Laurelin Labs · Information Architecture Audit & Proposal

From a content-type site to a topic-first knowledge base.


Retire the blog. Make the topic the spine. Every article, framework, service, lab build and project gets one obvious home, anchored to the Four Pillars framework Laurelin already sells. This is the case for the restructure, and an interactive map of the proposed architecture.

Method: PST / TT demand-weighted L1 · L2 Method: Music Ally two-layer nav Frame: The Four Pillars Data: DataForSEO UK live · 9 Jul 2026 Scope: laurelinlabs.com

01 · Where the current site fights itself

Two organising systems, and the topic one is hidden

The live site is organised by content type in the primary nav, while a parallel topic taxonomy exists but only surfaces as a filter on one page. That split is the root of everything below.

Nav is type-first

Primary nav: Services, Frameworks, The Lab, Insights, About. A visitor has to already know whether the thing they want is a "service" or a "framework" to find it. Content type is an internal distinction, not how anyone searches.

"Insights" is not a blog, and grows nothing

The Insights page aggregates existing services, frameworks and lab builds by cluster. There is no article or editorial content type at all. There is nowhere to publish and grow topical content, which is the one thing that moves Pillar 2.

The topic layer is subordinate

Five clusters exist in the code (AI search visibility, entity authority, technical SEO, measurement, paid media) but they live one level down, as a filter. The thing that should be the spine is a side-feature.

The clusters do not match the framework

Laurelin sells the Four Pillars, yet the site clusters do not map to them: Topical Content (Pillar 2) has no cluster at all, "AI search visibility" spans everything, and measurement and paid sit outside the pillars. The site does not practise what it preaches.

The test that fails. An IA is only as good as the share of real demand and future content it can house without a rethink. Right now, ask "where does a new case study go?", "where does an article on E-E-A-T live?", "where does the next lab tool attach?" and there is no single correct answer. That is the failure a topic-first restructure fixes.

02 · The move

Make the topic the spine, anchor it to the Four Pillars

Borrow the exact model that worked for PST/TT and Music Ally: a demand-weighted topic spine as the primary structure, with an editorial or framework layer sitting behind it. Here the framework layer is Laurelin's own Four Pillars, so the site literally demonstrates the methodology it consults on.

Now · organised by type

ServicesFrameworksThe LabInsights (filter)About

Topic clusters buried one level down. No article type. New content has no obvious home.

Proposed · organised by topic

TopicsServices (lens)The Lab (lens)About

Topics becomes the hub: one hero topic, four pillars, three cross-cutting tracks. Services and The Lab remain as cross-cutting lenses over the same content, not competing silos.

Why it works for search demand

  • Query fan-out makes the cluster the unit. AI search decomposes one question into 20 to 50 sub-queries. A topic hub that owns a whole semantic neighbourhood is what gets cited, not a single blog post. The IA should mirror the retrieval unit.
  • Demand-weighted spine, live data. Same as PST/TT: each node carries the summed UK volume of its sub-topics, so nav order and internal-link weight follow real demand. Commercial-intent ranking: AI Search Visibility 28,440, Paid 12,560, Technical 11,260, Authority 6,150, Measurement 2,550, Content 2,680, Guardrails 2,510, Trust 1,540 searches / mo.
  • Data-backed apex. The positioning bet paid off in the numbers: once navigational product terms (google ads, GSC) are set aside, AI Search Visibility is the single largest genuine cluster, led by 'ai overviews' (12,100) and 'ai search' (5,400). The apex is justified by demand, not just strategy, and it still has near-zero competition.

Why it future-proofs content

  • One home rule. Every future article, framework, service, lab tool and project classifies into exactly one pillar node. No orphans, no "which section?" debate.
  • Coverage becomes visible. A topic-first tree exposes where content is thin. Pillar 3 (Trust Signals) currently houses zero pieces: the restructure surfaces the gap instead of hiding it.
  • Defensible and self-demonstrating. The site's own structure is the Four Pillars, so the architecture is a live proof of the framework and compounds topical authority around Laurelin's proprietary IP.

03 · The proposed architecture · click any node

The topic tree

One hero hub, the four pillars beneath it, three cross-cutting tracks. Click a node to see the demand read, its pillar anchor, and every existing piece of content that moves into it.

Hero hub · the money page

AI Search Visibility
AEO · GEO · LLMO

The four pillars · how visibility is earned

Cross-cutting tracks

04 · Every current page, rehoused

Migration map

All 23 existing content items map into exactly one node. Nothing is lost, nothing is orphaned. The one node with zero existing items, Trust Signals, is the priority content gap the new IA reveals.

Content itemTypeTodayNew home

Demand data. Volumes are live UK monthly searches from DataForSEO (Google Ads search volume, United Kingdom, English), pulled 9 Jul 2026 across an 80-term set mapped one-to-one to nodes. Two cleaning passes were applied so the weighting reflects real intent: 'google eat' (1.22M, food-intent noise) was removed from Trust Signals, and the navigational product terms 'google ads' (165k) and 'google search console' (110k) are reported separately rather than driving the spine. Bars are relative to the largest cluster.

Laurelin Labs · IA restructure proposal · draft for agreement · built 9 Jul 2026 · no volumes fabricated