Strategy — internal · 20 May 2026
Verified property information, built on a community data substrate. Where the market is, who's already in it, where Sail wins — and where we still have to prove ourselves.
Huw Thomas · Evan Williams · Richard Scott · Status: working assessment · Awen Weave Limited (17229089)
01Why now — the law is already here
The transaction is being forced upstream. The duty to collect, verify and present material information now sits with the seller and agent — before marketing. Sail is built for exactly this regime.
Every fact verified 20 May 2026. Law-in-force stated as fact; reform, logbooks and statistics labelled as proposals/projections.
02The opportunity
Property data is fragmented, unverified, and re-gathered from scratch every transaction. The regulation now demands it be assembled, verified and disclosed at the point of marketing.
The gap
Less than 1% of property data is truly digital today (OPDA). 82% of consumers support digital property packs; 77% would use one. The demand and the legal duty both exist — the trusted, verified spine does not.
What Sail is
Verified, source-cited property information assembled at listing (TA6 pre-fill, material-information pack) plus a living UPRN-linked logbook that persists and accrues across the whole ownership lifecycle.
Market size — the honest version
Fund the thesis on the regulatory tailwind (mandatory TA6, DMCC duty), demand signals (82% consumer support), and the verification moat — not on a market-size number that hasn't formed.
The field
| Player | What they are | Hole | Raised |
|---|---|---|---|
| Sprift | Data breadth (300+ points, 30M properties) + CRM distribution; closest on the listing-stage data play | AI is cosmetic; no provenance, no ground-verification, no logbook | ~£1.8m |
| Moverly | Closest single competitor — AI + listing-stage + open-standards (OPDA) + £50k govt Smart Data prize | Document-assembly — "nothing is actually checked"; no living logbook | ~$1.2m |
| Coadjute | Bank-backed network rails + provenance language + UPRN | Pivoted to managed AML; provenance of checks, not property facts | ~£23m |
| Landmark | Contract Pack Vault — AI doc-validation (conveyancer stage) | Tied to paid-search revenue it must cannibalise | DMGT |
| RLBA / Chimni | Living UPRN logbooks (~500k registered) | Records only — no AI assembly, no verification at source | — |
Above them sits OPDA — a bank-backed standards body (NatWest, HSBC, Nationwide, Lloyds, PEXA) publishing the open property-data schema. The substrate is becoming a public good; the value migrates to verification.
05Funding & the real risk
Money is flowing
UK proptech raised £230.4m in 2025. Closest comparable: Conveyd's £2.5m (Eka Ventures, Mar 2025) — same niche, impact-aligned thesis. Banks are co-building via OPDA, not just acquiring.
The cautionary tale
Coadjute — ~£23m, backed by Lloyds, NatWest, Nationwide and Rightmove, after 4+ years — still publishes no transaction volume; its flagship "national rollout" is still a 12-branch pilot.
Pilots flatter; volume is the chasm. Conservative, sticky estate agents and the "second tool" problem are where good proptech dies — not in the build. We cannot out-spend that path, so we must out-manoeuvre it.
The wedge
| Capability | Who holds it | Sail |
|---|---|---|
| AI assembly | Conveyd, Moverly, Sprift | ✓ |
| Provenance + UPRN | Coadjute | ✓ |
| Listing-stage open-standards pack | Moverly | ✓ |
| Living UPRN logbook | RLBA / Chimni | ✓ |
| Ground-truth human verification | Nobody | ★ |
| Welsh-sovereign, non-extractive community model | Nobody | ★ |
| Construction / refurbishment value pool (Olive) | Nobody | ★ |
Every competitor holds a piece. The three starred rows are uncontested — and they are precisely the things the conveyancing-only, commercial, licensed-data incumbents are structurally unable to build.
07The trust card — our big win
The product is computable trust — provenance, validation, ground-truth — applied to open public data and the customer's own. Because the product is the verifying and not the data, the data stays sovereign and the model is structurally non-extractive.
"Rightmove can build the tool. They cannot buy five years of verified curator decisions."
08The operational unlock
The verification moat was also the bottleneck — slow, costly human curation. This dissolves it: the agent's marketing visit is the cheapest ground-truth event in the entire property lifecycle, and it already happens.
This makes "every user is a curator" literal and zero-extra-work — the curator doesn't need a separate visit; the marketing shoot is the visit.
From capture to construction
The same capture event that verifies the property profile yields construction-relevant data: fabric, condition, materials, dimensions. One visit, two value layers.
Sail — the verified pack
TA6 pre-fill, material-information pack, living logbook. Cited, source-linked, compliant at marketing.
Olive — the construction layer
Turns the captured condition + fabric data into refurbishment cost intelligence — "what would it cost to extend / convert / rewire?" — anchored to the property's actual construction type, age, condition and listed status.
None of the conveyancing-only competitors touch this value pool. Sail + Olive extends from the transaction into the whole ownership-and-improvement lifecycle — an entire adjacent market the field ignores.
Beyond conveyancing
Insurance — strongest commercial pull
~93% of UK properties are mis-valued for insurance. Insurers are racing for verified attributes at point of quote (Verisk, Cytora, Concirrus). They pay for verification specifically and want the fabric/condition data ground-verification produces — possibly a faster pull than agents.
Welsh retrofit — funded, mission-aligned
Welsh Government's Warm Homes "passport" + ORP mandate a per-home living record for social landlords (75% by Sept 2026). The logbook concept, validated with Welsh public money — and a natural CIC entry point.
Building safety — the precedent
The Building Safety Act golden thread is the legal precedent for a verified per-building living record. Narrow stock today; strong concept-validation.
The playbook
In diffusion terms: a strong p (the DMCC duty + mandatory TA6 = a regulatory forcing function) and a strong q (CIC community word-of-mouth). The funded competitors have neither working together.
Where Sail is — honestly
The two Low-confidence dimensions — adoption and curator economics — are holding the score down, and they are exactly what the deeper research flagged as make-or-break. 73 is a compelling hypothesis, not yet a decision. The next phase is converting those two from Low to High confidence.
The verdict
A real opening and a defensible wedge (verification + living logbook + Welsh-sovereign + the Olive construction layer). But adoption is the war, and the moat is the bottleneck — so the strategy is to out-manoeuvre, not out-spend.
Re-architect verification
Around the marketing-capture event — this de-risks the moat-is-the-bottleneck problem and makes the curator model real.
Treat insurance as a parallel market
Hungrier, pays for verification specifically, wants the condition data we produce.
Ride existing community infrastructure
Colouring Cities + Welsh LAEP networks — don't rebuild what's already convened.
Prove next (Low → High confidence): model the curator economics, and run a one-town beachhead. …but the seat is in the far corner, and the floor is slow to cross.