The problem
She was managing prospects by hand — forgotten follow-ups, lost contacts, no overview of her pipeline.
The solution
Custom-built site, automatic LinkedIn lead qualification, scheduled follow-ups, and sector monitoring that feeds her content with human validation. Nothing slips through the cracks anymore.
Approach
The existing showcase site, a single block hard to evolve, was rebuilt in Next.js (TypeScript, Tailwind): offers, about, articles, contact, structured data, sitemap, GA4 audience measurement triggered only after cookie consent, and Search Console monitoring.
Around the site, a private admin area gathers five modules: prospecting (a base of over 2,000 sector organizations, daily list, daily LinkedIn target, e-mails sent from the site), publishing (ideas, assisted writing, calendar), a knowledge base fed by monitoring of sector websites, visibility statistics, and settings where Audrey enters her own AI keys, encrypted, on her own accounts.
Automations run in a self-hosted n8n: daily monitoring, indexing of sources into a vector database so that drafts rely on real documents, follow-ups, supervision. Each module was delivered step by step and validated by Audrey in real conditions before moving on to the next.
Site SEO / GEO audit
Scores measured on October 4, 2026 across 14 public pages of aecsconseil.com, with Waycraft's in-house audit tool (the same one used on waycraft.studio). They describe the site's current state.
- 86
Overall
86 / 100
- 81
Technical
81 / 100
- 93
Content
93 / 100
- 83
GEO (AI assistants)
83 / 100
What the score is based on
- Overall: the average of the three scores.
- Technical: page title (20 to 65 characters), meta description (70 to 160), canonical tag, page language, a single h1, indexable page, Open Graph tags, mobile viewport.
- Content: length (at least 200 words for editorial pages), image alt text, internal links, h2 heading structure.
- GEO: valid JSON-LD structured data, dates in that data, AI crawlers allowed, presence of llms.txt, llms-full.txt, ai.txt and ai/summary.json.
This is not a Google rating. It is a transparent checklist: each criterion is weighted and the score is the share of pages that meet it. It measures the site's technical state, not its ranking or traffic.