Klue — Instant Battlecards (2023)

AI-first feature, "The Compete Composer"

Exploring LLM's potential to reduce customer onboarding by creating push-button battlecards that generate competitive intelligence in minutes instead of hours or days

Role & Deliverables

MVP Product Design + UX: Led the design of prototypes from concept to testing for enterprise beta users, validating AI-powered content generation while learning how to integrate emerging AI capabilities into established workflows.

They say it takes a village: This initiative connected me with many distinguished and talented colleagues at Klue—an exciting time of collaboration that extended well beyond the project itself.

"Push a button, get a Battlecard... Can we be meaningfully better than ChatGPT?"

— User Experience challenge

Context & Problem

• Company: Klue, a G2 industry-leading SaaS platform for competitive intelligence and market analysis, serving enterprise sales teams at companies like Dell
• Core Challenge: Manual competitive intelligence curation was creating long customer onboarding times and churn risk, while ChatGPT posed an existential threat to the core business model

Goal & Strategy

• Business Impact: Dramatically reduce customer time-to-value and churn risk by automating competitive battlecard creation for faster seller enablement
• Product Strategy: Leverage GenAI to automate the "good enough" lower-value tasks while elevating curators to higher-value strategic work, proving superior value over standalone ChatGPT
• Key Constraints: Had to overcome AI trust issues with enterprise users, work within existing data pipeline limitations, and deliver measurably better results than $20/month ChatGPT alternatives

Approach & Execution

• Rapid Prototyping & Validation: Built functional prototypes using real customer data and ChatGPT APIs via Azure, conducted extensive user testing with 8+ beta customers to validate AI-generated content quality and user trust
• Cross-Functional AI Integration: Partnered closely with data science team on RAG systems, ML engineers on prompt engineering, and platform teams to integrate AI microservices while maintaining data quality and reducing hallucinations
• User-Centered AI Design: Developed novel UX patterns for AI transparency (source verification, edit capabilities) and conducted research to understand how competitive intelligence professionals interact with AI-generated content
• Data-Driven Insights: Analyzed thousands of existing customer battlecards to identify content patterns, leveraged G2 reviews and buyer intelligence data to create AI-powered strengths/weaknesses analysis

Outcome & Impact

• Massive Time Reduction: Customer Success achieved new benchmark of <48 hours to rollout system access with 3 competitors (down from 90-120 days), with battlecard generation reduced from 90 days to 90 seconds
• User Adoption & Satisfaction: Beta users reported significant time savings for lower-tier competitors, with curators freed up for higher-value strategic tasks. Users recognized potential for value over ChatGPT alternatives
• Business Impact: Feature was spun off into dedicated team with additional designer, demonstrating company's doubled-down investment. Positioned Klue as AI leader in competitive intelligence space
• Platform Innovation: Created reusable AI microservice architecture and improved text editing experience that was promoted across entire platform, establishing foundation for future AI integrations

Key Features

• Push-Button Battlecard Generation: AI-powered system that creates comprehensive competitive battlecards in 90 seconds using proprietary data blend and advanced prompt engineering
• Intelligent Source Verification: Built-in transparency features allowing users to verify AI-generated content sources and edit misinformation, addressing enterprise trust concerns
• Smart Content Import: File management system that ingests existing competitive content (decks, docs) and feeds it into RAG system for more accurate AI responses
• Multi-Platform Distribution: Seamless integration allowing AI-generated battlecards to be distributed across sales stack (Slack, MS Teams, Salesforce, etc.)

Compete Composer was an early foray into a fully AI-automated competitive intelligence tool that could generate enterprise-grade battlecards instantly while maintaining the data quality and trust standards required by Fortune 500 sales teams.

— Compete Composer innovation

Reflection & Next Steps

• Key Learnings: AI UX design requires realistic data for effective validation - ChatGPT mockups and Figma prototypes were insufficient compared to functional prototypes with real customer data. Trust remains a critical factor in enterprise AI adoption.
• Future Positioning: This experience established deep expertise in AI product design, cross-functional collaboration in AI development, and understanding of enterprise AI adoption challenges - directly applicable to future AI product leadership roles
• Evolution: The success opened doors for expanded AI exploration across the platform, with ongoing work on AI insight correction and source data inspection for skeptical enterprise customers

Surface Design Possibilities

I examined various design directions and appreciated the breadth of opportunities available.

Flipbook Prototype