UX Research

Building a customer feedback loop to guide product development

Senior Product Designer
2021 - 5 Months
User Research
Feedback Automation
Data-Driven Personas
Presenting User Insights


Novoresume is a tech startup with +6 million users, so lot’s of different opinions and user requests. The challenge was making sense of all these opinions to help prioritize what we should build next.


The outcomes of this project consist of more tangible outcomes, like 5 data-driven personas and a semi-automated feedback loop build in Airtable. The most important outcomes are the customer insights and a higher level of customer-centricity within the whole company.

Data-Driven Personas

After extensive qualitative and quantitative research, the first outcome is a personas spectrum consisting of 5 different personas within the area of job-searching. The personas are grouped in terms of work-experience, and all have their own unique ‘Jobs to be Done’, pains and gains. For a deeper dive into the construction of these personas, scroll down to see the research process.

Semi-Automated Customer Feedback Loop

The primary outcome of this project is a semi-automated system that gathers and groups user feedback that is coming in daily. Around 25 people write their opinions to us each day, which are gathered in a tool called ‘Get Feedback’. With a ‘Zapier’ integation, I made this data flow into Airtable, which gives the overview of all the feedback. Each month I present a 'Voice of the User' presentation to the whole team to let these user insights flow into the organisation. After this feedback loop was set up, it became a crucial guide for deciding which features need to be changed or built next.

Design Process

My role in this solo project was of the UX Researcher. I initiated the project and handled all the steps from finding the needs in the team, to building personas and finally a Feedback Loop.

1. Interview all team members

Before doing any user research or design activities, I wanted to know what the current level of ‘customer centricity’ was in the organisation. How could my research activities and insights help each member of the team? Instead of keeping the UX insights within the design team, I hoped to find ways to support marketing, development and product strategy as well.

My approach was to interview all 12 team members and ask them questions about what they know about our users. Do the employees truly know who the users are, what they struggling with and what we can do to support them?

*I have blurred the post-its as the insights are confidential

2. Mapping the needs

The interviews resulted in very interesting findings and opportunities how UX research and customer feedback could help each individual team member in their current projects. The quotes were all ordered and analysed by categorising them in different ways. The final mapping shows how each individual project could be moved forward by user insights.

*I have blurred the post-its as the insights are confidential

3. Make a roadmap for the CX team

After presenting the outcomes to the team, a roadmap was made with a clear prioritisation of what the CX team should focus on first and what could be potential future steps. The most crucial things to do first appeared to be the construction of Data-Driven Personas and a Feedback Loop that lets users provide feedback to the team. The rest of this page is going to focus on these two projects.

4. Digging through old quantitative data

In order to build the data-driven personas, I would need a solid data set consisting of both quantitative and qualitative data. Before gathering this data myself, I checked the existing data sources to find if they could be re-used for the personas. Checking our LinkedIn + Facebook engagement and Google Analytics Demographics provided the first step into understanding our users better. Still, it was hard to overlap these data sets to derive insights from them, so new data was required.

5. Gather quantitative Survey data

In order to gather more quantitative data, I created a survey that would be shown as a pop-up to our users. 5.004 respondents filled in the survey; people from all over the world, age groups, industries, and work-experience levels. Certain countries were way overrepresented in comparison to the distribution of our users, so the data-set had to be cleaned up, so it represents our real users correctly. This resulted in the fact that I could only use 3.080 respondents.

6. Derive insights from quantitative data

Going through different data mappings led to the idea of separating our personas by the level of ‘work experience.’ The quantitative data helped to highlight basic demographics such as country, industry, paying or free customers, etc., but the ‘juicy details’ were lacking. What motivates these people? What are they struggling with? What is their ‘Job to be Done’? These questions could only be answered by adding a solid amount of qualitative data to the mix.

7. Interview users to gather qualitative data

In order to find qualitative insights, I decided to interview a number of our users. First, I had to outline what kind of participants we would need to represent the diversity of our user group as closely as possible. Through our customer support, we recruited 16 users that agreed to participate. The questions were all prepared, and the interviews were conducted in sessions of 40 min.

8. Derive insights from qualitative data

This was the most time-consuming phase of this research process; mapping all the quotes and finding patterns and differences between the different people. Several iterations of re-structuring, grouping, and discussing were held. This enabled us to write user stories, ‘Jobs to be done’ and find the main pains, gains, and goals between the different personas.

9. Combine all data to form data-driven personas

The qualitative insights were added to the quantitative ones, giving a solid understanding of the differences between our personas. In the end, this research process led to the definition of 5 different personas, categorized by their level of work experience; Students, Juniors, Mid-Levels, Seniors, and Freelancers (who are a bit different than the rest). The five personas were designed into nice-looking structured templates to be easily shared and used by the others in the team.

