Build vs Buy — PropTech

What this means for PropTech

• Listing data quality is structurally broken — MLS feeds vary by market, photos are mislabeled, square footage is contested, and pricing fields update on different cadences across the 600+ MLSs the platform consumes.
• Transaction velocity is throttled by paper-and-PDF closing — title, escrow, lender, and brokerage workflows still pass documents around in email when the consumer-facing product promises a tap-to-buy experience.
• iBuyer/instant-offer pricing models are under post-Zillow-Offers scrutiny — the algorithm needs explainable pricing, defensible market assumptions, and operational discipline that survives the next housing-cycle inflection.

Where it pays to act

• Listing data quality and enrichment — AI on photo classification, room labeling, square-footage validation, and cross-source reconciliation that finally produces a defensible single source of truth per property.
• Automated valuation models (AVMs) with explainable pricing — modern AVMs that beat the legacy comp-based models, with explainability layers that survive regulator and lender scrutiny.