AI in Product Design: How Figma Make speeds up prototyping

AI is changing how product design work starts. Tools like Figma make it easy to generate layouts, test ideas, and move from concept to an interactive prototype much faster than before.

This article brings together the main observations from an internal AI meetup at Kellton Europe, where a Senior Product Designer shared how she had been testing Figma Make in practice. The observations are based on tests with different types of design tasks, from standard e-commerce flows to concepts prepared for client proposals.

In product design work, it is common to start by exploring different approaches to a problem. With Figma Make, this exploration can happen directly in an interactive prototype. Several versions of the same flow can be generated within a relatively short time, each presenting a slightly different solution.

At this stage, the work often shifts from designing to evaluating. Instead of refining a single solution, the team compares multiple versions and tries to decide which direction should move forward. This change in focus slows the process and makes feedback less precise.

Generating more versions does not necessarily make the decision easier. In design workflows, AI-generated variations can create exactly this problem. Figma Make works better when the designer first understands the project, chooses a specific part of the product and gives the tool clear instructions.

A more predictable workflow separates these stages and introduces a clear sequence. This was also the approach presented during our internal meetup.

The process shown during the session involved four stages:

Before opening Figma Make, it is useful to understand what the product does, who will use it and what the selected flow needs to achieve. When a project includes several documents, spreadsheets or feature descriptions, AI tools can help organise the information and explain the product in clearer terms.

For example, NotebookLM can be used to analyse the available materials and prepare an initial product overview.

This makes it easier to identify the part of the application that should be presented in the prototype. The analysis still needs to be reviewed by the designer. AI can summarise the documentation, but it does not know which requirements are critical or which assumptions should be questioned.

AI can support early exploration by helping identify different ways a flow could work. It is usually more effective to generate one specific part of the application than to ask for the entire product. This could be a product listing, checkout flow, dashboard or configuration screen.

Working with a smaller scope makes the result easier to assess. The designer can check whether the flow is logical, whether important steps are missing and whether the concept is worth developing further. It also helps avoid spending time refining parts of the product that are not relevant to the proposal or current discussion.

Once a direction makes sense, define how it works. This involves:

During one of the tests presented at the meetup, Figma Make was asked to create a modern e-commerce platform for shoes, including a product listing, filters and a customisation panel. The first result contained the requested elements, but the customisation options were arranged vertically, which made the screen difficult to use. The composition had to be described more precisely so the tool could adjust the layout and position individual elements correctly.

A screenshot, wireframe or existing Figma design can also be used as a reference. This helps the tool understand the expected composition, although the final result still needs to be evaluated by the designer.

After the structure is defined, returning to Figma Make becomes more effective. At this point, AI helps with execution by generating layouts, maintaining consistency, and handling repetitive elements.

The design follows a defined direction, which makes the output easier to evaluate and refine.

Keeping multiple directions active for too long introduces uncertainty. Design files become harder to navigate, feedback becomes less focused, and decision-making slows down.

A single, clearly defined direction simplifies collaboration. It provides developers with a stable reference and allows product teams to move forward without revisiting earlier assumptions.

This part of the process often gets overlooked when the focus stays on tools and output. In practice, alignment, ownership of decisions, and the way teams communicate have just as much impact on the outcome as the design itself. We explored this in more detail in our article on the human side of product design.

A more focused context also makes the AI-generated output easier to control.

Introducing a structured approach shifts effort to the earlier stages of the process. More time is spent on understanding and defining the flow, which reduces the need for rework later.

Design files become easier to understand, feedback becomes more specific, and handover to development becomes more predictable. The role of AI here is mainly to speed up selected parts of the process rather than replace the decisions behind them.

Figma Make works best when the direction of the flow is already clear and the work moves into execution. In UX design practice, this means building screens based on a defined structure. The logic is agreed, and the focus shifts to consistency and completeness.

At this point, Figma Make can help generate layouts faster and extend familiar patterns across screens.

Figma Make can shorten the time between an idea and an interactive prototype. It is particularly useful for client proposals, early product exploration and standard interface flows. Its limitations become more visible when the project requires custom branding, a scalable design system or a file prepared for development.

One of the clearest takeaways from the tests presented during the meetup was that Figma Make is currently more useful for presenting and winning a project than for delivering the final product. It helps designers present an idea quickly. It does not replace the need to understand the product, make informed decisions and prepare the final solution for implementation.

If you want to explore how this looks in practice, you can take a closer look at how we approach product design in our work at Kellton Europe!