Airblox AI Onboarding with CRM Integration
Onboarding was a wall, not a welcome
Before the redesign, onboarding was blocking more than it unlocked. To deliver tailored quoting, Airblox needed to gather 20+ critical inputs with minimal friction.
Lack of structured user data prevented personalization
Without structured inputs, users got a generic, irrelevant experience.
First impressions broke brand trust
The onboarding felt clunky, undermining the premium feel Airblox needed.
Lack of role-based personalization
Forwarders, GSAs, and carriers weren’t distinguished in setup—hurting personalization.
CRM data existed, but was not utilized
Warm referral data existed—but wasn’t prefilled, wasting a key opportunity for delight.
Users couldn’t find their team
Company discovery was disjointed, isolating users from their existing orgs.
Payment fields came too early.
Asking for sensitive info before users saw value created drop-off and suspicion.
Prioritize onboarding essentials with a card sort
We began by mapping 20+ required inputs across user, company, and compliance needs. A card sort helped identify which details were essential to personalization—and which could be deferred or inferred.
Analyze UX risks of using AI in onboarding
Using AI was a business requirement—but I was concerned we would lose the user so I presented my concerns to business with a workaround. My solution: let AI act as the greeter, not the form. The onboarding flow was built as structured UI components dynamically sequenced by AI based on context and CRM data. This preserved clarity, reduced friction, and showcased our technical sophistication without sacrificing usability. The result is a novel hybrid onboarding paradigm—AI-personalized, but never unstructured.
Mapped out two flow story arcs: warm referrals & cold signups
I rebuilt the onboarding journey from scratch—anchored in psychological principles. The spark effect, Aha! moments, and completion reinforcement.
"We've met before" auto-prefilled company and user data from CRM data via warm referral links.
A built-in “Aha moment!”, our AI dynamically assembles the journey using CRM data, acting as a live demo of our product’s intelligence.
Persona-based branching dynamically routes forwarders, carriers, and GSAs into the right verification path.
WhatsApp opt-in delivered an AI-powered quoting assistant, instantly showing value.
Incomplete profile captures the inessentials, and payment is taken in context of first transaction
Through instant KYC, AI sequencing, and WhatsApp handoff — the onboarding effectively drives time-to-value to zero. The flow preloads CRM data, verifies credentials in-stream (IATA/IAC), dynamically adapts to user type, and ends with a WhatsApp opt-in that delivers immediate quoting value. Users go from sign-up to operational in minutes—with no waiting, no guessing.
The <hi-fi prototype> delivers value before onboarding ends
The flow verified credentials, surfaced lane recommendations, and launched WhatsApp quoting—turning setup into conversion.
Funnel visualization in Heap showed a 49.50% conversion rate on warm referrals 🙌
Even with warm referrals, we saw natural drop-off across stages — but over 49% completed onboarding and 40% fully personalized their profile.
Business outcome
We piloted the AI-assembled, CRM-integrated onboarding flow as a sales enablement tool for the launch of Airblox RFQ and RFP in Q2 of 2025. Delivered via warm referrals, the pilot exceeded expectations—accelerating onboarding, boosting engagement, and proving out product-market fit with high-intent users.
Conversion rate from CRM match to onboarding complete
Minutes for time-to-value, via WhatsApp opt-in and instant KYC
High-impact users fully onboarded in Q2 2025
Users became WhatsApp conversions
Up next
Continue exploring my work below.
.jpg)
.jpg)
.jpg)
