Which tool to choose
You don't need all of them. Choose based on where you're stuck.
If you have a completely blank page and you need a volume of ideas to filter: start with a conversational assistant like ChatGPT. Free users can search the web, analyze data, upload images or files and generate images — enough for serious brainstorming without paying.
If you already have the topics but don't know which ones land: go to Google Trends. It's free, with no search limits, and it tells you what's rising right now in your niche.
If you want the exact questions your audience types: AnswerThePublic visualizes the real questions around a keyword. Free accounts have a daily search cap (currently a few per day, check the current value before starting): use it sparingly on broad seed words.
If your target is specifically YouTube: VidIQ gives you ideas tailored to your channel. On the free plan, though, results per category are limited (matching keywords, related ones and questions). Good for exploring, tight for working on it seriously.
The zero-cost combination that covers most beginners: ChatGPT to generate + Google Trends to validate. The rest are upgrades for when the channel grows.
How to do it
From PC or smartphone the path doesn't change: both tools work in the browser and in the app.
Open ChatGPT (chatgpt.com or the app). Don't write "give me ideas for a blog": it's too vague and you get clichés. Instead paste the prompt below filled in with your data.
The working syntax:
You are a content strategist. I need to fill the editorial calendar
of my [blog/YouTube channel] on the topic [e.g. vegetarian cooking for
people short on time]. My audience is [e.g. workers 25-40 who cook
in the evening, beginners].
Generate 25 content ideas divided into 4 categories:
- 7 "evergreen" ideas (always valid, searched all year round)
- 7 ideas that answer a concrete question or problem of the audience
- 6 "contrarian" ideas that challenge a common belief of the field
- 5 ideas tied to seasonality or recent trends
For each idea give me: proposed title, the specific angle (why it's
different from the usual), and exactly who it speaks to. Avoid generic titles.
Read the 25 ideas and mark the 6-8 that spark something for you. Discard without mercy the ones anyone could write.
Check: if more than half of the ideas sound generic ("10 tips for..."), reply in the same chat with this:
These are too generic. I want more specific and original angles, raise the level and cut everything the competition has already published.The model corrects course.
Now validate. Open Google Trends (trends.google.com), type the topic of one of the ideas you like best and press enter.
Scroll to the Related queries section and filter the entries on "Rising": these are the hot topics and the questions people are asking right now.
If you make videos, change the search type from "Web Search" to "YouTube Search" in the dropdown menu: you'll see what's being searched on the video platform, which often differs from the web.
Check: if a "Rising" query talks exactly about your angle, you have a topic with growing demand and probably little competition. If instead Trends shows "not enough data", the topic is too niche or no one searches for it: keep it only if you have a strategic reason.
Concrete example
Marco wants to start a YouTube channel about balcony gardening but he has only three vague ideas and hasn't launched for weeks.
He pastes the prompt above with topic "balcony garden" and audience "thirty-somethings in the city, zero experience, small balcony". ChatGPT returns 25 ideas. Among the contrarian ones one stands out: "Why you should stop watering every day to grow better tomatoes". It strikes him, but he doesn't know if anyone cares.
He goes to Google Trends, types "balcony garden", switches the dropdown menu to YouTube Search and scrolls to the Related queries. He sees "how often to water tomatoes" marked "Rising". Confirmed: the demand exists and it's alive, and his counterintuitive angle intercepts it.
He opens his notepad and in twenty minutes he has five validated titles, not made up at random. The first video — "You water your balcony tomatoes too often (here's the proof)" — answers a real search. Marco is no longer guessing: he's answering questions people are already typing.
When it does NOT work (and how to fix it)
If the AI churns out only generic and predictable ideas
This happens when the prompt is poor on context. Add a precise audience, tone and a creative constraint ("no title that starts with a number", "angles the competition ignores"). Then ask in the same chat to combine two distant ideas: unexpected crossings are where original content is born.
If ChatGPT stops halfway through brainstorming
On the free plan there's a cap on the number of messages with the most capable model within a window of a few hours. Past the threshold, the chat automatically switches to a lighter version until the reset. The exact values change often, so check the current ones. Solution: concentrate everything into a few rich messages instead of many short exchanges, or continue with the lighter version (it holds up well for brainstorming) or pick it back up after a few hours.
If Google Trends says "insufficient data"
The topic is too specific for the volumes of Trends. Broaden the seed word: instead of "organic fertilizer balcony tomatoes" try "balcony garden". Then work on the related queries that emerge, going back into detail from there.
If you've used up your daily AnswerThePublic searches
The limit changes over time, so check the current one. Before starting, make a list of your seed topics and spend the available searches on the broadest possible words: each one generates dozens of questions, so a few well-chosen queries cover a lot. For the rest, Google Trends' related queries do a similar job without limits.
A tip from someone who actually uses it
The number-one mistake is treating the AI's output as a finished product. It isn't: it's raw material. Good ideas almost always arrive on the second or third round, when you tell the model "no, too obvious, try again sharper".
And never skip validation on Trends. Keep in mind that Google Trends doesn't show absolute search volumes: it uses a scale from 0 to 100, where 100 indicates the peak of popularity in the selected period. It's not a traffic figure, but it tells you the direction — and the direction (up vs down) matters more than the exact number when you decide what to bet on.
Frequently asked questions
Which prompt works best for ideas?
The one that gives context and constraints. A prompt that specifies audience, format, categories and a creative limit always beats the generic "give me ideas". The prompt in the "How to do it" section is ready: copy it and change only topic and audience.
Can I use Google Trends for videos as well as blogs?
Yes, and it's one of its most underrated functions. Enter the topic and use the dropdown menu to switch from "Web Search" to "YouTube Search": you'll see the top or rising topics in your niche directly on the video platform. YouTube trends often differ from web ones, so check the right source.
Do I need a paid tool like VidIQ to generate ideas?
Not to start. VidIQ becomes useful when you grow and want ideas tailored to your channel's data, but the free plan already covers limited keyword research, daily ideas, basic analytics and a Chrome extension. For pure brainstorming, ChatGPT plus Google Trends cover the same need at zero cost. Consider the upgrade only when you publish consistently and the limit really holds you back.
Do AI-generated ideas make my content "copied" or unoriginal?
No, and it's the most widespread misunderstanding. The AI gives you the starting angle, not the content: originality is born from your direct experience, your voice and the examples only you can tell. Thousands of creators start from the same tool — what distinguishes your video or post is what you put into it. The idea is 5% of the work; the execution and real expertise are the rest, and those stay yours.