In an ideal scenario, the user is so good at prompting that they can reach a great outcome on their first try. Realistically, that's unlikely to be the case. This is why wayfinders are so critical as patterns: they help users get to great outcomes with less friction.

Follow ups catch users on step two, after they have given their initial prompt to the AI but before the AI has started to build its response.

The intent is not to hide a long questionnaire behind an open input, a dark pattern that takes a seemingly simple input and makes it unnecessarily complicated. Instead, follow ups are short clarifications to build trust by showing the user that the AI is on the right track, or let the user intervene proactively.

Successful follow ups act as sample prompts, borrowing context from the initial request, and giving the user the sense that the AI is moving with them. Consider combining follow ups with a plan of action to provide even more upfront transparency and control to the user.

Details and variations



Agency and accuracy

Complicated prompts can be hard to put together, especially for less experienced prompters. Like templates, follow ups can break up what might need to be a multi-stepped prompt into smaller pieces. Additional questions give the user the sense that the AI is responding to the user's input from the get go, instead of simply collecting more information.

Potential risks

Use shortcuts where possible

The introduction of RAG allows users to give the AI instructions through primary sources, references, parameters, and other tools instead of relying on open text to get their intent across. Wherever possible, offer shortcuts for the user like these advanced techniques to save them time and effort
Use when:
If the user's initial prompt is likely to be insufficient to get a great outcome, use follow ups to capture additional data up front instead of relying on the user to guide additional iterations.


Perplexity shows follow ups by default if the initial prompt was not clear
Figjam includes followups as part of the template to build the initial prompt
Jasper breaks each complicated prompt into several follow up steps
Jasper seems trained by its foundational prompt to check in with follow ups if the direction of the generation seems likely to veer off the rails takes a novel approach to followups by allowing users to define details like an outline as a follow up to the initial prompt
No items found.