A multi-source market analysis of seed-stage valuations for vertical AI software, the multifamily-AI sub-category specifically, and the founder/team premium evidence. Companion to the PropFlow seed deck. This document is reference material for cap-rate conversations — public market data on where comparable rounds have priced, not a self-pricing argument.
The headline software seed median sits in a $18M–$24M post-money band across credible sources, with notable methodology-driven dispersion. Carta runs high (cap-table-software customer base; software-heavy); PitchBook runs lower (broader sector mix). Both are reporting all-time highs in late 2025 / early 2026.
| Source | Date | Cohort | Median post-money cap |
|---|---|---|---|
| Carta — State of Pre-Seed[1] | Q4 2025 (ATH) | US software seed, all raise tiers | $24.0M |
| Carta — Top 15 benchmarks[2] | Full-year 2025 | US software seed, $4M median raise | $20.0M |
| AngelList — State of Venture H1[3] | H1 2025 | US seed, $4M raise typical | ~$24.0M |
| PitchBook–NVCA Venture Monitor[4] | Q3 2025 | All US seed (broader sector mix) | ~$19.0M |
| Metal.so (cites Carta + PitchBook)[5] | 2025 | US SaaS seed specifically | $17–20M |
| Defensible center | 2025–2026 | US vertical software seed, $2–4M raise | ~$20M |
Three caveats that move a round within the band. Geo: 66% of top-decile seed valuations cluster in SF Bay / NYC[2] — non-coastal companies typically price below the published median. Bifurcation: top 1% of seed companies captured ~⅓ of all 2025 seed capital; bottom 50% got ~7%[6]. The "median" disguises a barbell. Dilution norms: priced-seed median dilution is now ~20%, down from 23% in 2019[7]. Competitive software seeds run 12–15%; 18%+ signals weakness.
The widely-quoted AI seed premium is a category-wide median dominated by foundation-model labs and AI-native infrastructure. Vertical SaaS-with-AI features (the relevant cohort for PropTech) captures a much smaller and compressing premium. Three sub-segments, three different premiums:
Source triangulation. SVB State of Markets 2026 puts AI seed premium at +10% (the lowest stage; widens to +222% at Series D)[10]. PitchBook Q4 2025 ecommerce-vertical-SaaS analysis shows seed AI premium of +44%, compressing from +86% in 2024[11]. Carta-implied (AI median $19M vs. non-AI $13M) lands at ~+27%[12]. The honest range for vertical SaaS-with-AI is +15–30%; treat anything above as a strategic-lead premium, not a category premium.
Multifamily-AI seeds are bimodal in 2025–2026: a generalist-led tier at $3–6M raises ($13–25M implied caps), and a specialist/branded-lead tier at $9–10M raises ($40–55M implied caps). Where caps were not disclosed, ranges are computed at the typical 18–25% seed dilution band.
Agentic AI for PM — leasing, renewals, maintenance, tenant comms across PMS/email/SMS. 5K+ units in IE/UK.
International near-twin of PropFlow scope. Pre-seed floor.
AI assistant "Emma" for residential property management automation. NYC-based.
Closest stage match by raise size. Generalist-led tier floor.
Agentic tenant comms automating 90%+ of inbound across STR + LTR. 20K+ properties.
YC pedigree + agentic-AI thesis. Adjacent vertical (STR-first).
AI utility billing for multifamily; up to 52% cost savings. Founder ex-Modern Message.
Highest-conviction multifamily seed of 2025. The "RET specialist premium" benchmark.
Real-time AI coaching for human leasing agents. "Humans beat bots" thesis.
Highest-profile 2026 seed. ICONIQ + Asymmetric is the new vertical-PropTech-AI seed syndicate.
AI copilot generating CRE deal decks/financial models. 5 of top 10 US brokerages.
YC + brokerage-adjacent. Generalist-tier comp.
AI front-office for home-services contractors. SMS/CRM automation.
MetaProp-participating PropTech-adjacent comp.
AI-powered FSBO / commission-free home selling.
Fifth Wall-led — second category-specialist benchmark alongside RET.
"AI-native operating partner for the built world" — agentic AI + CRE domain experts for real estate development.
Same ICONIQ + Asymmetric syndicate as Grotto. Pattern signal: this pairing is the new vertical-PropTech-AI seed lead.
Agentic AI for CRE asset management. 35 countries, 2B+ sq ft.
Cleanest disclosed Series A in the category. Trajectory marker.
Agentic AI for housing + healthcare. NMHC Top 30 enterprise customers. $100M+ ARR.
