The Next-Gen Designer (2): Prototyping & The Full-stack Designer

I remember an incident a few years back that changed the way I looked at the prototyping process itself. I was working on a dashboard redesign for a client using a popular prototyping tool. It was supposed to show sales-metrics for their product, and I had to whip up a quick prototype for the presentation. Since we didn’t have the real data yet, I used some placeholder numbers—just to get the look and feel right. It was the usual stuff: "Product A: 100 units sold," "Product B: 50 units sold." Basic stuff, nothing fancy.

The design looked good. Interaction and the workflows were well thought through with a clean layout on point, and everything lined up perfectly. The stakeholders from pre-sales and product owners side loved the flow as well as the minimal looks, so we went straight for a quick user validation. I clicked through the prototype, feeling pretty confident, until one of the users squinted at the numbers and said, "Wait… why does 'Product A' only show 100 units? We sell thousands of units a day."

That’s when it hit me: the dummy data I used didn’t remotely reflect their actual sales figures. Worse, the way I had laid out the design didn’t accommodate larger numbers. The numbers overflowed in the cards, and the graphs were compressed beyond recognition. What looked crisp and organized with placeholder data now seemed chaotic and unrealistic for their actual use case.

The participant looked at me, visibly confused. "If this is how our sales data will look, this won’t work at all."

Well this was one of the many common challenges we face during the design process, especially while translating the concept to something tangible - a prototype.

Prototyping is a crucial phase in product development, but conventional prototyping comes with its own set of challenges. There are many factors, time constraints, resource limitations, balancing fidelity, iteration fatigue along with the cost constraints.

Deadlines are relentless, and time is often in short supply during the prototyping phase. Teams are under pressure to produce functional prototypes within tight timeframes. Rushing to meet deadlines often compromises the quality of the prototype and leaves little room for exploration or creativity.

Whether it's the lack of skilled personnel, materials, or technology, resource shortages are a common hurdle. Smaller teams may not have access to the tools or expertise needed to create high-fidelity prototypes, leading to delays or lower-quality outputs. This scarcity can force teams to make tough trade-offs between scope and quality.

Prototyping requires a careful balance between speed and detail. Low-fidelity prototypes are quicker to produce but may not fully capture the intended user experience. On the other hand, high-fidelity prototypes take time and resources but provide more accurate feedback. Striking the right balance is a constant challenge.

The prototyping process involves multiple rounds of iteration, and teams can face burnout. Constant tweaking and fine-tuning can drain energy and creativity, especially when it feels like no iteration is “the final one.” Teams may struggle to know when to stop iterating and move forward.

Budget limitations can cap the scope of prototyping, particularly when it comes to high-fidelity or physical prototypes. Costly materials or technologies might be out of reach, forcing teams to compromise on quality or scope, which can ultimately affect the final product.

Traditional prototyping methods often fall short as they fail to accurately reflect user interactions in the absence of real data. Features that depend heavily on real data, such as personalized recommendations and data visualizations, are particularly challenging to prototype effectively without access to authentic datasets.

with the increasing adoption of AI in web applications, the necessity for data-driven outcomes to be thoroughly tested during the prototyping phase has become paramount.

Prototyping in enterprise settings often faces serious challenges due to the lack of API details and incomplete understanding of data points during the design phase. Designers frequently don’t have access to these technical insights, leading to a gap between the design and actual implementation. This disconnect means that during engineering, adjustments are made to accommodate the real data or API constraints, often resulting in a compromised or broken user experience.

To avoid this, it's crucial for designers to collaborate closely with development teams early on, ensuring the design can adapt to real-world data and technical limitations.

The above two slides showing a bare-bone example of the gap identiofied post implementation.

With the growing role of data and machine learning in application ecosystems, it’s becoming essential for designers to evolve into "Full Stack Designers." This means going beyond the traditional scope of design and building a deeper understanding of the tech stack, including APIs, data flows, and backend systems. In today’s landscape, data literacy is no longer optional—it's critical for designers to grasp how data drives user experiences. By understanding these technical aspects, designers can contribute more effectively, ensuring their designs align seamlessly with engineering and deliver cohesive, data-driven experiences. This not only enhances collaboration but also future-proofs their role in the industry.

With the growing role of data and machine learning in application ecosystems, it’s becoming essential for designers to evolve into "Full Stack Designers." This means going beyond the traditional scope of design and building a deeper understanding of the tech stack, including APIs, data flows, and backend systems. In today’s landscape, data literacy is no longer optional—it's critical for designers to grasp how data drives user experiences. By understanding these technical aspects, designers can contribute more effectively, ensuring their designs align seamlessly with engineering and deliver cohesive, data-driven experiences. This not only enhances collaboration but also future-proofs their role in the industry.

In the context of data and AI/ML, a Full-stack designer would take care of the following in prototyping:

The first in this list is "Data Accuracy & Integrity". Prototypes often use sample or dummy data that doesn't always match the real data that will be in the final product. This can create problems during user testing or when validating the design because users may react differently to the actual data. If the data in the prototype is incomplete or inaccurate, it can lead to wrong decisions, as the design might look good in the prototype but not work well when real data is used. Later, during development, this can cause issues that might need adjustments, potentially delaying the project and affecting the overall user experience.

The next is the the simulation of real time data. The lack of real-time data simulation can make it difficult to test and validate time-sensitive features and interactions. For instance, when a prototype relies on real-time data from APIs, sensors, or live user input, tools like Figma often fall short in simulating that dynamic behavior. Without actual data flowing through the design, it's hard to accurately assess how certain features will perform, especially when timing and responsiveness are crucial. This limitation can lead to gaps in the design process, where issues only surface later in development, causing delays or requiring last-minute fixes.

In this regard the increase amount of data-points also play additional challenges. As the number of data points increase, manually managing the data points in prototype becomes significantly prone to error.

In data-intensive applications, actual dynamic data is essential for real user testing. Without live data, it's impossible to replicate the true experience of the application, which can result in inaccurate feedback during testing. Users need to interact with real outcomes, such as live metrics, personalized results, or real-time updates, to fully understand how the app will function in practice. This helps uncover usability issues, performance bottlenecks, or unexpected behavior that might not surface with static or dummy data, ensuring a more reliable and realistic testing process.

Will continue this discussion in the next post .. be in touch and feel free to share and comment.