Houston Co-Living Pathway — Land Dashboard

Every candidate is scored 0–100 on seven weighted variables. Current weights (best-judgment, labeled so everyone can see): PadSplit demand 22% · rent opportunity-cost 18% · land fit & economics 15% · jobs proximity 13% · transit & walkability 13% · growth & expansion 12% · flood / risk 7% (= 100). Each variable is scored 0–100 on the exact rule below, then combined by these weights. Price is not a variable — it's a gate. Weights are easy to retune.

The seven variables — exactly how each is measured (0–100)

• PadSplit demand = 0.55×occupancy + 0.25×rent + 0.20×supply. Occupancy = (blended occ% [both shared + private bath] − 30) ÷ (95 − 30); Rent = (avg private-bath weekly room rent − $150) ÷ ($260 − $150); Supply = (# of 5+BR listings ÷ 8). Each clamped 0–100. Why the split: occupancy uses all bath types so a ZIP's demand sample isn't too thin, while rent uses private-bath only — that's the product we build and underwrite (falls back to blended rent only if a ZIP has no private-bath data).
• Rent opportunity cost = 20 + 0.55 × weekly savings, where savings = (ZIP 1BR rent × 12 ÷ 52 + $42 utilities) − the co-living room's weekly rent. Rewards lots where a room badly undercuts renting an apartment (plus ~$258 move-in vs ~$1,100–$1,600 for an apartment).
• Jobs proximity = 100 − (miles to nearest of Texas Medical Center / Downtown / UH-TSU − 1) × 13. PadSplit's stated #1 (“proximity to where people work”). Upgrade to transit-time next.
• Transit & walkability = 0.5×Transit Score + 0.3×bus-proximity + 0.2×Walk Score, +5 if rail ≤1 mi. Bus-proximity: 0.0 mi = 100, 0.1 = 95, 0.2 = 85, 0.3 = 70, ≤0.5 = 50, else 30.
• Growth & expansion = gentrification stage → early ≈ 85, early/mid ≈ 75, mid ≈ 60, mid-late ≈ 55, late ≈ 50; nudged by new-construction / permits & motivated-seller signals.
• Land fit & economics = f($/fourplex): ≤$150k → 100; $150–250k → 100→70; $250–300k → 70→45; over $300k drops steeply.
• Flood / risk = X-Unshaded 90 · unknown 60 · X-Shaded 50 · AE/high-risk lower. Deferred — harvested from listings; won't hang us up.

The land gate — “does it even work?” (pass/fail)

• Lot ≥10,000 sf per fourplex (structures OK — demo assumed). Capacity = lot sf ÷ 10,000.
• ≤$300k per fourplex (ideal ≤$250k) = asking ÷ fourplexes. All 16 below clear this; Fulton and the $300k+/fourplex premium block were screened out.

Buy-box qualifiers — straight from PadSplit (Audrey)

• Proximity to jobs is #1, and “stay within ~25 miles of the metro.” → scored (Jobs); every inner-Houston lot clears the 25-mi rule.
• “Assume tenants rely on public transportation.” → scored (Transit); a bus stop ≤0.5 mi matters.
• Parking must work. → per-lot diligence flag (can't fully automate from data).
• Never buy in an HOA / a deed restriction that bans co-living. → hard red-flag.
• “Don't be the highest-priced rooms in the area.” → captured by opportunity-cost + the ZIP's rent level.
• ~90% occupancy is the health line; avoid oversaturation. → the demand score; ZIPs under ~70% get a “soft demand” flag on the card.
• No PadSplit in the ZIP isn't a “no.” — “if it follows the buy box, be the pioneer.” We don't hard-exclude no-data ZIPs; we just can't score their demand yet.

Sources

• PadSplit Houston external report (06/12/2026, 2,202 rooms) · WalkScore (Jul 2026) · RentCafe/RentHop/Apartments.com · Zillow/HAR MLS (lot sizes) · PadSplit buy-box interviews (Audrey).