Suggested prompts

Why does suggested prompts matter?

One of the main problems with AI tools has been staring at a blank canvas and not knowing where or how to start. Blank screens can be intimidating, and not because they don't have ideas, but because they wonder:

What can it do? What should I try? Where do I start?

This can kill adoption before it begins.

Suggested prompts solve this by removing the friction of first action

They give users a way in: context-aware examples, preset queries, or one-click shortcuts that show what’s possible

For example, it's easier to just click a button “more diplomatic”, instead of thinking how to prompt AI the right way to improve an email. Quick actions like these help users try out things faster, achieve their intent, and learn the tool's capabilities.

Time to take a closer look now.

Key takeaways

1. In the right context

The most useful suggestions are embedded directly in context. Users don’t have to pause, switch modes, or guess what the system can do, the prompt appears exactly where it’s needed.

Grammarly drops actions like “Reply quickly” right next to an email, or "Rewrite with Grammarly". These appear right next to the content, so users can try what’s possible with one-click.

Monday.com offers “Generate formulae” at a second level, when user wants to generate a formula for the table.

 

2. One-click test

Suggested prompts lower the cost of experimentation. Users don’t have to commit - they can just test an option, see what happens, and move on if it doesn’t fit.

Apple’s writing tools and Grammarly, you can try it in a single click. If they are in the right context, they should give the results in one click.

Apple’s writing tools keep things simple too. "Friendly" or "Professional" is just one click. They are easy to act on and in context, never breaking the flow.

No prompt writing brings no confusion and faster results.

Setting tones in Grammarly, "make it assertive", is a great example of one-click outputs. You get the results, keep it if you like, undo if you don't.

 

3. It doesn't have to be buttons

Figma uses a visual tone slider as a quickstart tool. Users drag a dot across options such as Professional, Concise, or Casual to see instant results. Pre-built tone preset makes the complete experience even more intuitive and quick.

 

4. Show the system is working

Figma's “Adjusting tone...” status message reassure users that the system is actively working. Apple's shimmering effect on the block of text that's being iterated upon is another way to show the processing state.

 

5. One-click undo

Users try more when nothing feels permanent. All changes should be reversible with one click.

Apple’s 'Revert' and Figma’s undo make it safe to experiment with suggested prompts.

 

6. Pre-built templates

Presets don’t stop at one-off actions (at the start of the flow). They can be integrated throughout the user flow.

Monday.com integrates suggestions like “Write me…”, “Summarize doc”, “Complete text”, and "Generate formulae" directly into the user's workflow. These actions are placed beside items, formulas, or documents through intuitive touch points.

Designing suggested prompts isn’t just about adding convenience. The challenge is to guide and teach without overwhelming or creating dependence.

 

Editor's notes

As a practitioner, I like to write notes — key takeaways and questions — to ask myself whenever I'm designing suggested prompts interaction in the future.

It’s a tangible gut-check for myself and for you to steal, if you see fit.

 

Future play

Suggested prompts today are mostly static hints. The future is in adaptive and evolving prompts that shape themselves to the user, the context, and even the team. Imagine opening a document and seeing not just “Summarize this” but context-aware prompts like “Compare this with last quarter’s results” or “Draft a client update from these notes.”

Suggested prompts in the future will be personalized shortcuts that anticipate what’s next in your flow (closely linked to Proactive suggestions AI pattern). Teams, too, will get their own shared libraries of prompts, turning one person’s best practice into everyone’s quick start.