Executive synthesis — tactical GTM recommendation

Date: 2026-06-30 Repository source of truth: construction-strategy-context-2026-06-30.zip

One-line recommendation

Choose Path B, but enter through Path A: sell a paid, bounded Commercial Recovery Audit to the warm UK firm, then convert only if the same Commercial Event workflow can become a capped retainer. Do not start a full task-management SaaS build in 2026.

This is not a generic agency recommendation. It is tied to the bundle’s own product thesis:

README.md says the core object is the Commercial Event, not the diary/tender/dashboard.
research/06-build-spec/output/10-current-state-and-roadmap.md says the first paid SKU should be a Commercial Recovery Audit, that the first product is not a diary app, and that the moat is recovery outcomes + trusted cross-firm cost/event benchmarks.
research/dossiers/_LANDSCAPE-SYNTHESIS.md says field capture is crowded, while the missing loop is field evidence → Commercial Event → entitlement/quantum/recovery → cross-firm historical-cost/outcome data.

Filled risk/reward table

These are operating ranges for a two-person team with 1–2 warm UK firms. They are decision numbers, not TAM claims.

Path	Expected reward	Probability it works	Downside	Time-to-cash	Decision
A. Consulting / Commercial Recovery Audit	£12k–£25k per diagnostic; £30k–£120k first-year cash if 1–4 paid packs close	60–75% for first warm firm if one named matter has ≥£100k at stake	4–10 founder-weeks; £1k–£3k travel per UK visit unless reimbursed; reusability may stay <30%	2–6 weeks from scoped matter	Use as entry motion only
B. Agency / retainer	£6k–£15k/month per firm; £72k–£180k/year per firm; 3 firms = £216k–£540k/year	35–55% to reach 1–3 retainers via warm firm + intros; 15–30% to reach 10–15 without a channel partner	Service trap; 3–6 FTE/contractors needed at 10 clients; bespoke demands can erase product core	1–3 months after first audit	Recommended posture
C. SaaS task/field platform	£25k–£100k/year per firm only after repeatable workflow; £1m–£3m ARR requires 20–50 paying firms	10–20% from current state without funded design partners; 25–40% only after ≥10 paid matters and 2+ weekly active firms	12–24 months founder time; clone/bundle risk from Procore/Autodesk/Trunk Tools/Gather; negative cash while building	9–18 months to meaningful ARR	Do not start now

Why this wins on risk-adjusted earnings

The value anchor is commercial loss, not software-seat savings. UK adjudication is fast and evidence-heavy; public sources put the adjudicator decision timeline at 28 days and common adjudicator fees around £12,001–£14,000, with hourly rates frequently £301–£350. Late-payment pain is also current and material: UK government sources put economy-wide late-payment cost at almost £11bn/year and 14,000 business closures/year. These make a £12k–£25k audit rational when it protects or improves a £100k–£300k commercial matter.
The software benchmark ceiling is real but not sufficient. PlanRadar publishes UK per-user prices from the low tens to low hundreds of pounds per month; Fieldwire publishes $39–$89/user/month; Procore prices by annual construction volume with unlimited users; Autodesk quotes flexible user/project/account pricing. A new entrant cannot justify Procore-style platform pricing until it proves money recovery. It can justify £6k–£15k/month when tied to one avoided write-off, recovered variation, accelerated payment, or better final-account position.
The wedge is not full task management. Your bundle repeatedly rejects broad capture/task UI as the first product. The wedge is a QS/commercial workflow: detect and reconstruct NEC/JCT Commercial Events from photos, voice notes, WhatsApps, RFIs, diaries, drawings, programmes, and cost records; produce evidence gaps and draft commercial actions for human review.
Defensibility must be contracted now. The only realistic moat for a two-person team is not model quality. It is a permissioned event/outcome dataset, UK contract-specific workflow rules, and named references in a tight commercial-construction network. That requires IP/data/reference clauses in the first SOW, not later.

Pricing recommendation

Offer	Price	Packaging	Rationale
Paid diagnostic: Commercial Recovery Audit v0	£12k–£25k fixed fee for 4–6 weeks	1 named project or 1–3 named commercial events; 50% upfront, 50% on delivery; travel + cloud/LLM pass-through	Below the total cost of many adjudication/legal-prep episodes; small relative to one £100k–£300k issue
Larger recovery/reconstruction pack	£25k–£60k fixed fee	Includes deeper document ingestion, event register, evidence matrix, missing-evidence log, draft notices/RFIs/variation narratives, QS/commercial review pack	8–20% of a £300k protected/recovered position is defensible if the buyer names the money outcome
Capped retainer	£6k–£15k/month, minimum 6 months	Includes hosted event ledger, weekly commercial review, agreed pack quota, maintenance, and workflow adaptation	Sized below platform-replacement procurement but high enough to fund product R&D
Later SaaS module, not now	£750–£2,500/project/month or £150–£350/commercial seat/month	Field users cheap/free; commercial/QS users and projects pay	Competes above PlanRadar/Fieldwire seats only because it is money-recovery workflow, not field capture

Numeric decision triggers

Advance from audit to retainer only if all are true after the first 1–2 paid audits:

Value: at least one matter has ≥£100k claimed/lost/preserved value or ≥£150k retention/payment exposure.
Payment: the firm pays ≥£12k fixed fee plus reimburses travel and exceptional cloud/model costs.
Access: the firm supplies at least 3 messy evidence sources: diaries/photos, WhatsApp/email, RFIs/instructions, drawings, programme/lookahead, payment applications, cost/labour/daywork records.
Reuse: ≥60% of event schema, prompts, evidence rules, and export templates can be reused for firm #2.
Hours: pack production falls to ≤40 expert hours after the third matter.
Channel: the client agrees to 3 named introductions and one reference call after accepted delivery.
WTP: cold-referral WTP is ≥70% of warm WTP.

Start SaaS only if all are true:

≥10 paid matters.
≥50 normalized UK Commercial Events.
≥3 observable outcomes: recovered cash, accelerated payment, preserved entitlement, avoided write-off, or final-account improvement.
≥2 paying firms use the system weekly without founder attendance.
≥75% of reviewable events have source-linked contemporaneous evidence.
≥70% of build is reusable across firms.
≥5 prospective firms are willing to pay ≥£25k/year or ≥£1k/project/month before seeing a custom build.

Kill SaaS if any of these occur:

First warm firm wants a full task-management replacement before paying for a money workflow.
No firm pays ≥£12k for the audit.
No real records are shared under NDA.
Third pack still requires >40 expert hours.
Client refuses vendor-owned reusable core IP.
Benchmark/data-rights opt-in is <30%.
Support exceeds 6 founder hours/week/client after month 2.