Hey IH,
Posted here once before about Foldif — a Chrome extension that adds
highlight, margin notes, and cross-AI handoff on top of ChatGPT,
Claude, and Gemini. That post got 2 likes and 1 comment
I want to share the actual numbers and what I tried, because every
"0 to 1000 users" post I read skips the ugly middle. This is the
ugly middle.
THE NUMBERS
Launched: May 26, 2026
Today: June 2, 2026
Active users (last 7 days, GA): 30
Active users (last 30 days, GA): 149
Currently active in extension: 30
Total installs (Chrome Web Store): https://chromewebstore.google.com/detail/foldif-chatgpt-claude-fol/bfohkghddnennjbkhblikdplccneemfd
Paying customers: 0
Revenue: $0
Top countries (last 30 days):
WHAT I TRIED
Twitter DMs (~100 sent over 3-4 days)
Result: maybe 4-5 installs. Most DMs were similar templates, which
in hindsight is exactly the kind of pattern X flags. Felt like work,
returned almost nothing.
Reply tweets to AI productivity accounts (~50 replies)
Result: hard to attribute, probably 2-3 installs. Burned profile
visibility for a week.
Welcome page (homepage of foldif.com)
Result: had a layout bug for 5 days. People
landed, couldn't figure out the install path, bounced. I have no
idea how many users I lost here because I noticed the bug late.
Fixed now but waiting on the next extension release to fully ship
the corrected flow.
Indie Hackers post (May 26)
Result: 2 likes, 1 comment, 4 installs.
Direct outreach to ~10 people
Result: a few installs, but I felt gross doing it and the conversions
weren't worth the energy.
WHAT I LEARNED (THE HARD WAY)
Mass DM-ing on X is the worst ROI activity I tried. Spent
~6 hours on it total across the 100 messages. Result: ~5 installs.
That's over an hour per user, and most of them won't stick.
I confused "traffic" with "growth." The Foldif site had 149
unique active users in 30 days. Only 30 are actively using the
extension. That ~80% drop-off is my real problem, not lack of
traffic. Marketing harder while the funnel leaks doesn't help.
The homepage bug taught me to instrument my own funnel before
launching marketing pushes. I don't even have install-event
tracking properly set up between Chrome Web Store and my backend,
so I'm flying half-blind on which channels actually convert.
The "1 comment" on my first IH post was a domain shill — someone
pretending to give product feedback while suggesting I rebrand
to a domain they probably own. Watch for this pattern if you
post here. The praise-then-pitch structure is the giveaway.
WHAT I'M DOING NEXT
WHAT I'M GENUINELY ASKING
At 30 active users, 0 paying, 5 weeks in — when is this "keep
grinding" vs "the positioning is wrong and you need to pivot"?
I can't tell from the inside.
Has anyone successfully gone from "Chrome extension with no
payment yet" to actually charging? What was the trigger moment?
I have Pro features built but I'm scared to flip the switch with
this few users.
What's the one channel you'd try that I haven't?
Will be in the thread all day, no canned answers.
Hey, checked the site. Honestly, I think the idea is legit. I’ve definitely had the “losing important stuff in endless convos” problem. So yeah, the pain is real. Endless AI chats become a swamp very quickly.
That said, I wouldn’t take 5 weeks + 30 active users + 0 paying customers as “the project is doomed.” That’s more of a signal. Not necessarily “pivot everything right now”, but something in the first-time visitor story probably needs tightening before you push harder on distribution.
My honest read opening the site cold:
The product has a lot of useful stuff, but I had to work too hard to understand what category to put it in. Is it a productivity tool? A note-taking app for AI chats? A knowledge base? A folder system for ChatGPT/Claude/Gemini?
All of these are close, but they create different expectations.
Right now the homepage feels a bit like a feature buffet. Highlights, margin notes, TODOs, folders, book builder, cross-AI handoff, exports, etc.
Individually, those sound useful. Together, very quickly, my brain goes: “Okay so what’s the main thing I’m buying here?”
Which is annoying, because I think there -is- a good answer. It just needs to show up faster.
I’d start with the hero section tbh. The illustration looks nice, but as a first-time visitor it didn’t really help me understand the product. I’d rather see a clear statement before anything clever/visual.
Something like: “Foldif turns messy AI conversations into saved notes, tasks, and organized knowledge across ChatGPT, Claude, and Gemini.” Not perfect, but at least my brain gets a shelf to put it on.
I’d also answer this objection much earlier: “Why wouldn’t I just copy-paste important AI messages into Notion, Notes apps, Obsidian, whatever?” Because that’s probably what a lot of people are doing already.
So I’d show a very concrete before/after:
Before:
After:
That would make the product feel more like “oh, this fixes the exact annoying thing.”
On the “should you pivot?” question, my gut says no, not yet. I’d first try narrowing the category and the audience. “Anyone who uses AI chats” is probably too wide. Casual users won’t feel enough pain to pay. But people who use AI for actual work might:
For them, the pain is more like “I keep losing work I already paid attention to.” That’s more painful. And more paid.
Also small thing: the side popups were getting in the way for me. Maybe it’s just my screen, but they made the page harder to read, which is not ideal when the whole job of the page is to reduce confusion.
For channels, I’d probably stop cold DMs too, especially if you feel disgusted by it. I’d search for people already complaining about the problem:
And all that kinda keywords.
Social Searcher, Reddit search, X search, maybe even Google with recent filters could help. Then you’re not trying to convince random people to care, but finding people who are already annoyed.
So, I wouldn’t pivot yet.
I’d make the product easier to describe, pick one painful use case, prove it faster on the homepage, and then do distribution around people already showing the pain.
Right now it feels like there’s a real product in there, but the page is trying to introduce every feature before making me feel the main wound.
PS: Forgot to mention — if anyone's curious about the technical
side (keeping highlight anchors stable while ChatGPT and Claude
stream and re-render mid-response was genuinely the hardest
part), happy to write that up separately. Just say so in a
comment.