Boucher & Jones IA recommendation: Pattern A + hub-and-spoke, NOT enterprise matrix

The specific recommendation for the Boucher & Jones website (regional Petro-Canada lubricant distributor, 4 dominant verticals: Agriculture / Fleet / Construction / Residential).

Pattern

• URL convention: Pattern A — /industries/<vertical>/ (see [[url-structure-patterns-by-pattern]])
• Architecture: Hub-and-spoke. One page per vertical with shared-services blocks pulled in via partials/CMS includes — not a full Industries × Services matrix.
• Top nav: Industries (mega menu) + Services (mega menu) + About + Contact. Two axes orthogonal at the top.
• Mobile nav: Accordion or sequential — not split buttons (see NN/g: split buttons unreliable on touch (fat-finger conflicts) — use sequential menus or accordions on mobile).
• Schema: One Service per vertical when offerings genuinely differ (see [[rule-one-service-per-vertical-when-offerings-differ]]); one Service with multiple audiences when offering is materially identical.
• Forms: Single-step under 7 fields. Multi-step only for the longer "request a quote" flow with a vertical-selector first step (see [[zuko-multi-step-only-helps-above-7-fields]]).

Why this specifically

• B&J's 4 verticals have distinct vocabulary (hydraulic oil ≠ fleet engine oil ≠ furnace oil) — generic IA underperforms.
• B&J's team size doesn't support an enterprise matrix — hub-and-spoke maintains shared-service blocks once, in one place, then renders into each vertical hub.
• E.H. Wolf ([[eh-wolf-petro-canada-distributor-ia-pattern]]) is the closest existing analog and uses a very similar pattern (with the Pattern B flat URL convention — Candid can choose Pattern A for future-proofing).
• RelaDyne ([[reladyne-industries-pattern-19-verticals]]) shows what happens at 19 verticals; B&J doesn't need that fan-out.

Migration sequence (if existing site)

1. Phase 1: Stand up vertical hub pages with substantive content (case studies, regulations, vocabulary) — see [[rule-vertical-pages-need-substantive-content]]
2. Phase 2: Build shared-services blocks; pull them into each hub
3. Phase 3: Decide whether to expand into a full matrix (Pattern E) — only if specific matrix queries justify the content investment
4. 301 redirects from old paths preserve URL equity throughout — see [[sej-892-migrations-523-day-recovery]] for the migration cost reality