Many online shops that have not been successful did not fail because of their advertising or their website. They failed simply because there was no demand for their product and the founder didn’t realize that until after he’d already paid the money. Based on a CB Insights report, 42% of startups fail since they have no market need. Meaning more businesses failed because nobody wanted their product.
Validation allows you to detect that problem early, before you’ve invested a cent into product, style, or advertising that end up getting nowhere. The measures below are not about market research reviews or spreadsheets packed with assumptions. They’re about getting real proof – quick, low-cost, and often unpleasant – that informs you to either create the product or go out.
Step 1: Write down the exact problem and who has it
Before you start using a keyword tool or a store builder, condense your idea into something you can test: this specific group of people has this specific problem, and it’s painful enough that they’d pay to solve it.
Bad answers here ruin everything after. “People who like fitness” is not a customer segment. “Runners training for their first marathon who keep getting blisters from generic socks” is. The second version gives you something you can actually go test. If you can’t name who has the problem and why it’s bad enough to spend money on, there’s nothing to validate yet – you’re still just guessing.
Write this hypothesis down somewhere you’ll see it again. Every step that follows should either prop it up or knock it down.
Step 2: Check search data before you check your enthusiasm
Your passion for a problem doesn’t count. What counts is if people are searching for a solution.
Go on Google Trends. Type in some terms related to the problem you’re hoping to solve. What you want to see is one steady line. One upticking, steady, gradually increasing line over months or years means sustainable demand. One big spike and then it goes down means you found a fad. Near-zero demand in Google probably means nobody else was all that interested in your obscure fascination. A fad is how people wind up with garages full of unsold inventory.
Pair this with basic keyword research. Look at search volume for the problem itself and for solution-oriented terms. If almost nobody is searching for anything related to your idea, that’s not proof the idea is bad, but it is a signal that you’ll be creating demand from scratch instead of meeting existing demand – a much harder and more expensive game. Near-zero search volume should make you cautious, not more determined.
Step 3: Let competitor reviews do your research for you
You can do the work of a paid research firm by browsing Amazon reviews and those of competitors for free.
Locate products in the marketplace that solve part of your problem. Instead of good reviews, which won’t be helpful in this case, scroll to the reviews where people are complaining – the ones that describe what didn’t work for them, what was missing, what broke, what seemed overpriced, or what they wanted instead.
Complaints are a key part of validation. If two dozen people say a product is too small, arrives in thin packaging, or is missing a feature, you now know an opportunity: Build a better version of something people are already buying in the market. This step also helps confirm demand while helping you come up with a specific product.
Step 4: Build a one-page test store, not a full one
This is the part most first-time founders overspend. They start building a full store – collections, multiple products, a custom theme, blog pages – before they know if a single soul wants what they’re selling.
Skip that. Build one page. One product image (even a mockup), one price, one clear offer, and either a “pre-order now” or “add to cart” button. That’s the entire test. You’re not launching a business yet – you’re launching a question, and the page needs to look real enough that a visitor treats it like a real buying decision.
You don’t need a fully built out Shopify store to run this test, and you shouldn’t spend the money making one look polished at this stage. If you’re eventually ready to set up the real thing, it’s worth looking into how to start Shopify for less so you’re not overpaying for infrastructure you don’t need yet either. Keep this early version lean. A single product page with a working buy button will tell you more in a week than a beautifully designed ten-product store would tell you in a month, because the page isn’t the point – the response to it is.
Step 5: Decide what “pass” looks like before you see any results
Many individuals unknowingly destroy their own test results by first conducting the experiment and then establishing the criteria for success based on the outcomes. This is not validation. It’s an elaborate form of self-justification.
Determine your benchmarks before you go live. Perhaps it’s a 3% email sign-up rate, or a 2% conversion rate for the buy button. Set these percentages beforehand. If your plan is a good one, and the actual numbers fall short, and you catch yourself rationalizing why the lower numbers are acceptable, take heed. That’s the signal of someone desperately trying to persuade themselves their concept is a success despite the proof to the contrary.
Writing the pass/fail line in advance removes the temptation to move the goalposts once you’re emotionally invested.
Step 6: Send a small, targeted stream of traffic to the page
Now you just need some visitors – not many, just the right kind.
A small budget on Meta or Google Ads, targeted specifically at your identified customer segment, and you can buy somewhere between 100 and 200 targeted visitors for not a lot of money. That’s enough data to read real behavior. It’s also far more valuable than a hundred friendly conversations with people who already like you and want to be supportive.
Here, targeting matters more than volume. Sending 500 random visitors tells you less than sending 100 people who actually match your problem hypothesis. Keep the ad copy honest and specific – describe the actual pain point, not a vague benefit – so the traffic you attract is already primed to want what you’re offering.
Step 7: Trust money over compliments
This is the step people resist the most because it’s the least comfortable. Customer development interviews – the classic “would you buy this?” conversations – are useful for shaping an idea, but they are not proof of anything. People are polite. They say things sound “cool” or “interesting” to be nice, especially when they know you. None of that predicts whether they’ll actually pay.
What predicts it is action. A pre-order. A completed checkout. An email address volunteered by a total stranger who has no social obligation to support you. Waitlist signups from people who found your page through an ad – not through a group chat – are a far stronger signal than a dozen enthusiastic verbal endorsements.
If your test page generates real pre-order revenue or a strong list of cold signups, you’ve got something. If it generates a lot of “I’d totally buy that!” texts from friends and zero strangers taking action, you don’t have validation – you have a support network.
Step 8: Run the unit economics before you celebrate
Achieving conversions on the test page may seem like a victory, but that’s just the emotional win. Next, you need to stay in the black once you have to pay for all those customers.
Calculate what that observed conversion rate translates to in terms of orders, then multiply by your anticipated order value. Next, subtract your total product cost, total ad spend, and total fulfillment cost. If you don’t have at least some profit left after those three cost lines, or some unusually favorable costs on your side, your idea doesn’t work as a business. It’s just a fun test.
This step catches a specific trap: mistaking attention for a business model. A high conversion rate on a low-margin, expensive-to-ship product can still be a losing proposition. Do the math before you scale anything.
Bringing it together
Validation is not a single test – it’s a series of low-cost sieves, each meant to catch a bad idea as early as possible. A well-defined problem thesis keeps you from testing something too abstract to work. Search and review data keeps you from playing with shiny objects or a market no one is interested in. The one-page test, the budgeted thresholds, the limited ad buy, and the insistence on actual money rather than enthusiasm – these are all meant to push you to the decision to kill fast or proceed while there’s still muscle on the bone.
It was never the aim to be right. It was to determine, as inexpensively and as fast as possible, if you are. Once real people you have no previous relationship with are slapping down credit cards – not compliments, credit cards – you are in the clear to begin building the real store, and you’ll go into that build with real evidence to back you instead of a feeling.
