What Is a Test Card? Definition + How to Use It

A test card is a one-page tool for designing an experiment before you run it. Popularized by Strategyzer in Testing Business Ideas, it forces you to write down four things — a hypothesis, the test, the metric, and the success criteria — so your experiment is deliberate and falsifiable, not a hunch you defend afterward.

Quick Answer: A test card is a one-page template from Strategyzer's Testing Business Ideas (David Bland and Alexander Osterwalder) for planning an experiment up front. It captures four fields: the hypothesis you believe, the test you will run, the metric you will measure, and the criteria that decide whether you were right. By committing to a pass threshold before you start, it stops you from moving the goalposts once results arrive. It pairs with the learning card, which you fill out afterward.

The value of a test card is discipline. Many teams "test" an idea by building something, watching what happens, and then telling a story about the numbers that flatters the decision they already wanted. A test card blocks that move by making you commit — in writing, before any data exists — to what you believe, how you will check it, and what result would prove you wrong.

It is the planning half of a validation experiment. If you are new to the discipline, start with what a validation experiment actually is; the test card is the tool that structures one before you run it.

The Four Fields of a Test Card

A test card has four fields, and each begins with a fixed sentence stem that makes the field hard to fudge. Filling them in order turns a vague ambition ("let's see if people want this") into a specific, testable plan.

The table below defines each field and its prompt. It is a structural reference, not a scorecard.

Takeaway: The four fields march from belief to proof. The hypothesis names what you assume, the test says how you will check it, the metric says what you will count, and the criteria set the bar in advance — so the outcome is decided by the data, not by hindsight.

The hypothesis must be falsifiable. "Users will love the onboarding" cannot be proven wrong, so it is not a hypothesis. "We believe that at least a meaningful share of trial users will complete setup unaided" can fail — and a good test card demands a belief that can.

The criteria field is the one that hurts. Writing "we are right if…" before you run the test is uncomfortable, because it lets the experiment tell you no. That discomfort is the point: the threshold set in advance is what separates a real test from a demo you narrate favorably.

A Filled-In Test Card Example

Here is a test card for a common early assumption — that visitors to a landing page will act on an offer, not just admire it.

• Hypothesis: We believe that visitors who read our pricing page will sign up for early access.
• Test: To verify that, we will publish the pricing page with an "Get early access" button and drive a defined batch of visitors to it.
• Metric: And measure the share of pricing-page visitors who submit the early-access form.
• Criteria: We are right if the sign-up share clears the threshold we agreed on before launching.

Notice what the card refuses to skip. You cannot run this test without naming a metric, and you cannot claim victory without a pass bar set first. Whatever number arrives, you already know what it means.

Keep the hypothesis atomic. One card, one belief. If you find yourself testing pricing and messaging and demand in a single card, split them — a blended result leaves you unable to say which assumption passed. For how these cards sequence across a project, see the complete guide to startup idea validation.

Test Card vs. Learning Card

A test card and a learning card are two halves of one loop, used at different moments. The test card plans the experiment before you run it; the learning card captures what happened after. Confusing the two — or skipping either — breaks the loop.

Takeaway: Fill the test card first and the learning card second. The test card locks in the belief and the bar; the learning card compares reality against that bar and turns it into a decision. Together they make a single experiment auditable from intent to action.

If you run tests without learning cards, evidence evaporates and the same debates recur. The companion tool is covered in full in what a learning card is.

When and How to Use a Test Card

Reach for a test card whenever you are about to spend real time or money to check a belief your business depends on. In a corporate setting especially, the card doubles as a governance artifact: it makes a bet legible to stakeholders who were not in the room.

Start with your riskiest assumption. Do not card the thing you are already confident about. Rank your assumptions by how much damage a wrong answer would do, and card the most dangerous one first.

Set the criteria before you look at data — and honor it. The threshold is a pre-commitment. If the result falls short, the honest response is to treat the hypothesis as failed, not to relitigate the bar. A card whose criteria you rewrite afterward offers no protection.

Keep tests cheap relative to the risk. A test card does not require building the product. Fake doors, interviews, and landing pages all fit inside one, and the card helps you match a lightweight test to the assumption rather than over-investing in proof.

Key Takeaways

• A test card is a one-page planning tool. It comes from Strategyzer's Testing Business Ideas and designs an experiment before you run it.
• It has four fields. Hypothesis ("We believe that…"), test ("To verify that, we will…"), metric ("And measure…"), and criteria ("We are right if…").
• The criteria field is the crux. Committing to a pass threshold up front is what makes the experiment falsifiable and stops post-hoc goalpost-moving.
• The hypothesis must be atomic and falsifiable. One card, one belief that could actually be proven wrong.
• It pairs with the learning card. The test card plans the experiment; the learning card captures the result and the decision afterward.
• Card your riskiest assumption first. Spend the discipline where a wrong answer would cost the most.

Frequently Asked Questions

What are the four fields of a test card?

A test card has four fields, each with a fixed sentence stem: hypothesis ("We believe that…"), test ("To verify that, we will…"), metric ("And measure…"), and criteria ("We are right if…"). Filling them in order turns a vague ambition into a specific, falsifiable experiment with a success bar defined before you begin.

What is the difference between a test card and a learning card?

A test card is filled out before an experiment to design it — capturing the hypothesis, test, metric, and criteria. A learning card is filled out afterward to record what you observed, what you learned, and what you will do next. Together they form a loop: the test card plans, the learning card captures and decides.

Who created the test card?

The test card was popularized by Strategyzer and appears in Testing Business Ideas by David Bland and Alexander Osterwalder. It builds on the broader Strategyzer toolkit of one-page canvases and cards designed to make business assumptions explicit, testable, and cheap to check before you commit real resources.

Do I need a test card if my test is small?

Yes — small tests benefit most. A test card takes minutes to fill and forces you to name a metric and a pass threshold before you start, which is exactly the discipline informal tests skip. Even a five-minute interview or a fake-door button is sharper when you have written down what result would prove you wrong.