Category ceiling. Distinct buyer (enterprise-only); mid-market is structurally separate.
| Tier | Lead profile | Raise size | Implied cap range |
|---|---|---|---|
| Pre-seed floor | Pitchdrive (intl) · regional accelerator | $1.4–5M | $5–20M |
| Generalist-led seed | TenOneTen, Pi Labs, Susa, Wilshire Lane, Pathlight | $3–6M | $13–28M |
| Category-specialist seed | RET Ventures, ICONIQ, Fifth Wall, MetaProp | $6–10M | $26–55M |
| Branded-founder pre-seed | Ribbit + Afterpay co-founder team (Breezy) | $10M (oversubscribed) | $40–55M |
| Series A | Resolve Growth, Maverick, Notable, Haymaker | $12–22M | $55–100M |
Two reads. The generalist-led seed median for multifamily-AI is ~$20M post; the specialist-led tier roughly doubles it (~$45M). The $25M gap between tiers is paid by the lead investor's category conviction — not by anything intrinsic to the company. Without a RET/ICONIQ/Fifth Wall lead, a PropTech-AI seed prices in the lower tier; with one, the upper. The mid-market vs. enterprise split is the second axis: EliseAI ($2.2B Series E) and Cambio ($100M Series A) are both NMHC-Top-30-only — a distinct buyer profile from mid-market plays.
Founder/team profile is the single most consistent driver of seed valuation premium that the data supports. The First Round 10-Year Project — the most-cited primary VC dataset on this question — quantifies the multi-founder premium directly; and a recent ACA n=3,774 dataset confirms that working product alone does NOT command a premium (it's a qualifier, not a multiplier).
The cleanest current precedent for what branded operators + crossover-tech-investor leads pay at the earliest stage in residential real-estate AI. $10M pre-seed (oversubscribed) implies a $40–55M cap — pre-seed pricing matching the upper end of the multifamily-AI seed comp set. The signal: founder/team profile and lead-investor identity are the dominant inputs to early-stage PropTech-AI valuation. Branded-exit founders + crossover lead = ceiling tier; ex-Big-Tech engineering leadership + generalist lead = upper-mid tier; first-time founder + smaller raise = the floor of the band shown in §03.
How these stack. Premiums are correlated, not additive. The defensible stack for a comparable team profile (multi-founder, ex-Big-Tech engineering leadership, domain operator co-founder, technical AI co-founder) is +15% to +30% over the data-driven baseline. The Breezy precedent (Afterpay co-founders + OpenAI execs + crossover lead) sits well above this band — it's the ceiling tier, not the comparable. Without category-specialist or crossover-investor conviction in the lead position, the stack lands in the +15–30% range, not the +30–60% range that brand-name founder + brand-name lead profiles command.
Two independent ways to compose the inputs from §01–§04 produce the same observed band for a multifamily-AI seed in 2025–2026: stacking the multi-source software baseline with the vertical-SaaS AI premium, the team premium, and the stage discount; or starting from the multifamily-AI comp set median and adjusting for the lead-investor and team variables. Both converge at $19–24M for the generalist-led tier with a strong (but not brand-name) team profile, with the broader category band running from a $15–18M floor (no premiums) to a $30–40M ceiling (specialist-led).
| Step | Adjustment | Source | Running cap |
|---|---|---|---|
| Software seed P50, $2–4M raise | baseline | Carta + AngelList + PitchBook (§01) | $20.0M |
| + Vertical SaaS AI premium | ×1.22 (midpoint) | PitchBook + SVB + Carta (§02) | $24.4M |
| + Team premium (multi-founder + ex-Big-Tech + operator) | ×1.22 (midpoint of +15–30% stack) | First Round + NFX (§04) | $29.8M |
| – Pre-traction / 2-month stage discount | ×0.75 (midpoint of –25 to –40%) | Flowjam + Carta P25 floor | $22.3M |
| – Geographic / non-coastal adjustment | ×0.92 (–8%, midpoint) | Carta top-decile geo cluster | $20.5M |
| Composition 1 result | Category midpoint, generalist-led tier with strong (non-brand-name) team | ~$20.5M | |
| Step | Adjustment | Reasoning | Running cap |
|---|---|---|---|
| Generalist-led multifamily-AI seed median | baseline | Domos + Conduit + Henry avg, generalist-led tier (§03) | $18.5M |
| + Strong-team premium vs. comp leads | +$3M | Ex-Big-Tech engineering leadership + domain operator co-founder > most comp set founders | $21.5M |
| + Working-product qualifier | +$2M | Per ACA n=3,774 — product alone doesn't drive valuation, but it qualifies the team premium to apply | $23.5M |
| – Generalist (not specialist) lead | –$3M | Without RET / ICONIQ / Fifth Wall conviction signal | $20.5M |
| Composition 2 result | Comp-set adjusted, strong-team generalist-led seed | ~$20.5M | |
What moves a round within these bands. A signed term sheet from a category specialist (RET, ICONIQ, Fifth Wall, MetaProp) shifts pricing toward the ceiling. Additional paying customers beyond pilot stage and senior team additions from category leaders (EliseAI / AppFolio / Greystar alums) push the team-premium stack higher. Conversely, extended fundraises without lead conviction, macroeconomic seed-pricing compression, or further AI-premium compression below +15% pull pricing toward the floor.