A conversation analysis platform (desktop & mobile) focused on qualitative research interviews. My first product taken to market.
Project type
SaaS, Start-Up
Role:
Founder, CEO & Product Designer
Year:
URL:
Conversalis.com
The mobile app
Context
Conversalis is is a productivity tool that allows users to automatically transcribe their most important conversations and find key insights for qualitative research purposes. The platform automatically generates a shareable report that presents key findings.
Tasks & responsibilities
Product Design: End-to-end design process: Ideation, user research, competitive benchmarking, UX, UI, information architecture, prototyping, user testing, developer hand-off, iteration improvements & collaboration with developers
Leadership: Strategy, budget planning, fundraising, recruiting, contract management, legal work, cap table negotiations, updating investors, applying to incubator programs and grants
Project management: Product roadmap, user stories, bug testing, documentation, information organisation
Marketing: Branding, competitive analysis, promotional videos, landing page creation
Why I built this product
My desire to create innovative products coupled with cutting-edge technology is one of my passions. It all started from a personal idea, to a blank piece of paper in 2018. My goal with this project was to allow users to make better data-informed decisions based on the actual voice of the users to make the right strategies to achieve their goals. This allows our users to make better products, services, and organizations, making the world a better place.
Problem statement
Organizing and communicating qualitative data from research interviews was remarkably complicated and time-consuming. Yet,incredibly rewarding.
Launch of the MVP
Bringing an idea from 0 to 1 in 2020
The people who use my product
The ones who pay for my product
Promotional video
Marketing video for the legal sector
A 1 minute video is crucial when taking a product to market and social media. This is the video I created in collaboration with my friend Timothée Bart to showcase our product's key features for lawyers in a short format when we pivoted the product to the legal sector.
Product walkthrough
Walkthrough of the app with what we currently have online today
Research Methodologies
Research interviews: 10 (30mins-1 hour long interviews)
Co-building with prospects and users: Stakeholder workshops (2 brainstorming sessions)
Benchmarking: Exploring and discovering the state of the art of similar existing solutions
User testing: Prototyping, validating and iterating with end users
During the first 2 months of this project I did UX research fulltime. The idea to create this project came after interviewing 10 different UX designers, researchers and project leaders from around the world. Facing the problem firsthand when I needed to present my insights to other team members I saw the potential.
I searched for tools to help me properly do this task extensively and I didn’t find one so I decided to create it my self. I first built a prototype that I iteratively validated by industry experts and built a pitch deck, with which I raised the funds to start the company.
UI components
I created the UX of the entire product and collaborated with a graphic designer to build the UI of these high fidelity components
Search and replace widget
This design allows us to merge the "search" feature and the "replace" feature within a same widget effectively. During user testing, users requested for a "search and replace" feature numerous times.
Tool tip for tag creation
Highlights text and giving it a tag is the core feature of our app, we revisited the color scheme to make the color contrast more accesible
Adding and editing speakers
Manually adding a speaker to the transcripts is necessary and embedding speaker metadata is incredibly useful in the analysis phase when comparing diffrent interviews
Creation of a new speaker + metadata
Each speaker would have its own metadata, this form allows the user to add custom data fields and it allows them to filter content in an even more precise way to find insights .
High fidelity UI designs
Results
What was achieved
Took a product to market and made sales from real life paying customers. (Our first sales were made to university students who where doing qualitative research interviews for their final year master's thesis)
We got accepted to enter the most prestigious tech startup incubator in France called Station F.
Shortly after joining we were pre-selected by our program manager as one of the most promising startups in all of the campus.
Business deals
Contracts
I managed to start negotiations with high-level businesses to start pilot deals:
I did multiple workshops to co-build the product with the chambers of commerce of Bogota and Medellin, for them to integration our solution into for their operations.
Multiple meetings with the gendarmerie,( the French National Guard ) who also showed high interest in starting a pilot deal.
We had conversations to empower Deloitte with our services.
Development
Building a product from scratch
A remote-first dev agency from Belgium named Madewithlove.be was hired to assure solid quality and the right foundations for the product. They had good references and very good online reputation. We were doing innovation at the time and had no reference point on how to build this project from a technical standpoint. This made it challenging for the agency to accomplish the task, they under estimated the project and they were not able to deliver what they had promised on budget or on time. This turned out being a fatal mistake for our startup. We decided to pay for the premium price tag to build an MLP (Minimum Lovable Product) because we wanted to make sure we had solid quality foundations for our product to build on top of, to later find out that our product had structural flaws on a fundamental level which we were never able to recover from.
Learnings
Never start a tech project with a Devshop uness you are a developper yourself.
Never start a tech project without a CTO Co-founder.
Finding a good CTO is comparable to the task of finding the love of your life.
Building a startup is very much about the contacts one has.
Building a startup is building a company first, not building a product first.
It's all about the team.
Don't be afraid of sharing your work publicly even if its not perfect.
Being afraid competitors were going to copy me was a mistake.
Selling to big corporations is a long process that requires validation from many different people or departments.
Perfectionism is a weakness.
Competing with startups who have raised millions in funding shouldn’t t scare me.
At the end its all a money game. It starts with funds and it finishes with funds…
Challenges
North star metric
Our north star metric on this project was the number of times our users downloaded or shared the reports which contained the insights of the research project. This is because it signals that they are using our product to its full potential as they will be sharing valuable information crated with our app.
Heap
Product analytics
Product analytics
Web traffic analytics
Sleek plan
Customer feedback
Customerly
Intercom chat bubble
The iterative process
The product evolved and was redesigned multiple times as we made progress
POC using material design
MVP using ant design
V1 using tailwind
V2 using our own design system
Mobile app
React native mobile app designs
The original idea of the product was to have a system that allows the users to record conversations during face to face live interviews on a smartphone and analyze the interviews on a desktop computer.
Importing conversations from other apps like google meets or zoom with highlight tags
I designed the Plugins to be able to record conversations through google meets and zoom with the possibility to manually highlight or tag key moments on the spot
Branding & Naming
Logo + Naming iterations
The startup iterated its name several times, although we had bought the domain name: Conversate.app, it had to be changed for legal reasons to finally land on Conversalis.com & Conversalis.ai
Concept: Enable users to better aim and hit their targets thanks to understating key insights
Logo polishing by : Jeroen Van Eerden
Name iteration
We had to change names due to legal issues, we went for a friendly and approachable name while we found the right one
Final logo design
Initial prototype
After several months of research and wirframing I created a high fidelity prototype in invision to bring my ideas to life and communicate with early potential customers and ivestors.
