Good startup ideas rarely arrive out of nowhere. Most start as someone else's problem, and a handful of newsletters do the work of collecting those problems, naming who has them and sketching what might fix them.
The short version: look where ideas come with their sources attached. Our picks are Half Baked, Ideabrowser and Starter Story, plus Y Combinator's Requests for Startups and, for the Canadian picture, BetaKit. Read them for patterns rather than ideas to copy, mix in what you already know about your own trade, and test anything you like with a simple page before you build it.
Which newsletters are best for startup ideas?
The ones that name their sources. An idea with a real person's problem behind it is worth ten ideas generated from a keyword list. Here's what each one gives you.
| Source | What you get | Cost | Our take |
|---|---|---|---|
| Half Baked | One idea per issue, with the problem, a solution, and credit to the founder or investor who sparked it. Says it has 175,000+ subscribers | Free, daily | Our favourite. Short, specific, and often links to a quick demo build |
| Ideabrowser | A free "idea of the day" by email with a market report, from Greg Isenberg and his team. Free members get 24 hours to read each one | Free, paid plans available | Good for volume. Treat the market sizing as a starting point, not proof |
| Starter Story | A weekly "founder-tested business idea" plus a database of real businesses and what they earn. Says 200,000+ founders read it | Free newsletter | Best for boring, profitable ideas. Now owned by HubSpot Media |
| Trends.vc | Free 5-minute reports on emerging markets, by Dru Riley | Free, paid Pro tier | Strong archive. The reports aren't dated on the site, so check how current each one is |
| Exploding Topics | A weekly email of fast-growing search topics, from a tool owned by Semrush | Free, weekly | Trends, not ideas. Useful for checking whether interest is rising |
| YC Requests for Startups | Ideas Y Combinator's partners want to fund, currently for the Winter 2027 batch | Free (a web page, not a newsletter) | The most ambitious list here. Read it to see where investors think money is going |
| BetaKit | A weekly newsletter of the top Canadian tech and startup stories, from an outlet covering Canadian startups since 2012 | Weekly; cost not stated on its signup page | The only Canadian source here. Read it to see what Canadian founders are building and which problems they're chasing |
If you only subscribe to one, make it Half Baked. If you want ideas you could run from a laptop next month, add Starter Story.
What makes an idea source worth your time?
In our view, three things separate research from brainstorming. First, a real person or a real complaint sits behind the idea, named and linked. Second, you can see money changing hands somewhere, which is why Starter Story's revenue figures are so useful. Third, it's dated, so you know whether the moment has already passed.
A list of "AI for X" ideas with nobody behind them is a brainstorm. Fine for a spark, but don't plan a year around it.
What does a good idea look like in these newsletters?
Here are a few from recent months, to show the range.
Half Baked #675 (August 13, 2026) describes a monthly family newspaper: relatives send photos, voice notes and messages to a private number, and AI turns them into a printed and digital paper. Half Baked #631 (June 12, 2026) sketches a service that runs gift campaigns for sales teams, and credits the idea to Sam Blond, co-founder of Monaco.
Half Baked #723 (October 20, 2026) pitches a to-do list where adding a task sets an AI agent working on a first draft. And on Y Combinator's current list, partner Jared Friedman asks for AI that writes well, arguing that today's models got better at code and worse at writing.
Notice what they have in common. Each one starts with an annoyance people already have, not a technology looking for a use.
How do you turn inspiration into your own idea?
Use the newsletters for patterns, not for ideas to copy. Half Baked says it has 175,000+ subscribers, so the idea in today's issue lands in a lot of inboxes at once. Your edge is what those readers don't know: your trade, your city, your customers.
Take the shape of an idea and move it somewhere you know well. "Gift campaigns for sales teams" might become thank-you boxes for real estate agents to send after closing. "A to-do list that does the work" might become one built only for contractors' quotes.
Apart from BetaKit, none of these sources focus on Canada. Before you run with an idea, check that it works here: Canadian rules, smaller cities spread far apart, and two official languages. Sometimes that gap is the opportunity, because the Canadian version of an idea from elsewhere doesn't exist yet.
Then keep a running note for 30 days. Every time you or a customer says "I wish there was something that...", write it down. By the end of the month, you'll have a list nobody else is reading.
How do you test an idea before you quit your job?
Put a simple page in front of real people before you build the product. Describe the idea in one sentence, say who it's for, and ask for an email address or a deposit. If nobody signs up after you've shown it to a few hundred of the right people, you've saved yourself a year.
We've written about this habit before in Stop Waiting for Perfect: Why Testing Beats Overthinking. The gist: launching early gets you real feedback that no amount of planning can.
If you'd like help making that first test look credible, that's what we do at Loud Launch. A brand starter (three logo concepts, a brand guide and a brand voice) starts from $300, and a website from $1,600, each at one fixed price before we begin. If you can build the page yourself, do it. Either way, book a free call and tell us which idea you're chasing.
Drafted with AI help from Julian's notes and checked against the sources above.
