Breaking and building a SaaS
The mistakes I made, the lessons I learned,
and the product we designed.
Company
Avec
Context
Even as a startup in 2018, Avec was the largest Brazilian tech company in the beauty and wellness industry, with more than €9 billion transacted through our management tools (SaaS ERP). At that time, I was leading a small team of two other product designers with a strong UI drive.
What I did
— Problem investigation
— Ideation workshop
— Prototype
— Alignments with tech
The challenge
In Brazil, many beauty and wellness businesses founded by expert hairdressers and wellness specialists struggle with operations. Our solution automated tasks for ease. However, in 2018, Avec faced difficulties in transforming our 2 products from selling one-time packages to offering an online monthly subscription (SaaS) business model.
When we started this project, our POs were two people who had created some of the most successful ERPs for the beauty industry in Brazil. They had strong opinions regarding how we should work, what customers needed and wanted, and therefore, what we should deliver. Their requests included:
Prototypes must be high fidelity & sexy
Looking back, it's clear that they wanted people to buy into their ideas, rather than genuinely taking the risk of being wrong in a validation session.
Involve users with key modules prototyped
This created about six months of stressful, unclear, and hard work, significantly increasing the risk of bigger failure.
Recalculating
the route
This is a case about the mistakes I made, the lessons I learned, and the product we designed. And after six months of designing based on internal ideas, we found our designs didn’t meet our customers' needs. Recognizing our mistake, we implemented a UX process, adapting Google Sprints to involve users from the start, understand their needs, and validate designs in shorter cycles. This shift towards a user-focused approach was pivotal in our progress.
The marketing module
After implementing a more organized process, we revised and developed new modules for our ERP, with a focus on the marketing module. This module provided a solution for those lacking in-depth marketing knowledge, offering CRM features and automated marketing campaigns to help attract, engage customers, and increase profits.
Briefing
To start our marketing module, we discussed with our stakeholders why we were creating it, what our expectations were, and what the basic business requirements were. Now, it carries half the weight, being one of the perspectives that we will consider when designing the solution.
Shadowing and interviewing users
To build a digital system for businesses, we needed to understand their real-life operations. To do this, we observed and interviewed users, revealing that their practices differed significantly from our assumptions. Some users even ignored our systems to use Excel sheets.
Benchmarking
For this product, we didn't have a model to follow in our field, so we looked to other industries for ideas. We took inspiration from RD Station's use of conditionals and lists, but we kept it simple. We liked how MLabs made managing marketing campaigns easy for small businesses. We also learned from iFood, Brazil's leading food delivery company, on how they use marketing tools and present results to their partner restaurants.
Ideating
For this part, we would be getting together and shaping how our flows, and interactions would be. We did that with crazy 8s (yes, drawing and folding papers), and we created our flows using stickers through the office, it gave freedom to our team to explore different approaches to seam our ideas together. Being far from our laptops made the ideation more fluid and creative.
Prototyping
As we had just 2 product designers, it didn’t make sense to have a whole custom design systems, so we designed our prototypes with a library based on Google Material Design. It gave us speed to prototype things (with preset components), a good level of fidelity, and could ease our technology part as well with the adaption of a framework.
Validating
We invited customers for usability tests to validate our designs. The simple setup included a laptop with a prototype, a task list based on our hypothesis, and an Excel sheet for task accomplishment and notes. Using the success rate, notes, and user responses, we iterated our prototypes.
Shipping and measuring
Despite the implementation of shorter cycles, the limited tech resources and the stratety to hold the release until having more features, delayed the project. And before its release, I started my movements to move to the Netherlands, missing the final results. Determining success in such large projects is challenging, but general methods like CSAT, NPS, and surveys could be useful.
Lessons learned
Fail faster
If we had tested smaller pieces, we would have learned earlier about the things that wouldn't work out.
Structure helps
Even for startups with a super-fast pace, keeping things organized (but flexible) helps us work smarter.
Research more, execute less
At this time, much of my work was UI heavy, but I learned how important it was to keep our customers at heart and to research more about them before doing anything.
Ask more questions
Being a questioner, pushing back on bad decisions, surfacing risks, and taking time to think might sound like slowing down. But it’s flexing the knees to jump higher and in the right direction.
Did you liked it?
