From prototypes to business decisions - and why AI truly changes the game

Make it tangible

There is significant value in making an idea tangible and this is, after all, one of the core values of design. The prototype is where this value takes shape. 

Moving from an idea to a concept, and from a concept to a working prototype, brings a number of benefits:

• stakeholders align more easily around something visible and functional. Seeing what is being built helps build consensus and converge perspectives
• conversations become more concrete and less abstract, enabling discussions about how to actually build the solution
• ideas can be tested and validated by exposing the prototype to stakeholders or directly to users, collecting actionable feedback
• blind spots emerge more clearly: building a prototype forces reasoning about flows and interactions, clarifying areas that were still undefined
• it creates a sense of concreteness around the output, and therefore stronger commitment to the project

All of this has been known for years. Yet in many contexts, it still struggles to become the dominant way companies approach innovation in products and services from a strategic (rather than solution-first) perspective.

This is one of the areas where, as Tangible, we consistently bring value and method into projects.

AI: more speed, but not only

Speed is the most visible aspect, but it is not what truly changes the game.

Prototyping is a strong example of AI augmentation: it’s not just automation or efficiency gains, but an amplification of the activity’s value.

Reducing build time is only part of the shift. The real impact lies in shortening the distance between intuition, validation, and decision.

Let’s look at how.

Time

In the past, two of the main constraints in prototyping were time and effort. This required carefully balancing trade-offs between fidelity (how real it needs to feel) and time to build (and therefore cost), in order to find an effective sweet spot between concreteness and speed.

Today, with AI tools, the time required to build a prototype tends to collapse toward zero. This is significant because it also removes the constraint described above: there is no longer a real limitation to prototyping an idea.
You can iterate more quickly, explore multiple directions before committing, test faster.

In practice, some constraints (and excuses) for adopting truly iterative, prototype-driven innovation processes have disappeared.

Speed should not be interpreted as a proxy for value (less time = less cost = less value). On the contrary, it is an advantage that brings the moment of intuition closer to validation and business decision-making. Used as a strategic enabler, it increases value by shortening the path from idea to execution.

Level of fidelity

Until recently, fidelity was the second key variable in prototyping, both for practical and effectiveness reasons.
Low-fidelity prototypes are quicker to build and tend to focus conversations on macro aspects: flows, structure, and the core “reason why” of the product.

High-fidelity prototypes, on the other hand, create a stronger sense of realism. They are useful for user testing and for focusing discussions on micro-level details.

Today, AI tools allow us to reach much higher levels of fidelity in far less time.

Is this always a good thing? Not necessarily.
As discussed, fidelity influences other aspects - and more is not always better.

The risk is taking three steps at once and jumping straight into the solution space. This can lead to skipping proper problem exploration and shifting the conversation too early toward implementation details.

Highly realistic prototypes tend to anchor discussions on details. With AI, this risk is amplified: exploration may close too early, giving a false sense of solidity to something that hasn’t been fully questioned.

Tools matter less than we tend to think. What matters is the method.

Lovable, Claude Code, Google Stitch, or any other tool can be effective depending on context and goals. What makes the difference is how the prototype is used: to create clarity, surface questions, and enable meaningful decisions.

We’ve written more about problem framing and research elsewhere, both areas that are also becoming more effective thanks to AI.

Data and scenarios

Another less visible but equally important aspect is the role of data.

AI makes it possible to generate synthetic datasets, simulate realistic distributions, and test not only interactions but also system behavior in conditions closer to reality.

The prototype is no longer just a representation of an interface—it becomes a tool to explore scenarios.

Where the value really lies

The value of prototyping, both in the past and today, lies in the quality of the business decisions it enables. It allows teams to make ideas concrete, evaluate and validate them more effectively, create alignment and momentum, and reduce the uncertainty inherent in innovation.

When embedded in a solid design process, AI amplifies this mechanism: it accelerates validation and strengthens decision-making, creating strategic and not just executional value.

In corporate environments, having a semi-functional AI-generated product built in a few days often has limited value. It quickly runs into constraints such as compliance, security, legacy systems, or organizational dynamics.

Value emerges when a product or service vision (looking 1–2 years ahead) can be made concrete and testable. 
A prototype can align different functions, guide technology choices, and support investment decisions.

There is also a less visible effect: the process of building the prototype forces teams to enter the domain, clarify assumptions, and develop a shared language. This level of understanding rarely emerges through documents or presentations alone.

In this sense, the prototype shifts from output to input. Its value lies in the quality of the decisions it unlocks, and the speed at which those decisions can be made.

AI Product Vision

This need - to make better decisions in less time - is what led to our work on Design Vision.
Around 2015, we began structuring a process to work on product and service ideas more concretely, drawing inspiration from Jake Knapp’s Design Sprint and extending it to more complex contexts requiring deeper exploration.

Over time, we’ve applied this approach across many different products.
Today, AI pushes it further.

The ability to prototype faster, explore multiple directions, and simulate scenarios makes building and validating product visions even more effective.

This is what we now call AI Product Vision: an evolution of our approach, where AI becomes an enabler of decision-making.
Prototyping is easier than ever. Using it to make decisions remains a competitive advantage.

If this resonates and you’re working on the vision of a product or service, get in touch.