AI-driven onboarding: 67% faster

Company
Factorial
Year
2026
Status
Shipped

The problem

Factorial has historically been a sales-led business, with onboarding relying 100% on human specialists. As the company scaled, this model became increasingly difficult to sustain.

The onboarding model was becoming the bottleneck to growth.

45+ days
to onboard one customer
6h
specialist work per customer
Manual Excel
checklist per client, to track implementation

The signal

I immersed myself in the onboarding process, joining specialist sessions and observing customers firsthand to understand where the experience was breaking.

  1. 01

    Customers lacked visibility

    Customers didn't know what was happening, what they needed to do, or when onboarding would be complete.

  2. 02

    Working rules were difficult to translate

    Specialists had to turn each company's unique working rules into complex system configurations.

  3. 03

    The process relied on specialist intervention

    Customers couldn't complete onboarding independently, creating a dependency on specialist sessions throughout the process.

The opportunity

We needed a faster onboarding process — and a different operating model.

Before
CustomerSpecialistSystemsSpecialistCustomer
After
CustomerAI orchestration + onboarding skillAI agentsSpecialist when needed
The shift: from a relay through specialists to an orchestration the customer drives.

Exploration

I launched our first MVP to test the new onboarding experience with customers.

The MVP explored two complementary ideas: making progress visible and using AI to simplify complex configuration.

01

Get Started

I explored a checklist-based experience to help customers understand what they needed to do and how far they were through onboarding.

Factorial's first Get started page: a progress bar reading 0 of 10 modules completed, a Core card whose first task — how to create a legal entity — comes with a guide video and five more setup tasks under it, then a run of locked modules below the line 'Complete these mandatory steps first to unlock the rest.'
Get Started — the first MVP was based on a simple setup checklist to configure your account, connected to the help center and supporting videos.
02

Voice AI

Factorial's configuration was too complex to cover with a fixed onboarding flow. I explored AI as a way to dynamically understand what customers were trying to configure and guide them through the UI.

Voice AI — the first MVP was a voice framework connected with our AI system that reads the interface and guides the user through a red cursor on screen.

What we learned

The checklist worked. Voice didn't.

Get Started

  • Strong engagement

Voice AI

  • Low engagement
  • High error rate

Testing showed the guidance system worked — the AI understood tasks and directed users correctly. The issue was the voice interface, not the guidance itself.

Iteration

I kept what worked — and made it scalable.

The second MVP combined the strongest parts of the first experiment:

01

AI-guided system support

I kept the pointer system from Voice AI, but the voice itself was replaced by an AI system interface that reads the UI and guides customers with an on-screen cursor.

The system is connected with Factorial AI “One” and developed to guide customers through the process.
02

A simpler Get Started

I simplified Get Started based on what customers actually used. I removed low-engagement videos and the Help Center link, and moved optional tasks later — making it easier to scale globally.

The simplified Get started page: the left nav beside a progress bar reading 0 of 10 modules completed, a Basic configuration card with Add people expanded over five more setup tasks, and Module configuration locked below it.
New Get Started screen with a few steps displayed, providing clearer guidance and better understanding of the process.
03

An onboarding intro animation

I introduced a short animation at the start of onboarding to create a “wow” moment and make the first interaction feel more approachable.

The onboarding intro animation is the first thing users see after logging in, creating a sense of wow effect and a strong first impression.
04

A personalized onboarding start with AI

I used Factorial's AI to customize each customer's onboarding, tailoring steps from their first interaction and introducing the AI system early.

A co-creation experience with Factorial's AI system tailors each customer's onboarding path.

Results

67%
faster setup. 45d → 15d
~€ 1.5M
projected onboarding savings for the year
4.9
customer satisfaction with the onboarding

The impact was clear: enabling Factorial to scale strategically, onboarding more customers without increasing specialist headcount while keeping customers highly satisfied.