Strategy & Growth
How to Validate a Business Idea Using Technology Before You Spend on Building It
Validating a business idea means finding out whether people will pay, before you spend on building. The tests that matter all end with someone giving up something real: money, an email address, a deposit, or a booked slot. Everything else is encouragement. Run the cheap tests first, in order, and stop as soon as one fails badly enough. Most ideas that go nowhere did not fail at execution. They failed because not enough people wanted the thing, and that was knowable in two weeks for very little money.
An animated business dashboard for a sample hardware store, switching between today, this week and this month: revenue and order totals, a revenue trend line with its best day marked, and a bar showing how sales are split between three branches and the online store.
See how every branch and your online store are selling — today, not next month.
The only signal that counts
Friends saying it is a great idea is not validation. Neither is a poll, a like, or a comment saying 'interested po'. Those cost the person nothing, so they tell you nothing.
Validation is when someone gives up something they value. Ranked by strength: a payment, a deposit, a booked appointment, a phone number, an email address. Design every test so it ends in one of those.
Test 1: Is anyone already searching for this?
Start here because it costs an hour. Use keyword tools to see whether people search for the problem you solve, in the words they actually use. Look at the results that already rank. Weak, thin, or foreign-focused results against real search demand is a gap. Strong results from established Philippine competitors means you need a sharper angle, not necessarily that you should stop.
Zero search volume is not automatically fatal for a genuinely new idea, but it does mean nobody is looking for you yet, and your acquisition cost will be higher than you planned.
Test 2: What are people already complaining about?
Read the one and two-star reviews of your closest competitors on Google, Shopee, and Lazada. Read the Facebook groups where your customers complain to each other. You are looking for repeated, specific frustrations, not one-off bad days.
A complaint that appears in twenty reviews is a product brief. It tells you what to build and what to say in your marketing, in the customer's own words.
Test 3: The landing page test
Build a single page that describes the offer honestly, prices it, and asks for one specific action: join the waitlist, book a slot, or reserve with a deposit. Send a small amount of targeted traffic to it, ideally Facebook or Google ads aimed narrowly at the people you described.
Decide your pass mark before you look at the results. Write it down. Otherwise you will rationalise whatever number you get, and the whole exercise becomes expensive theatre.
- State a real price. A page with no price validates curiosity, not demand.
- Run two versions of the headline against each other rather than guessing which framing lands.
- Read what people type in the message box. The wording is worth more than the count.
- Reply to every single sign-up personally. Those conversations are the most useful data you will get all month.
Test 4: Take money before the thing exists
The strongest test available to a Philippine business is a pre-order or a reservation fee through GCash or Maya. It is easy to set up, familiar to buyers, and it converts an opinion into a commitment.
Be honest about the timeline and refund anyone who changes their mind without argument. You are buying information, not revenue, and a clean refund policy costs far less than a bad reputation before you have even launched.
Nothing tells you the truth about demand faster than asking someone to pay for it.
Test 5: Do it manually before you automate it
Deliver the service by hand for the first ten customers. Take bookings over Messenger. Track them in a spreadsheet. Do the work yourself. It does not scale, and that is the point: you are learning what the real process is before you pay to encode it in software.
Almost every business that builds a system before doing the work manually builds the wrong system. The details you discover in those first ten jobs are the ones that decide the design.
Once the tests come back positive, a proper landing page or first version can be live in days. See what a build includes.
See web design and developmentTest 6: The smallest real version
Only now do you build, and only the part that solves the core problem. One flow, done properly, that a customer can complete end to end. Not a stripped-down version of everything, which is just a bad product. One thing that works.
Get it into real use quickly and watch what people do rather than what they say. Where do they stop? What do they message you to ask? Which feature that you were sure about does nobody touch? Those answers set your next build.
What to ignore
- Enthusiasm from people who will never be customers, including most of your family.
- Feature requests from someone who has not paid. Weight feedback by commitment.
- Competitor count as a reason to stop. Competitors usually prove a market exists; a total absence of them is the worrying signal.
- Your own certainty. It is the most expensive thing in this list.
A two-week validation sprint
Days one and two: search demand and competitor complaints. Days three to five: build the landing page with a real price and a real ask. Days six to ten: run a small, narrow ad budget and reply personally to everyone who responds. Days eleven to fourteen: deliver manually for the first few customers who said yes, and write down every step you actually performed.
At the end of that you will know whether to build, adjust the offer, or stop. All three outcomes are wins. The expensive outcome is the one where you skipped this and found out a year later.
Frequently asked
Enough to get a clear signal and no more. A landing page plus a small, narrowly targeted ad budget over two weeks is usually sufficient. Set the pass mark before you spend, and stop the moment the answer is clear in either direction rather than paying for confirmation of what you already know.
That is usually good news. Competitors prove people pay for the category and save you from having to educate the market. Read their one and two-star reviews for the specific, repeated complaints, and build your angle around those. A category with no competitors at all is the more worrying finding.
It is a signal, not proof. Email addresses cost the person almost nothing. A deposit, a pre-order, or a booked slot is far stronger. If you only have a waitlist, go back and ask a portion of them to pay something small; the drop-off between those two numbers tells you what you need to know.
Build the smallest version that solves the core problem completely for one type of customer. One flow that works end to end beats ten half-finished features. Deliver manually first so you learn what the real process is, then encode only what you have actually confirmed people use.
It never fully is, but you can stop the formal sprint when strangers, not friends, are paying at a price that leaves a workable margin, and you can describe the customer and their problem in one sentence without hedging. After launch, validation continues as a habit: every significant new feature gets tested the same way.