The homeowner funnel, from first impression through signed agreement and activated listing, is the highest-leverage product surface in the business. Most growth product roles are about optimizing what already exists: faster page loads, shorter forms, better copy. This role is about building something structurally different.

Belong's homeowner acquisition funnel is being rebuilt as an AI-native system: conversational intake powered by LLMs, personalized onboarding that adapts dynamically to each homeowner's financial profile, predictive scoring that routes the right lead to the right moment in the Advisor workflow, and agentic follow-up that replaces manual sequences with intelligent, context-aware outreach.

The target is a funnel that learns, where every interaction generates signal that makes the next interaction more likely to convert. As Product Owner, Growth, you are the person building that system. You own the homeowner acquisition and activation funnel end to end, from first contact to listed home.

Responsibilities ================

AI-native intake and qualification layer

The first interaction a homeowner has with Belong, whether via belonghome.com, a paid channel, or a referral, is where trust is either established or lost. You will build conversational intake flows powered by LLMs that qualify, capture, and begin converting leads in real time.

These are not chatbots with decision trees. They are context-aware systems that understand the difference between a cashflow-positive homeowner who wants yield optimization and a cashflow-negative homeowner who needs a path to profitability, and adapt the conversation, the framing, and the call-to-action accordingly.

Personalized onboarding and trust architecture

A homeowner considering Belong is anxious. They are considering handing over their most valuable asset to a platform they found online. Conversion at this stage is not a UX problem. It is a trust architecture problem.

You will design onboarding sequences that adapt dynamically based on homeowner attributes: property type, cashflow profile, prior rental history, risk signals, and behavioral signals from in-session activity.

You will use LLMs to generate personalized content, market analyses, improvement ROI estimates, comparable listings, that makes the value proposition concrete and specific to their home, not generic.

Predictive lead scoring and Advisor routing

Belong's Advisors are the trust-critical human touchpoint in the homeowner funnel. Their time is finite and high-value. You will build the predictive infrastructure that scores every lead on conversion likelihood, property quality, and fit with Belong's ICP, and routes leads to Advisors with the context they need to have the right conversation immediately.

You will work with data science to train and evaluate these models, with RevOps to deploy them into the Salesforce workflow, and with Sales leadership to validate signal quality against actual close rates.

Agentic follow-up and nurture sequences

Most leads do not convert on the first contact. Today, nurture is a sequence of templated emails. The target state is an AI agent that monitors lead behavior, page views, document opens, return visits, session signals, and generates contextually appropriate, personalized outreach at the right moment, with the right frame, without a human initiating every touchpoint.

You will define the agent's decision logic, build the context retrieval pipeline, instrument the output quality, and iterate on conversion impact week over week.

Funnel instrumentation and the learning loop

An AI-native funnel without rigorous instrumentation is a black box. You will build the measurement architecture that makes every conversion decision traceable: which intake flow variant produced the lead, which scoring model routed it, which agent-generated touchpoint influenced the next action, which Advisor framing closed it.

You will design the feedback loops that push conversion signal back into model evaluation, prompt improvement, and scoring recalibration. The funnel gets smarter every week or it is not an AI-native funnel.

The activation gap: agreement to listed home

Signing the agreement is not growth. A listed home is growth. The conversion from signed agreement to activated listing is a product problem with high leverage: homeowners who do not complete inspection scheduling, who abandon the improvement process, or who sit in the pipeline without a live listing represent real lost revenue.

You will own the product layer that closes this gap, including AI-assisted improvement planning, proactive homeowner communication anchored to their cashflow profile, and predictive identification of homeowners at risk of churning before listing.

The AI Stack You Will Work With ===========================

• LLM-powered conversational intake with real-time lead qualification and cashflow profile detection
• Personalized content generation using property-level market data, comparable listings, and improvement ROI modeling
• Predictive lead scoring models trained on conversion, property quality, and ICP signals
• Agentic follow-up workflows with behavioral trigger logic and context-aware generation
• Retrieval-augmented generation for Advisor preparation: the right context, surfaced at the right moment before the call
• A/B testing infrastructure applied to AI-generated content variants, not just static copy

What Success Looks Like ====================== 90 days: The funnel is fully instrumented from first click to activated listing with conversion rates and drop-off points visible at each stage. An AI-assisted intake flow is in production and being tested against the baseline.

6 months: Lead-to-listing conversion is measurably above baseline. AI is integrated at a minimum of 3 funnel touchpoints with documented conversion impact per touchpoint. Advisor routing is scored, and the correlation between score and close rate is being tracked.

Year 1: The majority of homeowner outreach between first contact and agreement signing is AI-generated, with human Advisors focusing exclusively on trust-critical call moments. CAC on the supply side is trending down. Time-to-activation is compressing quarter over quarter.

Example KPIs You Will Be Held To ==================================

• Lead-to-listing conversion rate (the primary number)
• Cost per activated listing
• Time from first contact to listing live
• AI-assisted funnel touchpoint conversion impact, measured per touchpoint
• Advisor routing accuracy: scored lead close rate vs. unscored baseline
• Experiment velocity: instrumented tests shipped per month
• Homeowner CSAT at onboarding and inspection phases (the constraint: conversion gains cannot come at experience cost)

Who You Are ============ AI systems builder, not AI enthusiast. You have shipped LLM-powered product features in production. You understand prompt engineering, retrieval quality, latency tradeoffs, output evaluation, and model feedback loops. You think about AI systems the way a statistician thinks about models: with explicit assumptions, known failure modes

XML job scraping automation by YubHub