7 min read · Last update on Aug 8, 2022

One of the beautiful things about design is that there is never just one way to solve a problem.

During the creative process, it’s common for designers to have multiple UI designs and flow possibilities to accomplish the same goal. Which one should we pick? Which alternative users will like most? Which one will create a more positive impact on the users and drive more engagement and revenue?

To answer those questions, during my career, I had the chance to learn and practice different approaches and tools that led me to build my own interpretation of the design process, focused on getting the best of creativity, designer efficiency, and product performance.

Here I cover the main points and suggestions of this process that I’ve been improving as a Product Designer and leader in design and product teams.


  1. Principles
  2. Understanding the product and business goals
  3. Understanding the users
  4. Defining the design and data strategy


  1. Make tangible for businesses the importance of investing in Product Design.
  2. Help the designer deliver user experiences focused 100% on the users.
  3. Achieve high user satisfaction and long-term results.
  4. Leverage creativity without risking objectivity.
  5. Reduce time spent on tasks that have a low impact on the users and the business.
  6. Generate knowledge through every iteration to improve future hypotheses.
  7. Make design-with-data a fun, creative, and rewarding experience.

Understanding the product and business goals

Before getting creative and designing a new product or feature, designers should check if there are clear user-oriented goals and KPIs that will allow them to be creative without getting off-track from the company and users’ intentions, and be capable of measuring how the designs perform.

KPIs should be related to a leading indicator of revenue and repeated behavior.

Ex: % of customers that purchased after the first time, % of users that wrote a review after the first time.

The customer satisfaction KPI, such as NPS and CSAT, is fundamental to avoid situations where the UX increases growth metrics but decreases customer satisfaction.

Ex: Users are buying more but getting angry after purchase due to misinformation in the checkout, lowering satisfaction and recurrence.

Besides being a quantitative indicator, it can also become a qualitative source to identify pain points. If the survey form has an optional comments area, some users will write about why they are rating in a certain way, especially detractors.

If the product or feature has no way to be measured yet, designers should assist the company and teams to find the metrics before diving deep into the design.

Understanding the users

Market research

Useful to find context and understand which market the product/app belongs to, and which audience the competition and similar apps are targeting.

At this stage, designers should search for research and reports related to the audience and talk to the company founders (if accessible).

Talking to operations teams

CR, Sales, Logistics, and other operations teams are rich in user knowledge: their pains, incidents, and frequent doubts/questions. Members of those teams are constantly full of ideas and suggestions that can help build hypotheses for UX solutions.

It’s important to maintain a close relationship with those teams and remove any existing communication friction between them and the designers.

When their inputs transform into a design experiment, if designers share the results with them, they will be able to see the impact of their knowledge on users and get encouraged for future participation.

User Interviews

Well known in the design area, this stage helps understand the connection between the users and the problem that the product/feature/hypothesis is trying to solve, adjust the KPIs (if needed), and discover similar products they might be using.

The interview process needs to be consistent and based on exploration. It’s about knowing users better, not validating the business model, a feature, or a hypothesis.

For new products (MVP), it’s interesting to start with 1:1 interviews with the potential audience to get to know them better and, for existing products/apps, establish periodic interviews.

A big concern with User Interviews is confirmation bias. A simple question wrongly asked can express pressure or lead the interviewees to say what they believe the interviewers want to hear.

A good practice is to create a template with questions for triggering conversation and, when interacting with users, to do curiosity-related questions, such as: “Why?”, “What do you mean?”, “How did you feel?”, etc.


For a new app or early-stage startup, it might be better to keep audience profiling open and let the product engagement show who is the engaging and not engaging audience.

Looking at traffic and engagement data might bring new and surprising audiences that can drive the company in a different direction. Designers can group the audience into behavioral patterns and later schedule User Interviews to establish personas for each different group (if any).

For existing products/apps and companies with more resources, relying on research experts will reduce the chance of building a low-quality persona that could drive the team in the wrong direction and distract people from seeing other good profile possibilities.

It requires quantitative and qualitative data, time, and expertise to build high-quality personas.

Usability Tests & Session Recording

As essential tools to help identify friction, they are recurrently used in several stages of the process.

When doing usability tests, it’s effective to give users a simple and clear objective and observe how they will complete it. Designers should look for emotional signs, interaction behavior, and ask questions that will help to understand the underlying reasons for a funnel drop-off or low engagement of a feature.

Using session recording tools like Inspectlet and Hotjar are interesting additions to observe users engaging with the website/app in their environment, without the psychological pressure of Usability Tests that can lead to biased comments and behavior.

