
Staring at a blank page for twenty minutes before writing a single idea is a common problem, and prompts can help you break out of it. Now I open a chat tool, type one plain sentence, and have a messy list in under a minute. The list isn't always good. That's fine. A bad list is still easier to work with than nothing.
Here's the thing most people miss, though. The AI doesn't pick your ideas. It just starts the list. You're the one who decides what's worth keeping, and that part matters more than any clever prompt.
Below is a simple process you can use for blog posts, gift ideas, weekend plans, side project names, and pretty much anything else where you're stuck. It takes about fifteen minutes once you get the hang of it.
Short on time? Our top picks: ChatGPT, Claude, Google Gemini and Microsoft Copilot. See all picks

Before you type anything, figure out what you're actually trying to come up with. Then write it as one plain sentence. Not a paragraph. Not a list of background details. One sentence.
For example: "I need blog post ideas for a small website about houseplants for beginners." Or: "I want weekend activities for two kids, ages 6 and 9, who are bored of the park."
Why so short? Because when you cram in too much at the start, the AI tries to please every detail at once and the ideas come out flat. A clear, simple topic gives it room to stretch. You'll add details later, on purpose.
Tip: If you can't write the sentence, that's useful information. It usually means you're not stuck on ideas yet. You're stuck on the goal. Spend two minutes on that first.
Now ask for a list. Asking for ten is usually a good default. It's enough to get past the obvious stuff but not so many that I start skimming.
A prompt that works well for many people:
"Give me 10 ideas for [your sentence]. Keep each one to a single line. Include a few unusual ones, not just the obvious picks."
That last part helps. Without it, you'll often get the same safe ideas you'd find on page one of a search. Asking for a few odd ones shakes things loose.
When the list comes back, resist the urge to react. Don't delete. Don't say "no, not that." Just read it once. Quantity comes first here. The first few ideas are usually the most predictable, and the interesting ones tend to hide around number six or seven.
Tip: If the whole list feels boring, don't start over. Just type "10 more, weirder this time." I do this all the time and the second batch is often better.
This is the step that turns a generic list into ideas you could actually use. Limits make ideas better. It sounds backwards, but it's true. "Dinner ideas" is endless and useless. "Dinner ideas under 30 minutes with what's left in a fridge after a busy week" is something I can work with.
Pick one or two constraints that match your real situation:
Then ask again: "Same topic, but every idea has to work for someone with only 20 minutes and no special tools."
You'll notice the ideas get more specific and more doable. Some of the flashy ones from Step 2 will vanish, and honestly, that's often a good thing.
Tip: Don't pile on five constraints at once. A common result of that approach is ideas so narrow they barely count as ideas. One or two is the sweet spot.
By now you've probably got 20 to 30 ideas on screen. Scroll back through all of them and pick three you actually like. Not the ones you think you should like. The ones that made you go "oh, that's kind of fun."
Copy those three into your notes. Getting them out of the chat window helps more than you'd think. It makes them feel like yours instead of the AI's.
If you're torn, here's a quick gut check you can try. Ask yourself:
An idea that gets a yes on all three is usually a keeper.
Tip: It's fine if one of your three is a mashup. Sometimes idea #4 and idea #12 are both half-good, and together they're great. Combine them yourself before moving on.
By the way, if you'd like a few ready-made starting points for this kind of session, I put together a Free 25 Writing Prompts Starter Sheet you can grab by joining the email list.

Now take your single favorite from the three and go deeper on just that one. This is where brainstorming with AI really earns its keep. Getting a list is easy. Getting five different angles on one good idea is where it saves me real time.
Try something like:
"Here's an idea I like: [your idea]. Give me 5 different versions of it. Change the angle, the audience, or the format each time."
Say your idea was "a blog post about plants that are hard to kill." The variations might include a version for people in dark apartments, one for folks who travel a lot, a short checklist version, a "mistakes people make" story version, and so on. Same core idea, very different posts.
You can also ask follow-ups like "What would make this more surprising?" or "How would someone explain this to a friend in one sentence?" That second one is a great test. If the one-sentence version sounds dull, the idea might need more work.
