Architect - Effort Model (Internal)¶
Never client-facing. A living model: capture per-step, per-tier actuals on every Architect engagement and revise it (see Calibration).
Assumptions¶
- Conventions: website-pricing.md § Effort-Model Conventions; 2.3's mid hours are design mids at the same rate.
- Pricing basis - as Clarify: external prices carry a productised uplift over rate-card cost (27–43% across the sold tiers; the Baseline sits far above at 124% because it is a non-removable minimum carrying the Cornerstone, not a sold tier), validated against five ideal-site actuals. Cost bases and rates are defined once in website-pricing.md § Effort-Model Conventions. See Calibration flag 1 for the history.
- 2.3 is costed at the tier's default ambition band - Focused System-led, Core Expressive, Full Premium. Banding up beyond the tier default is a priced delta (see flag 5).
- Assumes the templated instruments are bedded in: the Cornerstone prompt, the boilerplate library, and a templated concept format. First runs will cost more (see Calibration).
The Architect Baseline (the Non-Removable Minimum on Every Build)¶
| Element | P hrs | M hrs |
|---|---|---|
| Light IA + content-object sketch (from the proposal and existing estate) | - | 3 |
| Three-state scope from the sold proposal, confirmed with the Decider | 1 | 1.5 |
| Technical definition: default-stack ADR, plan selection + upgrade trigger, integration map | 2 | 1 |
| Direction from existing brand guidelines, recorded | - | 1 |
| Brand evaluation against those guidelines | 0.5 | 0.5 |
| Two System-led concepts | 1 | 10 |
| One revision round on the chosen direction | - | 2 |
| Principal direction review + concepts playback | 0.5 | 0.5 |
| Cornerstone: prompt assembly, principal correction, internal review, written confirmation | 2.5 | 3.5 |
| Total | 7.5 | 23 |
Rate-card cost: (7.5 × £165) + (23 × £110) = £3,768. Staff cost: (7.5 × £119) + (23 × £82) = £2,779. Priced at £4,000 on every S build quote (non-removable - it carries the Cornerstone).
Why the Baseline carries concepts. It is the floor beneath every S build sold without a deeper Discovery, and its job is to carry the Cornerstone - the artefact that fixes price, scope and time. Without concepts a client signs that having seen no visual direction at all, so any mismatch surfaces in Design and the rework lands inside a fixed price. Two concepts and a round close that gap. They are not optional: an optional element inside a non-removable minimum is the first thing dropped under pricing pressure and the most expensive omission when it is wrong.
Focused (S/M Band · System-led · 1 Concept × 1 Round)¶
| Step | P hrs | M hrs | What drives it |
|---|---|---|---|
| 2.1 Experience & content structure | 1.5 | 5.5 | Noun-first content-model pass in one principal working session; mid develops the IA (≤50 nodes) and journeys for one or two audiences |
| 2.2 Technical & delivery | 3.5 | 2.5 | Platform confirmation ADR from the default stack, plan + upgrade trigger, S/M integration map, baselines from boilerplate, scriptable migration approach where an estate exists |
| 2.3 Visual direction | 1.5 | 13.5 | Brand evaluation, two System-led concepts, one round; principal reviews direction and attends playback. A third concept is a stated upgrade (see flag 9) |
| 2.4 Scope & priorities | 1.5 | 2.5 | One trade-off working session with the Decider; mid documents the three-state table |
| 2.5 Cornerstone | 4.5 | 3.5 | Part A gap audit + Part B assembly (mid runs, principal corrects), internal review, 60-min playback, written confirmation |
| Total | 12.5 | 27.5 | 40 hrs ≈ 6.5 tracked days |
Rate-card cost: (12.5 × £165) + (27.5 × £110) = £5,088. Staff cost: (12.5 × £119) + (27.5 × £82) = £3,743.
