You've been at this three months. You check the dashboard most days out of habit more than hope. There's been a sale or two, maybe none. You still don't know if you're one small change away from this working, or if you've been pouring evenings into something that was never going to go anywhere. Everyone around you says "give it six months" or "give it a year," but nobody tells you what you're supposed to be watching during that time to know if it's working.
Here's a way to actually answer the question instead of just waiting and hoping.
First, define "nothing" precisely
"It's making nothing" can mean three very different things, and they call for different responses.
It can mean literally zero dollars have come in. It can mean some money has come in, but less than what you spent on materials, ads, subscriptions, or tools. Or it can mean you made a small profit, but it took so many hours that the effective hourly rate is embarrassing. Write down the actual number. Total revenue, total costs, total hours. Most people who say "it's making nothing" haven't actually done this and are going on a feeling. The feeling is often wrong in either direction.
The advice to "give it six months" is usually a stall tactic
This gets repeated constantly and it's not useless, but it's also not a plan. Six months of what, exactly? If you spend six more months doing the same thing that produced zero in the first three, you'll likely have the same result, because time alone doesn't fix a broken offer or a nonexistent audience. Patience only pays off if something underneath is actually changing between attempts. If you're going to keep going, you need a reason to believe attempt four will behave differently than attempts one through three, not just more of the same thing repeated for longer.
Count your actual at bats, not your activity
Activity is posting, tweaking, researching, redesigning. An at bat is a real, specific moment where an actual stranger who could plausibly pay you was shown the actual offer, the price, and the ask to buy. Scrolling past your shop isn't an at bat. Someone reading your listing and leaving without adding to cart is closer, but still soft. Someone asking "how much for two" or putting it in their cart and then not checking out is a real at bat.
Go count these. If you can still count them on your fingers after three months, you don't have a failed idea, you have too small a sample to know anything yet. That's not a reason to quit, it's a reason to get more data before deciding anything.
If you have real at bats, split them into two piles
This is the part that actually tells you fix or drop.
Pile one: people who showed interest and then didn't buy. Questions asked, items added to cart and abandoned, replies to a pitch that went quiet after you named a price. This is a conversion problem. Conversion problems are usually fixable: price too high or confusingly presented, checkout has too many steps, no trust signal like reviews or samples, the offer itself is fuzzy about what they actually get. These are concrete, checkable, fixable in days, not months.
Pile two: total silence. No questions, no cart adds, no replies, even though real people saw the actual offer. This is a demand or targeting problem, and it's the harder one. It usually means the people seeing it aren't the people who want this, or the thing itself isn't something enough people want at this price in this form. You can't fix silence with a better checkout button. Fixing it means changing who you're showing it to or what you're offering, which is close enough to starting over that you should treat it as a real decision point, not a tweak.
If your three months has mostly produced pile two, more consistency won't rescue it. That's the uncomfortable part, but it's true.
Watch for the moving target problem
A lot of "three months, nothing" stories aren't actually three months of one test. They're three different ideas, or three different niches, or three redesigned offers, each run for two or three weeks before getting scrapped for the next idea. If that's what happened, you haven't proven the concept doesn't work. You've proven that you haven't let anything run long enough to know. That's fixable too, just not with a new idea. It's fixable by picking the current version, holding it still, and giving it four clean weeks with nothing changed except getting it in front of more of the right people.
The honest tell
Sunk cost is real and it lies to you both ways. Three months of nights and weekends makes quitting feel like admitting failure, so people keep going past the point of evidence. But burnout also makes people quit right before something was about to turn, because they're tired, not because the numbers say so.
The numbers are the tiebreaker for a reason. If you have real interest that isn't converting, that's a fix, and usually a fast one. If you have real exposure and real silence, that's a decision about whether to build something else, and staying busy on this version won't change that answer. Either way, you get to know instead of guess, and knowing is worth more than another month of checking the dashboard out of habit.