level of complexity depends on several key factors:
Simple, clickable prompts placed inside or near the input field work well for broad systems and low-complexity tasks.
Complex prompt suggestions are better suited for specialized systems and high-complexity tasks, where rich examples provide more meaningful guidance.
While users are more likely to engage with quick, actionable prompts, rich examples can help them develop a deeper understanding of how to interact with the system effectively.
Prompt complexity also affects how suggestions should be displayed. Simple prompts take up little space, allowing multiple suggestions to be shown at once. This is ideal for inline layouts or suggestion lists.
Complex prompts, however, require more screen real estate and are better presented one at a time, such as in a carousel, expandable section, or example library.
Ultimately, designers must balance two competing goals when choosing the complexity and format of prompt suggestions:
The primary goal of helping users understand what an AI tool or feature can do is to encourage them to start using it and ultimately increase adoption. This is especially important for unauthenticated users, as organizations hope to convert them into long-term, paying customers.
At this stage, users are typically:
To support this last decision, companies often display simple, curated examples in the preauthentication view. These static suggestions act as marketing nudges: they are quick to scan, easy to understand, and designed to clearly communicate the tool’s key capabilities, ideally motivating users to create an account and get started.
For curated use-case prompt suggestions to be effective, their content must be tailored to the user’s level of experience and likely objectives. This alignment helps showcase the AI system’s capabilities in a way that feels relevant and actionable, while also setting realistic expectations for interaction.
A critical consideration is deciding which strengths of the AI system to highlight. Depending on the tool’s purpose and audience, prompt suggestions might emphasize abilities such as:
The features showcased should reflect potential users' most common goals. The more closely these suggestions align with user needs, the more likely they are to drive engagement and convert new visitors into active users.
While most AI tools support a wide variety of use cases, it’s the product team’s responsibility to identify a focused set of high-impact capabilities. Highlighting the most broadly relevant strengths ensures that the prompt suggestions appeal to the largest possible share of users — especially in early interactions like the preauthentication experience.
These curated examples often function as push revelations — predefined content shown proactively, regardless of user context. While useful for showcasing features, push revelations have well-documented usability pitfalls: they can be intrusive, easily ignored, and quickly forgotten. If these suggestions disrupt rather than assist, they risk undermining their intended purpose.
To be effective, prompt suggestions in preauthentication views should follow the same usability principles as other types of onboarding. They must be:
Used thoughtfully, these static prompts can serve as lightweight introductions to system capabilities. But if overdesigned or poorly timed, they may frustrate users rather than support them.
There are several ways to visually present prompt suggestions in preauthentication views, each offering distinct advantages and tradeoffs.
Pills are often used to display prompt suggestions as single terms or short phrases. This format is ideal for highlighting a wide range of topics or system capabilities, helping users quickly understand what the AI tool can do.
Pills should be interactive — either triggering the prompt directly or inserting a longer prompt into the input field for users to edit before submitting. In preauthentication views, clicking a pill typically leads users to sign in or create an account to continue, in case they must be logged in to use the tool.
Using cards to display use-case prompt suggestions allows for in-depth examples of how to interact with the system. Using cards for prompt suggestions works well when:
Like pills, cards should be clickable and direct users to sign in or create an account (if required).
Videos are a powerful way to demonstrate how a system works and educate users on its capabilities and interaction patterns. They’re especially useful when users have little to no prior experience with generative AI tools, as they can offer rich context and realistic examples of effective usage.
Using a carousel is an effective way to showcase multiple curated examples of how to interact with the system. It offers both breadth and depth — allowing users to see a variety of use cases while presenting each one individually. Carousels make it easy to highlight complex interactions and rich examples of the system’s capabil