Sam

Building

[Sam @samshakk]

unpopular take: "just build it and see" is survivorship bias.

you hear it from the people it worked for. you don't hear from the thousands who built for 6 months and shipped to silence.

talking to 10 people first costs a week. building the wrong thing costs your year.

[Sam @samshakk]

why i'm building this:

every time i skipped talking to people first, i paid for it later. wedge is basically the step i kept skipping, turned into a tool.

gonna share the whole build here, wins, dead ends, all of it.

[Sam @samshakk]

i've shipped things nobody wanted. it's the worst feeling in indie hacking.

so i'm building Wedge: validate the idea before you build it.

now building it in public, and i want a handful of people also building in public to test it early, the kind who'll actually tell me what's broken.

here's what it looks like so far 👇

comment 'in' and i'll reach out

[Sam @samshakk]

if you'd rather just poke around: https://t.co/LKfet5cixs

still early, so expect rough edges, that's kind of the point. tell me what breaks.

[Sam @samshakk]

5/ Why this matters

When you only see the highlight reel, you think you're failing.

But that's just the edited version.

The struggle is what actually happens.

I'm building Wedge in public and I'm going to post the ugly weeks too.

Not just the wins.

If you want the real version of early-stage building, follow along.

[Sam @samshakk]

Most founders building in public in 2026 are lying to you.

Not intentionally - but they only post when things go well, when the metric looks good, when the story is clean.

Here's what actually happens when you build something from zero, unfiltered:

[Sam @samshakk]

4/ The loneliness nobody mentions

Building something from scratch is deeply isolating.

No team to blame.

No boss to ask.

Just decisions you're not sure about.

Every day you're guessing and hoping you're right.

[Sam @samshakk]

5/ The harsh part

This isn't about skills.

Most solo founders are smart enough.

They just pick the easy work over the awkward conversations.

Building without checking first isn't hustle, it's just expensive procrastination.

If you're pre-launch right now, stop and do the boring stuff.

Talk to people.

Map your competitors.

Write down your guesses and actually test them.

The ones who do this are the ones who actually ship something people pay for.

[Sam @samshakk]

Most solo founders aren't failing because they can't code.

They're failing because they never bothered to understand the market they're entering.

No competitor research.

No customer interviews.

No roadmap.

Just months of building something nobody asked for.

A thread🧵

[Sam @samshakk]

4/ What customer interviews actually reveal

Every time a founder finally talks to users, same story.

The problem they guessed?

Wrong.

The feature they spent weeks on?

Nobody cares.

The thing users actually want?

Takes 10 minutes to build.

The market will tell you everything if you just ask.

[Sam @samshakk]

most founders skip customer discovery entirely.

only 1 in 5 can clearly name the problem they're actually solving.

the feeling that your competitor's landing page makes no sense and their product has no real differentiation is completely justified.

[Sam @samshakk]

less than 1% of businesses ever talk to their first 10 customers before building.

and yet somehow that's the one thing that predicts whether the whole thing survives.

what are the odds.

[Sam @samshakk]

when pitching your startup idea, lead with what you do rather than what you are. not because it sounds more founder-y, but because it forces you to actually know if the idea makes sense.

[Sam @samshakk]

Wrote up the full process here, plus the interview questions I actually use → https://t.co/vpmFQC9Gms

Second guide on customer discovery questions specifically is up too → https://t.co/sf2iduhVhk

Building Wedge in public.

[Sam @samshakk]

Been building Wedge, an AI co-founder for early-stage founders.

Before writing a line of product code, I wrote down the actual process I use to validate an idea. Turned it into a guide instead of keeping it in my notes.

🧵

[Sam @samshakk]

The single best question I've found:

"Walk me through the last time this problem came up for you."

Not "would you use X", that's asking someone to predict the future. People are bad at that. They're good at describing what already happened.