Test it cheaply
Before you spend real money, find the smallest test that shows whether people will act, ideally by paying. Learn to set a pass mark and run an honest pre-sale.
VALIDATION
Evidence is what people do
Validation means collecting evidence that real customers will act, not just agree. The more something costs the person, the more it tells you. A compliment costs nothing. Giving you their phone number costs a little. Paying costs the most, so a payment is the strongest evidence you can get.
Fifty people saying 'I'd buy that' can turn into zero sales. Five people handing over a deposit almost always turns into five customers.
The ladder of evidence, weakest to strongest
- Compliments and likes
Nice to hear, costs them nothing, tells you almost nothing.
- Contact details
A phone number or a sign-up for updates. Some interest, easy to give.
- Time
They book a call, come to a tasting, try a free session. They gave up something.
- Money
A pre-order, a deposit, a paid first order. The closest thing to proof before you launch.
Check yourself
Mahsa is testing a home-organising service. After two weeks, which result gives her the strongest evidence?
- Her Instagram post about the idea got 300 likes and dozens of heart emojis
- 40 people filled in her form saying they'd 'probably' book
- 6 people paid a deposit for a first session next month
- A friend who runs a shop said it will definitely work
Show the answer
6 people paid a deposit for a first session next month
Yes. Six deposits are six people who acted with their own money. That beats hundreds of likes or 'probably'.
SMALLEST TEST
Do the smallest thing that could fail
A minimum viable product (a term popularised by Eric Ries's book The Lean Startup) is the smallest version that lets real customers use it and pay. Often it's not a product at all: you do the work by hand, sell a small batch, or post a one-page offer and take orders.
Want to launch a meal-prep app? First, take orders by message from ten colleagues and cook on Sunday. If they don't reorder, an app won't save it.
Mona's test: a paid weekend baking class
Room hire 150
Price per seat 30
Seats to cover the room 150 / 30 = 5
Pass mark, set BEFORE asking:
8 paid seats from 40 people invited
8 / 40 = 20%
Result: 9 paid from 40 = 22.5%Passed: 9 ≥ 8. She runs the class, then tests a second date.She decided what 'success' meant before she asked anyone, so she couldn't talk herself into a weak result.
Check yourself
Cheap test, or expensive bet?
- Take pre-orders for 20 T-shirts before printing any
- Sign a one-year lease on a shop to see if people come
- Offer three paid tutoring sessions to parents in one school group
- Order 1,000 units to get the lowest price per piece
- Sell a small batch at a weekend market
- Pay a developer to build the full app first
Show the answer
Cheap test: Take pre-orders for 20 T-shirts before printing any, Offer three paid tutoring sessions to parents in one school group, Sell a small batch at a weekend market
Expensive bet: Sign a one-year lease on a shop to see if people come, Order 1,000 units to get the lowest price per piece, Pay a developer to build the full app first
Run an honest pre-sale
- Write a clear offer
What they get, when they get it, the price, and what happens if it doesn't go ahead.
- Set your pass mark first
A number of paid orders from a number of people asked. Write it down before you start.
- Ask real prospects for real money
A full pre-payment or a deposit. Check your local rules on taking payments in advance.
- Deliver or refund, promptly
If it passes, deliver on the date you promised. If it fails, refund everyone in full. Your reputation is worth more than the test.
Check yourself
Parisa set a pass mark of 10 pre-orders from 80 people. She got 3 pre-orders and 25 'maybe later' replies. She should treat the offer, as she tested it, as having failed.
Show the answer
True
True. 3 is well below 10, and 'maybe later' costs nothing to say. That doesn't mean the whole idea is dead. It means this customer, price or offer didn't work, and she should change one of them and run another small test.
Check yourself
Babak dreams of opening a bike repair shop, which would mean rent, tools and a long lease. What is the best first test?
- Sign the lease now, before a good location near the park is taken by someone else
- Do paid repairs at customers' homes on weekends with a basic tool kit
- Run an online poll asking if people would use a new bike shop
Show the answer
Do paid repairs at customers' homes on weekends with a basic tool kit
Right. Weekend repairs cost him little, bring real paying customers, and show him demand, prices and the most common jobs before he commits to rent.
Lesson recap
- Validation is evidence people will act; the more it costs them, the stronger it is.
- Payment beats time, time beats contact details, and all of them beat compliments.
- Run the smallest test that could fail: by hand, a small batch, or a one-page offer.
- Set a pass mark first, pre-sell honestly, and refund in full if it doesn't go ahead.