Create a prototype to do what

Sinh Vũ often hears this from clients: "Creating a sample is a waste of time; just put it straight on the website for speed." It sounds reasonable, but this is precisely when the cost of mistakes is at its highest. A sample is not a redundant step; it is a decision checkpoint before real money and effort are invested.

A template is a hypothesis, not a draft.

The correct perspective on prototypes: this is a way to express a design hypothesis in a testable form. The hypothesis could be "users will understand what this button means," or "this three-step ordering flow will not cause customers to drop off midway." When creating a prototype, you are asking questions, not providing answers.

Because as a hypothesis, a model can always be wrong and needs to be revised. Ripping up a sketch costs nothing. Scrapping two weeks of programming is very costly.

Two levels of refinement and when to choose each type.

There are two main axes that need to be distinguished:

• Low-fidelity sample: sketches on paper, gray frames (wireframes), without colors or real images. Viewers immediately understand this is a work in progress, so they feel free to provide feedback on all aspects, including direct criticism.
• High-fidelity sample: clickable, almost a live product, with colors, fonts, and real images. Users behave more realistically because it feels like they are using a real system.

Four questions to answer before production

Before opening Figma or taking out a piece of paper, you need to clearly understand these four things:

• What hypothesis is this template testing? User flow, screen layout, call to action, or brand feel?
• Who will see or use this template? Internal clients approve, real users during testing, or developers reading to build? Each audience requires a different level of nuance.
• How much time and ability to revise? If it needs to be sent to programming next week, it is not advisable to use a rough draft and then not have time to elevate it.
• After the testing session, what will you decide? If there is no answer to this question, the sample does not have enough justification to be built.

Common mistakes when using the wrong template

• Polishing the sample too early: Making the sample look almost like the final version while many elements are still undecided. The consequence is that clients hesitate to criticize, designers hesitate to make changes, and both end up defending an unverified version.
• Sample operation errors: pressing the wrong button, dead links, interrupted flow. Users lose direction and no longer trust the testing session. Always conduct a pilot test at least once before.
• Create templates without making any decisions: The templates are reviewed, praised, and then stored in a folder. Each template should lead to at least one decision: keep this direction, fix that point, or discard that idea altogether.
• Select the wrong level of refinement: using a high-fidelity prototype when many revisions are still needed is labor-intensive. Using a very rough prototype when genuine feedback is required will not yield valuable data.

A template is a tool for learning, not a deliverable product. When a designer starts worrying about whether the template looks good instead of whether it answers the questions, the template-making process has strayed from its goal.

Practical experience, Sinh Vũ Studio

How does Sinh Vũ work?

In the digital product design service (P1), Sinh Vũ divides the templates into two clear layers. The first layer is the journey map and gray interface framework, used to clarify what the customer needs and what the business wants the customer to do on each screen. This is a rough template, intended for discussion and quick adjustments. The second layer is the complete Figma interface with colors, fonts, and real content, used for final approval before handing over to development.

These two layers should not be rebuilt from scratch, but gradually refined according to the certainty of the decision. When the decision is unclear, the template should be rough. When the decision is clear, the template should be refined.