Core (L Band · Expressive · 2 Concepts × 2 Rounds)¶
| Step | P hrs | M hrs | What drives it |
|---|---|---|---|
| 2.1 Experience & content structure | 5 | 14 | 50–100 nodes, structured/relational content model, journeys per audience including editors |
| 2.2 Technical & delivery | 7.5 | 6.5 | CMS comparison matrix (options considered), business-system integrations (CRM etc.), migration tiers + redirect matrix + SEO risk, moderated governance |
| 2.3 Visual direction | 4 | 26.5 | Three Expressive concepts × two rounds; principal direction steer and concepts playback |
| 2.4 Scope & priorities | 3.5 | 2.5 | More workstreams triaged; one bounded re-pass of an affected stream built in |
| 2.5 Cornerstone | 7 | 6 | Fuller assembly, internal review, 90-min playback |
| Total | 27 | 55.5 | 82.5 hrs ≈ 13.5 tracked days |
Rate-card cost: (27 × £165) + (55.5 × £110) = £10,560. Staff cost: (27 × £119) + (55.5 × £82) = £7,764.
Full (XL Band · Premium · 3 Concepts × 2 Rounds · From £20,000, Scoped Bespoke)¶
| Step | P hrs | M hrs | What drives it |
|---|---|---|---|
| 2.1 Experience & content structure | 8.5+ | 23.5+ | 100+ nodes, highly relational model, multi-audience / multi-region journeys |
| 2.2 Technical & delivery | 12+ | 10+ | Volatile/custom APIs, enterprise governance, migration at volume |
| 2.3 Visual direction | 5+ | 31+ | Three Premium concepts × two rounds |
| 2.4 Scope & priorities | 5+ | 4+ | Many workstreams, multiple decider-adjacent stakeholders |
| 2.5 Cornerstone | 10.5+ | 8.5+ | Staged assembly and playback, versioned governance |
| Total | 41+ | 77+ | 118+ hrs ≈ 20+ tracked days |
Rate-card cost: from (41 × £165) + (77 × £110) = £15,235. Staff cost: from (41 × £119) + (77 × £82) = £11,193.
The Concept Ladder¶
Concepts, not the ambition band, drive 2.3's effort: the time goes into producing each concept, whatever its band.
| Tier | Concepts | Rounds |
|---|---|---|
| Baseline | 2 | 1 |
| Focused | 2, upgradeable to 3 | 1 |
| Core | 3 | 2 |
| Full | 3 | 2 |
A System-led concept runs about 5.5 hrs, an Expressive one about 6.5. The Focused upgrade to a third concept is a stated +£1,000 (flag 7).
Price Validation¶
Does the breakdown of hours to price hold up? Markup is over rate-card cost, the basis the prices were set on; margin on price is shown against staff cost, which is what the work actually costs to deliver.
| Tier | P | M | Rate-card cost | Price | Markup | Staff cost | Markup | Margin on price |
|---|---|---|---|---|---|---|---|---|
| Baseline | 7.5 | 23 | £3,768 | £4,000 | 6% | £2,779 | 44% | 31% |
| Focused | 12.5 | 27.5 | £5,088 | £5,700 | 12% | £3,743 | 52% | 34% |
| Core | 27 | 55.5 | £10,560 | £12,500 | 18% | £7,764 | 61% | 38% |
| Full (floor) | 41 | 77 | £15,235 | £20,000 | 31% | £11,193 | 79% | 44% |
Focused, Core, and Full validate exactly against the markups quoted in flag 6: the hours, the rate card, and the prices reconcile to the percentage. The Baseline reads differently on purpose. Against staff cost it runs 124%, well above its neighbours, because it is a non-removable minimum carrying the Cornerstone - priced as a floor rather than as a sold tier. Read every percentage in this table as markup over rate-card cost; mixing in margin-on-price is what makes the Baseline look anomalous when it is not.
