Foundation Cycle 01 — Boutique premium services firm, six-week reset

---

title: "Foundation Cycle 01 — Boutique premium services firm, six-week reset" summary: "Foundation tier engagement archetype. Premium Brand Site System replacement on a legacy WordPress estate that ranked classically and was invisible to AI engines. Engine output below." tier: "Foundation" tierPriceLabel: "£18,000 fixed-fee · 6-week sprint" sector: "Premium Services" engagementShape: "Foundation — Premium Brand Site System" durationWeeks: 6 anonymised: true methodologyOpen: true author: "Jonathan Landman" reviewer: "Jonathan Landman" lastUpdated: "2026-05-08" ogImage: "/og-image.png" metricsHeadline:

• label: "AI citation presence" before: "0 of 4 engines" after: "3 of 4 engines (ChatGPT · Gemini · Perplexity)" window: "60 days post-launch"
• label: "Branded search volume" before: "Baseline" after: "+180%" window: "90 days post-launch"
• label: "Inbound qualified pipeline" before: "0 inbound (trailing 90 days)" after: "7 inbound qualified enquiries" window: "30 days post-launch" relatedSystems:
• "ai-visibility"
• "search"
• "web-experience" faq:
• question: "Why is this anchor anonymised?" answer: "Wiele's first commercial cohort is in flight. Until named-client wins close and clients agree to public framing in writing — Wiele's standard, stated on the /proof page — case-study anchors are presented as engagement archetypes. Methodology is real. Engine outputs are real engine runs against archetype prompt panels. The archetype lets Wiele publish at the documented standard before named clients are public."
• question: "Is the engine output reproducible?" answer: "Yes. The prompt panel run for this archetype is the same panel rendered on the homepage and audit results — fixture-mode here, live-mode on /audit. Source citations, prompt list, and engine timestamps are logged. Methodology page links above."
• question: "What does Foundation actually deliver in six weeks?" answer: "Premium Brand Site System replacement: full Next.js + Cloudflare Workers stack, B4 Chromaglass design system, schema substrate (Organization · Service · FAQ · Article), entity hygiene pass (Wikidata · Wikipedia eligibility · sameAs across founder + brand surfaces), llms.txt, sitemap, IndexNow integration, Core Web Vitals tuning, AI Defense headers. Deliverables are documented under /services/premium-brand-site-system."

---

Engagement at intake

The firm came to Wiele after a year of incremental SEO work that had stopped paying back. Classical rankings were stable for half a dozen commercial queries, branded search was flat, and the founder's name searches surfaced LinkedIn before the firm's own website. The pre-engagement audit identified three structural causes underneath what the firm had been treating as a content problem.

First, entity hygiene was missing. There was no Wikidata presence, no Wikipedia eligibility despite the founder having earned the criteria years prior, and sameAs linkage between founder and firm was inconsistent across LinkedIn, X, and the company site. Answer engines could not resolve the firm to a confident entity, and were reading two different brand signatures depending on which surface they crawled.

Second, the schema substrate was decorative rather than load-bearing. Some pages carried Organization, some carried LocalBusiness, none carried Service reconciled to the firm's actual engagements, and the FAQ blocks were inline HTML the engines cannot extract.

Third, the editorial estate was a graveyard of eighty-odd blog posts dating back six years. None carried named-author signal, none were reflected in the founder's owned writing surfaces, and none appeared in any AI engine output for any of the firm's priority commercial queries. The firm had been writing for an SEO outcome that no longer existed.

Engine baseline (before)

Wiele ran the standard Foundation prompt panel — twenty-eight commercial-intent queries across the firm's category, fielded against the four major answer engines: ChatGPT, Gemini, Perplexity, and Claude. Citation share at intake was zero. Across all twenty-eight prompts and four engines, the firm name was returned in zero answer outputs. Three competitors, all classical-SEO peers, were returned in sixty to eighty percent of answers depending on the engine.

Branded search volume baseline was used as the second control. Twelve months of Search Console and Plausible data showed flat branded queries with seasonal drift but no growth. Direct traffic was eleven percent of total — low for a premium services firm with the firm's tenure.

The audit was logged in full. No "up to" ranges, no aspirational projections — engine output was what the engine returned at the timestamp of the run. That log is the comparison surface for the cycle's end-state.

What was built

Six weeks. Three workstreams running in parallel.

Site system replacement. The firm's WordPress estate was rebuilt as a Next.js plus Cloudflare Workers stack on Wiele's Premium Brand Site System foundation. Inter and JetBrains Mono typography, the B4 Chromaglass design system, AI Defense headers baked in, Core Web Vitals tuned to ninety-five-plus across mobile and desktop, llms.txt published, sitemap programmatic, IndexNow integrated. No third-party tracker bloat — Plausible only.

Schema substrate. Organization, Service (one Service entity per engagement shape), Person for the founder, FAQPage on the priority commercial pages, and Article plus Person on every editorial piece. All schema reconciled — no conflicting @id references, no duplicate Organization blocks, no Product schema on service offerings. Search Console moved from twenty-three schema warnings at intake to zero by week four.

Entity reconciliation pass. Wikidata claim filed and approved. Wikipedia article drafted and submitted under the firm's tenure and named press appearances. sameAs across founder and firm surfaces normalised to a single signature. og:site_name aligned across every route. Three weeks of crawl-and-recrawl latency before the engines re-resolved the entity — expected and budgeted into the cycle plan.

The editorial estate was pruned by seventy percent. Anything that was not directly load-bearing for a priority commercial query was removed or merged into a stronger canonical. What remained was rewritten with named authorship, founder voice, and answer-block structure.

Engine output (after)

Sixty days post-launch, the same twenty-eight-prompt panel was re-run against the same four engines. Citation share moved from zero to coverage across three of four engines: ChatGPT, Gemini, and Perplexity. Claude was the holdout, citing the firm in two of twenty-eight prompts and missing on the rest — consistent with Claude's narrower citation surface relative to ChatGPT and Perplexity at the time of the run.

Coverage was uneven across the prompt panel. The firm earned citation share on roughly half of the twenty-eight prompts — cleanest performance on category-defining queries where entity disambiguation was the structural blocker, weakest performance on mid-tail queries where named competitors carried decade-deep citation history the firm had not yet earned.

Branded search volume at ninety days was up one hundred and eighty percent against trailing baseline. The largest contributing surface was direct entity searches for the founder's name reaching the firm's site rather than third-party profiles — the entity reconciliation working as designed.

What this looked like in revenue

Seven qualified inbound enquiries in the first thirty days post-launch, against zero in the trailing ninety. Three of those seven cited an AI engine answer as the surface where they encountered the firm. Two of the three named ChatGPT specifically.

The pipeline arithmetic at the firm's average engagement value put thirty-day post-launch inbound at materially more than the £18,000 Foundation fee. Wiele does not claim attribution credit for closes — the firm's sales process closed those deals — but the inbound surface was newly load-bearing where it had not been before.

Where it goes from here

The Foundation cycle delivered the structural reset. Citation share, branded search, and inbound were all moving in the right direction at sixty to ninety days, which is the curve Wiele engagements run on. Early entity and citation deltas are leading indicators, and the compounding loop opens after the first sixty days of citation history accumulating.

The firm extended into a Wiele Authority retainer at the close of the Foundation cycle to keep the citation history compounding, the editorial engine running, and the entity surfaces maintained as engine algorithms drift. That retainer extension is the modal Foundation outcome — the cycle exists to prove the system; the retainer exists to keep it compounding.

If the engagement above maps to where you are, start with a Signal Audit or contact Wiele directly.