In this episode of The Future of Consumer Marketing, host Andres Figueira interviews Xiao Zhang, Co-founder and CEO of Collov AI, a spatial AI platform revolutionizing the real estate and home renovation industry. Zhang’s company transforms the traditional staging process from a several-thousand-dollar, multi-day operation into a seconds-long, sub-$1 automated experience. Starting as a Stanford PhD in Applied Physics, Zhang pivoted from human-in-the-loop design services to fully automated AI generation, surviving near-bankruptcy and building a platform now serving over 10,000 real estate agents and hundreds of enterprise clients. His journey illustrates how technical founders can build consumer-facing platforms by maintaining relentless customer focus while leveraging breakthrough AI capabilities.

Topics Discussed:

• Transforming traditional real estate staging from manual to AI-automated processes

• Building domain-specific AI models versus general-purpose solutions

• Surviving startup pivots during the pre-ChatGPT era of AI development

• Scaling customer feedback loops to drive product iteration

• Balancing automation with human customization in consumer platforms

• Managing lean teams with AI-savvy talent in the generative AI era

Lessons For Consumer Marketers:

Simplify Complex Technology Into Single-Action Workflows

Instead of explaining “spatial design intelligence” or technical AI concepts to real estate agents aged 40-60, Collov AI created a three-step process: upload photo, select style, click generate. This streamlined approach eliminated technical jargon and reduced friction to drive adoption among non-technical users who needed results, not explanations.

Use Model Performance as Your Primary Growth Lever

Xiao identified that in generative AI startups, model performance directly correlates with growth rate. After fixing their core issue – maintaining room structure during generation – they saw dramatic improvements in user adoption and conversion rates. Similar to Cursor’s growth to $500M ARR by perfecting code completion, technical excellence became their marketing advantage.

Scale Customer Feedback Collection as a Competitive Moat

Collov AI institutionalized daily conversations with 5-10 customers, systematically categorizing feedback to drive product improvements. This created a feedback-to-iteration cycle that continuously improved their AI model’s applicability across different customer needs, turning customer insights into a sustainable competitive advantage.

Pivot Business Models While Preserving Core Technology

When their human-in-the-loop model created negative unit economics (more clients = more losses), Xiao maintained the underlying spatial AI technology while completely restructuring the business model. The 2022-2023 transition from contractor-managed services to fully automated AI generation demonstrates how technical startups can preserve R&D investments while fundamentally changing go-to-market strategy.

Hire for AI-Native Talent Over Team Size

Rather than scaling headcount, Collov AI maintains a lean team where each member understands both AI capabilities and customer requirements. This approach enables rapid feature development where team members can fine-tune models or engineer prompts to meet customer demands without traditional departmental handoffs.

Maintain Customer-Centricity Despite Technical Capabilities

Xiao emphasizes that regardless of AI model sophistication, customer satisfaction and service quality remain paramount. The most advanced technology fails without proper productization and user experience design that serves the target audience’s actual needs rather than showcasing technical capabilities.