I almost shipped the "smart" version. Users actually wanted the simple one.

Solo founder, Building Freenity for 6 months: bookkeeping on
autopilot for solo founders and small startups.

In 1 line: Upload your banque statemement or Connect your bank, forward your invoices, snap your receipts, and we hand your CPA a clean monthly report. No spreadsheet, no Sunday cleanup, no $300 surprise in March.

Here's what I almost got wrong.

Original plan for the capture flow: connect Gmail, AI scans your inbox,
invoices auto-detected, zero effort. The "magic" version.

OAuth working. UI clean. About to launch. I sent 3 weeks working on the automation.

Then I forced myself to do something I'd been avoiding: talk to people
before shipping. CPAs, indie hackers running SaaS, mentors at
Bootstrappers Breakfast (free online sessions you can find through
this community, genuinely useful advice).

Two things came back, and I didn't expect either.

1. Users didn't want the "magic" version.

When I described the Gmail-scanning flow to actual solo founders, the
reaction wasn't "wow." It was "you're going to read my entire inbox?"

When I asked: "would you prefer we scan everything, or you forward
what matters?" every single one picked forwarding. They wanted
control, not automation. Privacy, not magic.

The lesson hit hard: I had been building what was technically
impressive to me, not what was emotionally acceptable to them.

2. The professionals saw a wall I'd missed.

Gmail "restricted" scopes need a CASA Tier 2 security audit.
$15K-$75K. 2-4 months. Renewable yearly. Pre-revenue, it would
have killed the project.

I'd spent weeks building a feature I couldn't even legally ship
without a budget I didn't have. And I only learned about it because
two consultants and a CPA pointed it out.

So I dropped the Gmail integration and rebuilt the capture flow in
two days:

📧 Forward invoices to one address (Gmail, Outlook, Proton, anything)
📷 Snap a photo of paper receipts (OCR + auto-categorization)
📤 Upload from desktop or mobile
🔄 Everything auto-matches to your bank transactions
📊 Monthly export ready for your CPA, no cleanup needed

Less "magic," more control. Wider market (works with any email
provider). Simpler architecture. Zero Google dependency.

And the pitch is stronger: "we never read your inbox, you forward
what matters" beats "we scan your Gmail" every single time in a
real conversation.

The lesson I keep coming back to:

It's tempting to build the most automated, AI-powered version of
anything in 2026. That's what's fun to code. That's what looks
impressive on a landing page.

But customers don't pay for what's impressive. They pay for what
solves their problem in a way they're emotionally ready to accept.
Sometimes that means less automation, more transparency, more
control. Even if it costs you a few clicks of friction.

You can't figure that out from your laptop. You have to talk to
real users, real CPAs, real consultants. The cost of those
conversations is a few hours. The cost of skipping them is months
of work on a feature nobody wants.

Two asks:
-> Has talking to users ever changed your mind about a feature you
were sure was the right move? Genuinely interested in those stories.

-> Hunting my first 5 paying customers. Are you tired of chasing receipts,
fixing spreadsheets, and never really knowing your cash flow?
Founding deal is $29/mo, direct Slack with me, full roadmap input.

DM or comment.

Happy to roast or be roasted. www.getfreenity.com