HOW TO FIND PROFITABLE NICHES
I reached $12.5k MRR in a crazy crowded market (dictation apps)
How?
By finding an unserved niche - people with disabilites
Here's how you can do it too.
Step 1: Get early customers/data
When you launch, go after everyone, everywhere.
MARKET AS HARD AS YOU CAN.
Just get users in any way possible.
Reddit, Linkedin, X, ads, whatever.
Just get as many users as possible.
Step 2: Analyse your users
Now you need to know who signed up.
The best ways to analyse users (ranked):
1. User calls (you need to do these anyway)
2. Use their emails to find them online. Find out who they are and what they do
3. Check their product usage. What are they using your product for the most
Step 3: Ignore the most obvious use-cases.
In my video, I analyse the https://t.co/OoJSy4Y4kh users.
Obviously most users are SaaS and agency owners.
They want this product the most + I've been marketing in places where these people show up.
BUT YOU DONT WANT TO GO AFTER THE OBVIOUS USERS
Every single competitor is already going after them.
So go and find the 3rd most common type of user.
In my case, it's:
- Freelancers
- Fractional executives
- Coaches
Step 4: Find out if that niche is big enough to market to
There are a million ways to do this.
My favourite way? Check if those people search for my product
E.g. how many freelancers & coaches are searching for SEO solutions
How I did it: I connected to the https://t.co/Db5bGADJxv MCP
I told it about this niche I want to go after.
I asked it to find relevant searches and search volumes
Turns out: There are people who search for this!
The numbers aren't massive, but they're not tiny either.
IF YOU CANT FIND PROOF THAT THIS MARKET IS BIG ENOUGH, PICK ANOTHER MARKET
DONT GO WITH YOUR GUT HERE. GO WITH DATA.
Step 5: Create a plan to target these people
I'm an SEO guy, so I'll use SEO to reach these people.
I just asked Romeo to create me a content plan and it did.
For the next few weeks, it will create and post that content.
And hopefully, I'll get some traffic and customers.
But that's it! If any of the steps fail, just try again and again.
Hope that was helpful:)
Good luck, have fun ❤️
@psen102 Hey man, let's connect. For the X-to-LinkedIn flow, a preview of how links, images and line breaks look on both platforms would be a useful demo. Catching formatting surprises before publishing feels like the real win.
@diogocapela Hey man, let's connect. A six-year look back at the products that kept improving after launch day would be a useful read. The launch gets attention, but the long-term progress is harder to see.
@hamen Hey man, publishing the exact previewed draft is a useful fix. The tricky case is a timeout after the post is already live. Showing how the saved attempt gets checked before offering Retry would make the recovery flow much clearer.
@uzzy_dev1@officiallevi_c0 Hey man, congrats to you both on getting Fitness Pace live. With workouts, food logs and movement in one app, can people start with just one and add the others later? Keeping that first setup light feels important.
@WareDash Hey man, running memory makes sense here. The tricky next case is when “wo wala” fits two recent items. A demo showing when the AI resolves it from context and when it asks for clarification would make the fix much easier to judge.
So far, from the general use cases, Google's new frontier AI model, Gemini 4 Hydrogen, worked well. I'm going to soon use it for detailed tasks, then. I'll share the feedback soon, but for day-to-day tasks, it is one of the good ones, honestly.
#ai#gemini#Google
@saifcodes_exe@convex Hey man, live session revocation + the token inspector makes this more useful than a basic login demo. Showing what happens in a second active tab after revoking a session would make the security behavior especially clear.
@madebyabhinav Hey man, protecting the newer draft is a useful fix. A clear warning in the stale tab, with a way to recover its unsaved text, would complete the loop: no silent overwrite, and no edits lost while resolving the conflict.
@morgancacm Hey man, pausing email connections until verification clears makes sense—trust matters more than a discount here. A waitlist update showing the mobile flow and its permissions could keep those 24 users engaged without asking them to bypass a warning.
@codervicky_ Hey man, 1,325 visits with six listings is a useful starting point. Adding outbound clicks to each listed SaaS would make the next update even more interesting—traffic to Tinolaunch is one thing, discovery for the makers is the real test.
@Nkire_E_chi Hey man, the calendar sync is the part I'd want to see tested. Two guests trying to book the last room at the same time would be a useful demo, alongside how the agent hands off when availability can't be confirmed.
@toankhontech Hey man, the PR evidence is the interesting part. A recording plus logs makes the change reviewable, not just “the agent says it works.” Capturing failed flows as clearly as successful ones would make that even more useful.
@rahulbhadoriiya Hey man, getting users while still making room for family and friends is a solid win. Imposter syndrome is loud, but actual people choosing to use what you build is concrete evidence to put beside it.