Tip: Stop after one or two rounds of variations. It's easy to go down a rabbit hole of "make it more fun, now more serious, now shorter" and end up right back where you started, just more tired.
This is the part I care about most. The AI has no idea who reads your blog, what your customers keep asking about, or what your kids will actually sit still for. You do.
So look at your favorite idea and its variations and decide. Trust what you know about your people. If something sounds clever but you know your audience would shrug at it, skip it. If something seems plain but you've heard people ask for exactly that, go with it.
Picking "boring" ideas over flashy ones often pays off, because the boring ones tend to solve a problem someone actually has. The AI tends to reach for novelty. Real people mostly want something useful.
Tip: Write one line in your notes about why you picked it. Something like "picked this because readers keep emailing about watering." Future you will thank you when you're deciding what to do next.
If you end up doing this a lot, the Everyday AI Prompt Pack and the Writing Prompts Bundle below this post have more prompts like the ones above, so you're not starting from scratch each time. Totally optional, though. The steps here work fine on their own.
This happens a lot. Usually the topic sentence is too broad, or you didn't ask for anything unusual. Try adding "avoid the obvious answers" or ask for ideas from a specific point of view, like "ideas a retired teacher might suggest." Changing the point of view is my favorite fix.
Go back to Step 3 and tighten your limits. Tell it your real budget or real time. The AI might suggest a "pop-up event" even when you have only one free evening and no money. A quick "keep it realistic for one person with no budget" fixes that fast.
That's okay. Sometimes it's a sign you need a break, not a better prompt. Other times, telling the AI what you don't like helps. "These feel too salesy" or "these are too long" gives it something to steer away from.
Brainstorming tools can make things up, especially numbers, names, and dates. Treat anything that sounds like a fact as a guess until you check it yourself. For ideas, that's not a big deal. For anything you publish, double-check it.
Set a timer. Seriously. Fifteen minutes is plenty for one round. If you're still going at forty minutes, you're probably polishing instead of deciding. Pick one and start.
You don't need anything fancy to follow the steps above, but a few tools make the process smoother. Here are some well-known options, depending on how you like to work.
A good fit if you want a general chat tool for lists, variations and follow-up questions. Check which features come with the free plan and which need a paid one before you rely on them.
Suits people who like a conversational back-and-forth and plain, readable replies. Check the usage limits on the plan you pick, since long brainstorming sessions can run into them.
Handy if you already live in a Google account and want brainstorming close to your other tools. Check how it handles your data and what privacy settings you can control.
Worth a look if you work in Microsoft apps and want idea help nearby. Check which version you actually have access to, since options differ between personal and work accounts.
Suits anyone who wants one place to paste the ideas they like and note why they picked them. Check whether the built-in AI features are included in your plan or sold as an add-on.
Good for visual thinkers who want to drag ideas around on a board and combine them. Check how much of the board and AI help is available on the free tier before you invite a team.
A simple option for saving your three favorite ideas and the reasons behind them. Check how well it syncs across the devices you use before moving your notes over.
Please check current plans and prices on each brand's own site before you sign up.
No. The free versions of most chat tools handle brainstorming just fine. Paid versions may have extra features, so check the tool's own page if you're curious what's included.
Usually not. It's a lot like asking a friend to throw out suggestions while you think out loud. The judgment, the choosing, and the actual work are still yours.
Ten is my default. Fewer than that and you don't get past the obvious picks. Way more than that and I start skimming instead of reading, which defeats the point.
Yes, and you should. Once you have a few prompts that work for you, save them in your notes and just swap in the new topic. That's how I keep the whole thing quick.
Pick one thing you've been stuck on this week and run through the six steps with it tonight. Just one. If you want a head start, grab the Free 25 Writing Prompts Starter Sheet by joining the email list, so you've got some ready-to-use prompts on hand. And if you find yourself brainstorming often, take a look at the Everyday AI Prompt Pack and the Writing Prompts Bundle below this post to see if either one fits how you work.
Printable checklists and prompt ideas for everyday tasks. Always check AI output...
See details