Blacklane Help Center
Redesigning the help center to reduce inbound requests to Blacklane customer service, which currently drive higher costs and longer resolution times for guests. The new design will enable guests to resolve issues through AI-powered chat, minimizing the need to call customer service and improving efficiency.
Discovery & Engagement | September 2025 | Product Designer
iOS, Android & Web
Problem identifcation
My design process for this project
Data & Prioritisation
Solutions
Design
Test & Iterate
Problem
Our guests while loved Blacklane premium services were completely frustrated when they had to reach customer support.
Guests frequently preferred calling support instead of using chat.
Despite having a live chat feature, calls accounted for nearly 37% of all support contacts, resulting in:
15 mins average hold time
Higher Support Cost
Frustration for guests
As a high paying guest for chauffeur driven services, having help on tap of the finger is a part of the whole experience. Our current solution did not offer such experience rather kept them frustrated and issues kept getting bigger.
Data to support business use case
22%
Using Mixpanel analytics and support data, we analyzed 23,000 monthly Help tab visits
initiated a chat
13%
made a phone call
17%
of the users interact with articles to get information
The contact rate (calls + chats) was 34.8%, much higher than desired.
We aimed to reduce this to 18%, mainly by cutting down unnecessary phone calls to also cut down cost.
Approx 3$
cost per resolution via call
Solution milestone 1
1. Chat-First Help UI
Prominent button: “Start chat”
Moved phone contact to a secondary page (“For urgent issues only”)
Contextual entry for current or upcoming rides
Surfaced top 5 Help articles based on engagement data
2. Native Chat Integration
Replaced slow web-view with Intercom’s native iOS SDK
Added persistent threads, push notifications, file sharing
Pre-populated user data (name, booking info, location)
3. Phased Rollout & Testing
A/B test with 25% of users for 4 weeks
Compared chat vs. call behavior between Control & Variant groups
Tracked chat engagement, call rate, and CSAT
To drive business impact and reduce the inbound contact rate, I prioritized a chat-first support strategy. By introducing a sticky chat button and proactively surfacing high-engagement help articles, we empowered customers to resolve common issues through self-service. This successfully shifted volume away from traditional channels while ensuring phone support remained readily available for urgent escalations.
Delivery & Impact
Sticky ‘Start Chat’ button
Start chat in the first 2 weeks reduced calls by approx 50% but also increased chat by 150%.
Users were using start chat more than our expected behaviour but some were just curious to use the feature rather getting a real solution.
Recent ride card
Using recent ride as the entry point to chat as users often used mobile app when the ride was finished or is upcoming.
Recent ride card gave context to our AI enabled chat support.
Users did not have to go through finding booking numbers, it could be fetched from the recent ride data.
Milestone 1
From apps to web
Following a successful rollout on iOS and Android, we scaled the solution to the web platform with a tailored approach. To account for a primarily unauthenticated (logged-out) user base, we omitted phone support entirely and opted against a sticky chat widget.
Instead, we focused heavily on self-service, positioning high-engagement help articles and chat entry points to empower users to resolve queries independently.
Milestone 1.1
Want to make a quick change to the booking?
Partnered with Blacklane’s Support and Post-Booking teams to analyze ticket drivers. Discovered most inquiries involved minor booking modifications or confirmation checks.
Launched a self-service solution by surfacing the top 3 'Quick Edit' options on the Help screen, allowing users to resolve issues instantly without agent intervention.
Resulted in overall reduction in contact rate to 20% from 34% in the first two months.
Milestone 2
Help Center Reskin & Advanced AI Chat Integration
End-to-End Visual Overhaul: Reskinned the entire Help Center interface to improve usability, navigation, and brand alignment.
Transactional AI Enablement: Trained Fin AI to move beyond answering FAQs and execute live database actions, including complex multi-booking modifications.
Bulk-Task Automation: Empowered users to apply global updates across multiple reservations at once (e.g., instantly adding a chauffeur note for a baby seat across three upcoming trips in under 2 minutes).
90% Faster Resolution & 30% Chat Deflection
Upgrading our AI from informational to transactional transformed a 15-minute phone hold into a 2-minute self-service victory—compounding ticket deflection and skyrocketing CSAT.
Milestone 3
Improved AI chatbot
Leading the integration of a contextual AI chat experience within Blacklane’s native apps to streamline the guest support journey. By architecting contextual entry points and training the AI to handle complex, multi-intent booking modifications, I reduced friction and allowed guests to perform simultaneous edits autonomously. To further optimize the UX, I implemented a prompt-based interaction model, enabling guests to resolve issues with minimal manual input.