To constantly spot friction points, designers should book in their agenda recurrent 1:1s with users and potential users to observe them using the apps and ask specific questions. Opening every day a few recorded sessions to watch users interact with the app/website/feature is also a good practice.

Defining the design and data strategy

Case 1: New app/website from scratch

As a Product Designer building an MVP, in the beginning, rarely there is data to help assist in the decision-making.

Designers should focus on helping the user to execute the app’s main function as frictionless as possible and to launch the app quickly to start collecting data.

The overall process includes:

  1. Funnel definition
  2. Writing down a list with the essential elements on each page
  3. Wireframes
  4. Building a simple UI Kit or getting an existing design system (Material UI, for example)
  5. Embracing patterns and market standards.
  6. Transforming the wireframes into final designs
  7. Checking for missing error states, feedback messages, and loose ends.
  8. QA
  9. Usability tests

In the MVP, the app might end up similar to others in the market but if the design and development team established metrics from the start, later by understanding data, iterating, and testing, designers can start finding the sweet spot between aesthetics and function, making sure the app has personality and performs well.

Case 2: Improvements to an existing app/website

In this case, the objective is to mix quantitative with qualitative data and turn every idea and iteration into measurable experiments. That means incorporating A/B and multivariate testing into the data toolset.

When bringing A/B/n Testing to the design process, it’s important to analyze the number of users and daily visitors, as it influences the experimentation and data strategy.

Few daily visitors and users

If there are not enough users for having A/B tests with statistical relevance, Usability Testing becomes the main source of data to decide which UI and UX will be delivered to the users.

The design process is similar to the one for new apps but it might vary depending on the complexity of the improvement, ex: Redesign, text change, small feature, or a new flow.

The overall process includes:

  1. Writing down a list with the essential elements on page/feature
  2. Wireframes or quick sketches
  3. Reusing components from the UI Kit as much as possible
  4. Transforming the wireframes into final designs
  5. Checking for missing error states, feedback messages, and loose ends.
  6. QA
  7. Usability tests
  8. Metrics and funnel tracking tests
  9. Shipping to production as a feature flag, if possible

If the improvement is a new flow, it’s important to define the funnel and verify early if developers have the resources to track when in production.

One of the main benefits of A/B testing is to mitigate risks, being able to pause the experiment at any time and redirect the traffic back to the baseline version.

If there are no A/B tests, due to the lack of traffic and statistical relevance, it might be possible to ship it as a feature flag. That way, it can be sent to production faster and, if by any chance the experiment is broken, it can be paused to stop users from interacting with it and fixed later.

Shipping as an A/B test and analyzing the results through heuristics is also a possibility but it should be done with caution to avoid biased decision-making.

Many daily visitors and users

In this situation, is all about delivering changes with A/B/n testing. It’s a great way to measure impact and also to give anyone, not just the designers, a creative edge.

It helps gather knowledge about the users on each experiment, helping avoid spending resources on future ideas that are likely to fail, and helping to focus on the ones that have a higher chance to create impact and drive the KPIs.

When defining the design strategy with A/B Testing, I’m usually analyzing these two approaches:

Narrow down before A/B

Designers spend more time narrowing down the UI/UX options by doing Usability Testing and interviews before running the A/B test.

After having the experiment result, if it didn’t have the expected outcome, they should go back to UX Research to understand why and then iterate.

Don’t narrow down before A/B

Designers spend less time deciding which UI/UX alternative to choose and run a multivariate test instead.

After having the result, they should do Usability Testing and Interviews to figure out why it worked and why it didn’t work, and then iterate.

I tend to choose the second option if the development team is capable of fast iteration and the product/website has enough visitors to reach statistical relevance in a short period. Mainly because:

  1. Human assumptions are usually wrong and the unexpected outcomes from multivariate experiments can lead to other opportunities.
  2. Users interact with the experiments in their environment without the risk of biased data due to psychological pressure, common in user testing and interviews.


The best of the process is extracted through iteration and time. The team needs to get comfortable with the mindset shift and keep fueling it with ideas and better hypotheses. To reach that point, it’s important to keep recycling and reusing the following tools:

  • Usability tests
  • User interviews
  • Benchmarks
  • Talking to operations teams

Designers should reuse them regularly and nurture curiosity. Asking “Why?” it’s the most powerful tool.


This design process is a mix of well-established design and growth tools. The insertion of quantitative measurement tools is done in a way that helps designers to get more knowledge about their users and designs. It helps the impact on users and businesses to become tangible and to empower others to be creative.

If you are interested in discussing the process, sharing your learnings, or discussing how to implement it in your team or company, please don’t hesitate to reach out to me.

I will be happy to have this conversation.