How to Know If Your Startup Idea Is Actually Working
Is Your Startup Actually Working, or Are You Just Very Busy?#
Something is off and you can't name it.
Not in a dramatic way. In the quiet way. The kind that shows up when you're not doing anything and you actually have a second to think.
The external story looks fine. People are responding. Conversations are happening. Someone introduced you to a potential customer last week.
But underneath that, there's a question you keep not asking out loud: is any of this real?
Startup Traction vs. Activity: Why the Difference Is Hard to See#
Here's what makes this confusing.
Activity produces the same feeling as forward motion. Both feel like momentum. Both give you something to report. Both let you go to sleep feeling like you did something today.
The difference only becomes visible when you stop and ask a specific question.
That question is not "is this working?" That's too big. You can't answer it. It lets you argue with yourself forever.
The real question is smaller: has any user done something you didn't cause?
Not a meeting you set up. Not a response to an email you wrote. Not interest that appeared after you spent an hour on a call explaining why the problem matters.
Something they did on their own. Something that only makes sense if the idea is solving a real problem for them.
Alex Hormozi shared this pattern in a conversation with a stuck founder. His direct advice: change directions less. Stick with things long enough to see if they actually worked. The implication is the same one that surfaces repeatedly among early-stage founders at YC Startup School and across Reddit threads where founders are honest about what they're experiencing. Most of what looks like signal is just the echo of your own effort coming back to you.
The One Early Stage Founder Signal That Actually Counts#
Write down one specific behavior.
One thing you have observed a real user do, without prompting, that would only make sense if your idea was genuinely solving something for them.
It doesn't have to be big. It could be that someone shared it with a colleague before you asked them to. It could be that someone came back a second time without a nudge. It could be that someone complained when it wasn't available, unprompted.
The size of the behavior doesn't matter. What matters is that you didn't engineer it.
If you can write it down in one sentence, you have signal.
If you sit down to write it and what comes out is something you set up, you created, or you caused through your own persuasion: that's not signal. That's you working hard.
Both things are real. They just tell you different things.
How Long to Stick With an Idea Before You Change Direction#
This is where founders make the most expensive mistake.
When nothing is clearly working, the instinct is to change something. The pitch, the target customer, the feature set, the positioning. Change feels productive. It feels like responding to reality.
But if you change direction before you know whether you've actually found unprompted behavior or not, you are not responding to reality. You are responding to your own discomfort.
The honest version of this: most pivots at the three-to-six month mark are not strategic. They're emotional. The founder is tired of not knowing and changes something because that at least produces a new feeling of movement.
Hormozi's observation to a stuck founder was specific. The advice was not "try a different customer segment" or "rethink the product." It was: stop changing directions. Stick with things long enough to see if they actually worked.
That advice only lands if you have a way to know whether something worked. That's why the one-sentence behavior test exists.
What to Do With What You're Seeing Right Now#
If you can write down one unprompted user behavior in a single sentence: stay the course. You have enough signal to keep going in this direction without second-guessing the whole thing. Your job now is to find more of that behavior, not to reinvent the idea.
If you cannot write it down, or if what you write is something you caused: do not change direction yet. Not because the idea is definitely working, but because you don't have enough information to know. Changing now throws away whatever learning is still sitting in what you've already built.
Spend the next two to four weeks looking specifically for that one unprompted behavior. Watch for it. Design for it. Talk to users not to sell but to observe.
If you find it, then you know. If you don't find it after genuine looking, that's information too. If you're still unable to find signal, The Pivot Question: How to Tell If Your Idea Is Dead or Your Execution Is Broken will help you distinguish between a dead idea and flawed execution.
The Signals Everyone Mistakes for Real Signal#
Meetings: measure your ability to get on a calendar.
Responses: measure your ability to write a compelling message.
VC interest: measures your ability to tell a story.
Team growth: measures your ability to recruit.
None of these are evidence that the idea is working. They are evidence that you are working. That distinction matters enormously when you are about to decide whether to push harder or change direction.
A VC call is not traction. A waitlist is not traction. A glowing intro from a well-connected person is not traction. These things feel like progress because they are hard to get and they produce external validation.
But they measure you. Not the idea.
The One Thing to Carry Away From This#
You cannot think your way out of this confusion.
You cannot convince yourself, and you cannot logic your way to certainty. The only move is to write down what that one unprompted user behavior would look like, go find it, and look honestly at what you see.
Until you can write it in one sentence: don't change direction.
Until you find it: don't celebrate.
Write it down first.
If you want help separating what's real from what's noise before you spend another six months building in the wrong direction, Acrein Lab is built for exactly this moment.