The NodeScript visual workflow builder in action
Case Study
/
NodeScript
Role
Lead Product Designer
Timeline
2023-2025
Team
4 Engineers, 1 PM
Visual Programming for Backend Automation
NodeScript is a node-based tool for building backend logic visually—drag nodes, connect them, expose as API. Think n8n or Pipedream. Internally, it powers 300 million monthly automations for enterprise clients including Google and Trivago.
I spent 2.5 years as lead designer, shaping the product experience from structure to onboarding, right as ChatGPT changed what developers expected from tools like ours.
01
/
Impact
0
0
m
Monthly Automations
Powered by the platform for enterprise clients
0
0
m
Annual bookings in USD
USD Annual bookings facilitated for enterprise clients
02
/
The Approach
Mapping the Funnel
User journey from landing to payment, with stages, metrics, and design goals for each step
I mapped the user journey from landing page to paid subscription, defining stages, metrics, and design goals for each step.
The framework distinguished between someone who just signed up versus someone who'd built their first graph, versus a power user creating custom nodes. Each stage had different needs and different design challenges.
This became the foundation for how we talked about growth as a team.
03
/
THE DATA
What the Metrics Hid
The Grafana dashboard showed steady growth. New users activating daily. Everything looked fine.
On the surface, things looked healthy. We had 231 activated users. New signups coming in daily. The PM was creating viral TikTok videos that drove traffic. People were finding us, signing up, completing onboarding.
Cohort analysis told a different story. Every cohort dropped to near-zero.
Then I looked at retention.
"Our 'healthy' metrics were hiding a crisis. 96% of users never returned after week one."
Every cohort dropped to near-zero by week four. The "active users" were mostly internal team members and a handful of outliers. We weren't losing people at signup or onboarding—we were losing them after.
04
/
THE RESEARCH
The Uncomfortable Discovery
Snapshot of the workshop I organised to come up with onboarding improvement ideas.
I ran interviews with recent signups, including new colleagues who'd just learned the tool. I organised workshops to synthesise findings and generate solutions.
What we found was uncomfortable. Colleagues could learn NodeScript in a single day. The tutorials worked. The interface was learnable.
"The problem wasn't usability. It was utility."
We'd built generic backend tooling with infinite use cases—which meant we couldn't optimise for anyone specific. Developers would look at it, nod, maybe build one graph... and never come back. They didn't have a burning problem that NodeScript solved better than alternatives.
05
/
WHAT WE SHIPPED
Designing for Activation
With the retention problem clear, we shifted focus to activation. The hypothesis: if we could get users to value faster, maybe they'd stick around.
SHIPPED
Pre-built templates for common use cases
Guided onboarding flows with first-run experiences
Starter graphs so people could see value immediately
Quick start panel and documentation improvements
New users land in a workspace pre-populated with starter graphs and guided onboarding.
05
/
WHAT Happened
The Ground Shifted
The cohort data confirmed what the business already knew: we had two paying users. One was a friend.
Meanwhile, AI was disrupting the whole company, not just NodeScript.
"We started in a different era. ChatGPT didn't exist yet."
— Boris Okunskiy, CTO
"We got disrupted by vibe coding."
— Marcus Greenwood, CEO
Suddenly vibe coding changed what developers expected from tools like ours. Visual programming started to look like the old world.
The company refocused on B2B, where UBIO's core business had always been. NodeScript continues to power internal operations. The external growth play wound down.
We'd built the right product for the wrong era.
"The thinking process that Sjors delivered behind all of this adds the most value. Design is identifying the problems that everyone seems to be ignoring, and then think structurally about what can be done about them."
— Boris Okunskiy
/