Composed Discovery SKUs on the same bases:
| Composed | P | M | Rate-card cost | Price | Markup | Staff cost | Markup | Margin |
|---|---|---|---|---|---|---|---|---|
| Baseline floor | 10.5 | 35 | £5,583 | £6,000 | 7% | £4,120 | 46% | 31% |
| Focused Discovery | 23.5 | 46.5 | £8,993 | £9,995 | 11% | £6,610 | 51% | 34% |
| Focused Clarify + Core Architect | 38 | 74.5 | £14,465 | £16,795 | 16% | £10,631 | 58% | 37% |
| Core Clarify + Focused Architect | 48.5 | 79.5 | £16,748 | £19,200 | 15% | £12,291 | 56% | 36% |
| Core Discovery | 63 | 107.5 | £22,220 | £26,000 | 17% | £16,312 | 59% | 37% |
| Full Discovery | 98 | 165 | £34,320 | £45,000 | 31% | £25,192 | 79% | 44% |
Clarify hours are its effort model's tier totals; recheck this table whenever either chapter's hours or the rates move.
The Baseline floor is deliberately the thinnest rung. At £6,000 it runs close to rate-card cost, giving up the productised uplift every other tier keeps. That is a gateway decision, not an error: the floor sits on budget-constrained S engagements and is priced to win the build behind it. It still sits above the £5,000 baseline-only engagement in the validation set, with £1,000 of headroom rather than £2,000. Revisit once there are enough Baseline actuals to test it.
Calibration Flags (the Important Part)¶
- Pricing basis reconciled with Clarify - validated on real data. A rate-card-cost basis (the rates already carry ~40% margin) prices Focused at £4,295, Core at £10,500 and Full from £16,000. Validation against five ideal-site actuals shows clients already pay the uplifted level: two ideal Focused Discoveries sold at £11k and £12k against a £10k rate-card-composed price, and a between-tiers engagement at £20k prices exactly as Core Clarify + Focused Architect on the uplifted menu. Hence the productised-uplift basis, so both halves of Discovery share one pricing philosophy.
- The expected mix is 35:65 (P:M), set as the target and the pricing basis. The instruments (Cornerstone prompt, boilerplate, concept templates) are the delegation lever that makes 35% honest. Watch 2.2 especially: architecture and migration risk is where principal hours creep back.
- Capture actuals. Add per-step, per-tier time capture to the next Architect engagement and revise this document - the delivery-efficiency house metric (touchstone §14) applies here as in Clarify.
- 2.5 is near-fixed, as designed - it varies ~2.4× across tiers while 2.1 varies ~4.5× and 2.3 ~3.5×. If assembly hours blow past this model, the Cornerstone prompt and boilerplate aren't doing their job.
- Band-up deltas (calibrate against actuals): the ambition band was modelled as multiplying 2.3's per-concept and per-round effort at roughly System-led 1× · Expressive 1.5× · Premium 2×. Treat that multiplier as the least evidenced figure in this model. The effort goes into producing each concept, and a concept is largely a concept whatever its band, so concept count is the primary driver and the band a secondary one. Capture per-concept actuals across bands before relying on it. Banding a tier up one level ≈ +50% of that tier's 2.3 hours, priced at rates. The band also prices the Design chapter (3.1) - one band conversation, two pricing consequences.
- Prices (see productisation.md): Baseline £4,000 (priced, non-removable; 6% over rate-card cost - it carries the Cornerstone and two concepts, and is deliberately run near cost as a gateway), Focused £5,700 (12% over rate-card cost at the 35:65 mix), Core £12,500 (18%), Full from £20,000 (31% at floor), typically £20k–£35k. All + VAT. Every percentage here is markup over rate-card cost, on one basis - the Baseline's 124% reads high only because it is a non-removable floor, not because it sits on a different basis. Composed Discovery SKUs: Baseline floor £6,000 · Focused £9,995 (£4,295 + £5,700) · Core £26,000 · Full from £45,000; mixed tiers price as the clean sum (Focused Clarify + Core Architect £16,795; Core Clarify + Focused Architect £19,200). Per-tier margins live here and nowhere else; the cost rates behind them are defined once in the shared Effort-Model Conventions.
- The Focused third-concept upgrade is +£1,000. Focused ships two System-led concepts; a third is a stated delta, not a judgement call at quote time. It costs about 5.5 hrs and holds Focused's margin.