How to test a product before you make it

You've got a quote from a supplier. Minimum order 100 units, ₱62,000 up front, four weeks. Eleven people have told you it's a great idea and two of them said "I'd definitely buy that."

None of that is evidence. The only test that counts is somebody handing over money before the thing exists.

There are three ways to run that test, in ascending order of what they cost you. Have the price conversation with ten real buyers, which takes a day and costs nothing. Take pre-orders with a real date and a real refund promise, which takes two weeks. Or make the smallest batch anyone will sell you, which costs actual money and is what you do once the first two have passed.

Below is how to run each, what number counts as a pass, and how to do a pre-order without lying to anybody.

"Would you buy this?" produces lies, and not on purpose

Ask someone whether they'd buy a thing and you're not asking about the product. You're asking whether they want to be encouraging, and almost everyone does.

It's worse if they know you. It's worse again if you're visibly excited, because now saying no has a social cost and they can feel it. The polite answer costs them nothing and it costs you ₱62,000.

Likes, poll votes and comments have the same problem. Tapping "yes" on a story sticker takes half a second and commits nobody to anything. A hundred votes on a poll is a hundred people who were scrolling.

The gap between intention and payment is enormous and it's not a character flaw. People genuinely mean it when they say they'd buy something, and then a bill arrives, or they see something else, or the moment passes. Nine people saying yes and two paying isn't a bad conversion, it's the normal one, and it's exactly why the test has to involve money.

Test one: the price conversation

Free, and you can do it today.

Find ten people who look like your buyer, not your friends. Message them individually and don't ask whether they like the idea. Ask what they'd pay for it, and then ask what they currently do instead.

Hi, I'm thinking of making a canvas laptop sleeve, padded, fits a 14-inch, made here in Manila rather than imported. Before I do, two questions: what do you use now, and what would you expect to pay for something like that?

What you're listening for:

• A number they say without hesitating. Hesitation means they've never thought about buying this category, which means you're creating demand rather than serving it. That's a much harder business.
• What they use now. If the honest answer is "nothing, I just put it in the bag", you're not competing with a product, you're competing with a habit, and habits are harder to beat than competitors.
• Whether their number is above your cost. If ten people say ₱600 and your unit cost is ₱480, there's no business there and you've found out for the price of ten messages.

Ten conversations takes an hour and a half. It will not tell you whether people will buy. It will tell you whether the price you had in mind is in the same universe as theirs, and it kills a surprising number of ideas before they cost anything. If you haven't costed it properly yet, do that first, because you can't read the answers without it, and how to price what you make is where that starts.

Test two: the pre-order

This is the real test. Take money for something that doesn't exist yet, with a date attached and a refund you'll honour without argument.

The rules matter more than the mechanics.

Say plainly that it doesn't exist yet. In the post, in the listing, in the message. Not in small print. "These aren't made yet. I'm taking orders until the 20th, and if I get to 30 I'll place the production order. They'd be with you by the 18th of next month."

Give a real date and pad it. Whatever your supplier says, add two weeks. Every first production run is late, and a pre-order that arrives late is how you turn early supporters into people who warn others about you.

Promise a refund, in writing, with no conditions. "If it doesn't happen, or if you change your mind before I place the order, I'll send it straight back." Then honour it in a day, not a week.

Don't spend the money until you've committed to the run. This is the discipline that separates a pre-order from a mess. The money sits there, untouched, until you either place the order or refund everybody.

Ask for the money in the same message where you say it doesn't exist. If you split those across two messages, the second one never gets sent, and you'll find yourself with fifteen people who've verbally agreed and nobody who's paid.

A smaller version of this is a deposit rather than the full amount, which lowers the barrier and still filters out the people who weren't serious. Taking a deposit before you start covers the wording and the amounts.

If you already have an audience, running the pre-order as a proper launch with a date and a list will roughly double what you'd get from posting it once. Building a waiting list for a drop is that mechanism.

Decide what a pass looks like before you start

Write the number down first, in a note, with a date. Otherwise you'll rationalise whatever you get.

The number should come from arithmetic, not from hope:

Thing	Example
Minimum order from supplier	100 units
Cost of that order	₱62,000
Your selling price	₱1,200
Units needed to cover the run	52
Pre-orders that make it safe	35
Pre-orders that make it a no	Under 15

Then set a window. Two weeks is right for most things. Long enough for a second and third reminder, short enough to be a real deadline.

And decide in advance what you do in the middle. If you set 35 and you get 22, that's the interesting case, and deciding it in advance stops you from talking yourself into a run you can't afford. A reasonable middle rule: negotiate a smaller minimum with the supplier, even at a worse unit price, or extend by one week and message everyone who enquired but didn't order.