10. Introducing the personas to the team

The most crucial step in building personas is that they are actually being used. They shouldn't become a relic or dusty file inside a Google Drive; they should guide product decisions, help to target people in marketing, and assist in design discussions. Therefore the personas were presented to the team and examples were given on how the different departments could use them. Especially the marketing and design team understood the potential and started using the personas immediately in their work.

11. Learn about current feedback processes

After the personas were constructed and used, the last step would be to build the feedback loop. To do this properly, I first needed to understand thoroughly through which channels we currently receive user feedback and the differences between these channels. Next to that, I outlined our preferred channels and separated the unsolicited feedback from the solicited.

12. Research best practices in gathering feedback

Every customer-centric company uses user feedback to improve its service or product, so there is much knowledge around the topic available. I’ve done thorough research in best practices and case studies of how other companies gather, analyze and use user feedback. This resulted in a clearer understanding of how we could approach our feedback system and which tools could be handy.

13. Finding the right tools

After that, we needed to define which tools we should use to gather, structure, and use the feedback. A thorough comparison was made to find which tools have the highest quality, lowest price, integration capabilities, security, and interesting additional features. The conclusion was to go for 'Get Feedback’ as the tool to gather the feedback, ‘Zapier’ to make the data flow, and ‘Airtable’ to store, label, and categorize the feedback.

14. Building the Feedback Loop

After buying the subscriptions to the tools, it was time to set them up and make them talk to each other. The biggest challenge here was to build the Feedback Loop without any help from the developers, as they had other projects to focus on. Therefore I’ve built the integration with Zapier, as a ‘no-code’ alternative to making the data flow into Airtable.

15. Let the user insights flow into the team

Once the Feedback Loop was set up, tested and working, it was just a matter of waiting until the feedback was coming in. Currently, around 25 people leave feedback each day, which all goes automatically into the right spot in Airtable. The only manual action that needs to be taken care of is tagging the items on a monthly basis. This manual effort makes the feedback incomparably more valuable.

I presented the first results to the team in the form of a creative workshop. The biggest problems from our users were presented and in groups, our team went on to ideate and prototype solutions for these problems. This was one of the ways I hoped to let the insights flow into the team.

16. Guiding product development

The goal of this Feedback Loop was to guide product development. In order to achieve that, I arranged a monthly ‘Voice of the User’ presentation where I would go over the main feedback. This helped to raise awareness in the team about how the users perceive our product and what the next steps in product development should be.

"Pieter's ability to bring order to 'chaos' is impressive. I've had the pleasure of working directly with him during his stay with Novorėsumė.  He has a capacity to filter large amounts of data, analyse them and make sense of them.

He's highly structured, and dedicated to his work. He has helped us build a Feedback Loop, where large amounts of user input comes in and is then filtered and sorted. Pieter also conducted interviews and surveys to build our personas. His UX Researcher skills are not the only ones that shine though, as he has also brought great work into Interface Design and Design Systems.”

Cristian Letai

CDO & Manager, Novorésumé

Explore Other Work

A selection of Service Design, UX and UI projects.

A mobile-first redesign of a digital guestbook for hotel guests

Responsive Design
UI Design
I helped a booming startup in the hotel industry switch to a mobile-first approach of their product. Check how hotels can make custom mobile apps for their guests.
View Case

Prototyping for a personalized Volvo car-sharing service

Service Design
Personalized Prototyping
Academic Research
Through 4 rounds of prototyping and testing, a new car-sharing service was developed with Volvo. The focus was on personalization and resulted in integrating existing mobility services.
View Case

Harnessing the power of IoT in the marine industry

Digital Prototyping
UI Design
Using sensor data enables a range of new possibilities. One of them is this IoT dashboard to track and tune engine perfomance of big container ships.
View Case

Establishing a flexible design system

Design System
UI Design
A crowded user interface needed a new approach to be more aligned, better looking and ready for scaling up with new features in the future.
View Case

Visualizing Italian elections through social media activity

Visual Identity
Data Visualization
Responsive Design
Is it possible to predict the results of a political election by analyzing posts on social media? I helped visualizing an innovative research that examines the relationship between Twitter activity and election results in an Italian region.
View Case

Branding a freshly founded design studio

Visual Identity
Brand Design
Responsive Development
Logic Moon is a design studio of which I’m a founding partner. This project focuses on the branding and building of the company’s visual identity.
View Case

Digital prototyping for a financial Robo-Advisor

Design Sprint Facilitation
During my employment at Nordea Bank I was involved in several projects, of which one was prototyping parts of their Robo-Advisor called ‘Nora’.
View Case

Interactive storytelling on cultural diversity

Workshop Facilitation
Interaction Design
How is migration in Italy’s most culturally diverse city perceived by the youth who grows up there? Youth in the City results in an interactive story after a co-creative workshop with 48 local high school students.
View Case