Launching new feature for B2B SaaS and growing MRR in 2 months
This case focuses on user research. Alongside examples of how I did quantitative research and interactive prototypes, I convey the story of how I launched a new feature within 2 months and increased MRR at a B2B SaaS company.
- my role: product designer
- year: 2024
- timeline: 2 months
- team: me, 1 founder, 2 full-stack engineers
B2B SaaS for second-hand sellers
Imagine having 1000s items to sell e.g. vintage/preloved clothes, shoes or antiques. Manually listing them on marketplaces is a nightmare. That's where our product comes in to automate the process of cross-listing and auto-delisting.
We launched 3 years ago, have hundreds of paying customers, and 2,500,000+ listings were created using our product.
Here is our landing
People unsubscribe because we don't support important workflow
There is a way to get more sales: delete and list again stagnant items from your inventory. It is recommended that your listings must be not older than 3 months. We didn't support this workflow and it caused a lot of frustration among our users.
Let's confirm the problem
To confirm the problem I pulled two methods from my "user research" toolkit: questionnaire and in-depth user interviews. The former is a quick way to confirm the problem, the latter is a way to understand the problem better. I launched an email campaign. Users received a questionnare with open ended questions.
I received about 30 responses
Problem confirmed but people want different solutions
Speaking to users led me to discover that different users have different requirements. Some want to end items on one marketplace only, some want to do it on multiple platforms in one go.
Put simply here are workflows people have
Different requirements: all at once vs one by one
Time to test potential solutions
I had 3 interactive prototypes, 5 scheduled user interviews and 1 week to choose the best option. Not that I needed all that to launch the feature, but when you get locked in serious user research, the tendency is to push it as far as you can. Solutions that I tested but end up not using
Rejected option #1: relisting by marketplace
Rejected option #2: multi marketplaces boosting
These option didn't go any further because they were too complicated for users and engineering-wise. After presenting my findings to the team lead and founder, we decided to go with the simplest version: the one with less clicks and allowing to work with one marketplace at a time.
Final design
Launch, failure and comeback
Preparations for the launch included: handoff to engineers, drafting email campaign, and educating customer support about the new feature.
We launched the feature thinking it will be a success. But it wasn't. At least not at the beginning.
On the day of the launch, we started to receive phone calls from users. What happend is while syncing our system and their shops, for some users we emptied their shops. Created emergency hotline and fixed it within couple of days.
To drive users into the new feature I came up with an idea of a new type of notification. If a user has any items older than 90 days, a yellow clock icon will automatically pop up. When clicking on a notification, a tooltip with an explanation and link to the FAQ will appear.
In the end, customers found the new feature very useful and we saw an increase in MRR.
Lessons learned
- Start recruiting users for interviews as soon as possible. It took me 2 weeks to complete the interviews because of users' tight schedules. Next time I will start scheduling interviews the same moment I start working on prototypes. It works especially for B2B products and/or if you have a small user base.
- Manage users' expectations from the beginning. The first interview went a bit off track because the user expected my prototype to be fully functional and they focused on the wrong thing. To avoid this, it's essential to communicate the task and constraints from the start. For the rest of the interviews, I followed this approach, and everything went smoothly.