The photo problem, and how to solve it honestly

You need a picture of a thing that doesn't exist. There are honest ways to do this and dishonest ones.

Honest: a sample you paid for, a hand-drawn sketch with the measurements on it, a photo of the fabric and hardware laid out, a photo of a similar product with a clear caption saying it's a reference and yours will differ, a 3D render clearly labelled as one.

Not honest: a supplier's stock photo presented as yours, another brand's product with the logo removed, an image that implies you have stock.

Get one sample if you possibly can. Most suppliers will do a single unit for a premium, and a real photo of a real object is worth more than any amount of explanation. It also catches problems, because the sample is where you find out the zip is wrong.

While testing, do not invent scarcity. No "only 5 left" on a thing that doesn't exist. People remember, and the whole value of a pre-order is that the people who buy trust you enough to pay in advance. Spend that carefully.

Marco's laptop sleeves, in Manila

Marco wanted to make padded canvas laptop sleeves in Manila. Supplier minimum 100 units at ₱620 each, ₱62,000 up front, four weeks production. He planned to sell at ₱1,200, with ₱150 delivery in Metro Manila and cash on delivery available.

He'd posted the idea and got 140 poll votes and 31 comments saying they wanted one. On that basis he nearly placed the order.

Instead he ran the price conversation with 12 people. Nine gave a number without hesitating and the median was ₱900, not ₱1,200. Three said they'd never buy a sleeve at all. That was the first useful thing he learned, and it cost him one evening.

He adjusted to ₱1,050 and ran a two-week pre-order, with 50% up front, ₱525, refundable. He said clearly it wasn't made yet and gave a delivery date six weeks out.

Signal	Number
Poll votes saying yes	140
Comments saying they wanted one	31
People who opened the pre-order link	88
Paid a ₱525 deposit	24
Cancelled and refunded before production	2
Held at production	22

Twenty-two deposits from 140 votes. He'd set his pass line at 20 and his no line at 10, so it was a pass, and a narrow one.

Two things he'd do differently. The 12 price conversations should have been 25, because nine data points on price is not much to move ₱150 on. And he ran the pre-order without offering cash on delivery, which is the default expectation in Manila, so an unknown number of the 88 who opened the link were never going to pay a stranger in advance for something that didn't exist. He has no idea how many. That's the cost of not asking.

The refunds were the part that surprised him. Two people changed their minds, which he'd expected, but sending the money back took him nearly an hour across two evenings because both were bank transfers and he had to chase account details. On manual rails a refund is a job you do by hand, one transfer at a time, and refunding on a manual payment rail is worth reading before you take twenty deposits rather than after.

Digital products and services are cheaper to test

If what you're testing is a file, a course, a template or your own hours, the test is even simpler and there's no excuse for skipping it.

Sell it before you build it. Write the sales page, name the price, open it for a week, and take payment with a delivery date two or three weeks out. If eight people pay, you build it and you're paid for the work. If one person pays, you refund them with an apology and you've saved yourself six weekends.

For a service, the version is even smaller: offer four slots at a real price next month. Slots that fill are demand. A waiting list with no money attached is not.

What a fail actually tells you

If the pre-order flops, you've learned one of four things, and they're not the same:

• Nobody wants it. The clearest and most useful result.
• They want it, not at that price. Test the same thing ₱300 lower before you abandon it.
• They want it, and they don't trust you enough to pay in advance. The product is fine and the problem is credibility. That's fixable with reviews, a physical sample and time.
• Nobody saw it. Check how many people actually opened the link before you conclude anything. Twelve views is not a test.

Only the first one means stop. The other three mean adjust and rerun, and rerunning costs another fortnight rather than ₱62,000.

The limitation worth knowing before you take deposits

Sailo will carry the listing, the options and the order, and on the manual rails, bank transfer, cash on delivery, or an order that arrives through WhatsApp, it never touches the money. That's the right shape for a pre-order, because the deposits sit in your own account where you can send them straight back.

It also means the work is yours. Nothing tells you a transfer arrived and nothing sends a refund. If your test fails with 22 deposits in hand, that's 22 manual refunds and about an hour of your evening. Card payments, where refunds are a button, need a Stripe account that Stripe has cleared, which takes days rather than minutes and is not something to start the week of your launch.

Have three conversations tonight

Not a poll. Three individual messages to three people who look like your buyer, asking what they use now and what they'd pay.

If the numbers come back near your price, write down your pass line and your no line, then open the pre-order for two weeks. If they come back low, you've saved the production run, and that's the test working exactly as intended.

And if you've never sold anything at all yet, run the pre-order to people who already know you rather than to strangers, because that's where the first orders come from anyway and getting your first ten orders is the list to work from.