Building a Chrome extension for AI context management has been one hell of a ride. When you're a solo founder (or close to it), every user feels like a small victory, and every plateau feels like you're pushing a boulder uphill. I'm at about 2000 users now with AI Context Flow, and honestly, getting here wasn't about one magical growth hack - it was about trying everything, failing at most things, and doubling down on what actually worked. Here's the honest breakdown of how we grew from 0 to (almost) 2000 users.
Here's a breakdown on how it worked
First 50 users = Friends and family
Went on calls with my school friends, college friends, earlier colleagues, and anyone who was interested in AI prompts and memory and was willing to test out a shitty first version of the product. Nothing fancy, just showing people what I built and asking them to try it.
50-250 users = Prompt engineering workshops
I offered prompt engineering workshops to universities and offices. Did 50+ emails and then did 20+ sessions. The smallest session brought me 5 users, the best one brought me 60. This was exhausting but honestly really valuable - you learn so much about what people actually need when you're teaching them.
250-1000 users = Product Hunt launch
We did Product Hunt and came #1 product of the week and #1 productivity tool of the week. This gave us the biggest inflow of users (~900 users). There was a significant churn afterwards too but it settled at about 1000 users. The spike was insane but keeping them was the real challenge.
The real slog: from 1000-2000 users
This phase... man, this was tough. The movement from 1000 to 2000 was slow and sluggish and no quick growth hacks were working here. So we did two things:
1. CSO (Chrome Store Optimization)
We focused on our Chrome store listing to make sure we were being discovered by the right people. The key here was capturing user intents and what "problems" users are actually trying to solve. Not just keyword stuffing, but really understanding the pain points people type into the Chrome store search. A good trick is to put yourself in the shoes of a user that is trying to solve the problem your product is built for but has no idea your product (or any product in that category) exists, what would they type? YES! Add that to your short and long description. Even better if you could front load the right keywords in your title (ask any AI agent if you want to know what front-loading means).
THIS IS THE HIGHEST LEVERAGE TASK YOU COULD DO FOR YOUR PRODUCT!
Don't half-ass it please :)
2. SEO/GEO (and yes, GEO is a thing now)
I started writing blogs. I've never done SEO before and now there is this new monster of GEO (i.e. how to rank on LLMs). Anyways, a founder wears many hats so I went deep into it.
Since my website domain was new-ish, I went for long tail queries. Like really long tail. The strategy was to target super specific problems that people search for - the kind of queries that big sites don't bother with.
Some of my long tail articles really hit it:
These started getting picked up in AI overviews too! That's when I realized GEO is real and it's wild how LLMs are surfacing content differently than traditional search.
2 months later, 30% of our traffic is coming from blog posts!
Also tried a few other methods like social posts, cold DM'ing people etc... had some results but nothing major.
This feels like a challenge now - how do we reach 10k users by end of this year? I don't know the answer to this yet, but once I do, you might see me writing again :)
Now, back to fixing bugs. If like me, you are also tired of re-typing prompts with correct context a gazillion times a day, would love you to try AI Context Flow.
Adios!
Top comments (2)
Bravo on the number growth! Looking forward to see your share when you reach 10k this year
10k let's goooooo