Layout, copy and image slots for the new .ai site. These are static mockups for review, not the built site. Nothing here is live, indexed, or linked from anywhere public.
Copy comes from site-builder/WIREFRAME-AND-COPY.md. Dashed accent blocks are image
placeholders — each is labelled with its ID, aspect ratio and what it has to show.
Opens on a question the reader can't answer — "which one of these numbers is right?" — then shows the contradiction as an evidence graphic: three systems reporting three different numbers for the same month. The turn is the emotional core: you became the integration layer, and that's the job we're taking back. Centred FluidIQ.AI lockup above the headline. Four colourways: 8A–8D.
Barely argues at all. Opens on the Growth Check output itself — three findings ranked by what they're costing, each with its fix attached and the number it's meant to move. The claim is one line; the artifact does the persuading. Modelled on how Linear leads with real product UI instead of a value proposition. Four colourways: 7A–7D.
Leads with the qualifying symptom straight from the ICP doc — fifteen tools, fifteen numbers, not one of them deciding anything — so the reader self-identifies before any claim is made. Carries the large FluidIQ.AI lockup above the headline at display size. Four colourways: 6A–6D.
The same V5 page in six registers: 5A Editorial Luxury · 5B Dark Enterprise · 5C Swiss Precision · 5D Split Cinematic · 5E Product-Led · 5F Marine. Structure and copy are held constant so the comparison is about palette and visual register, not about the argument. Q-SB-08 (the palette) is the gate these exist to close.
V4's ideas, subtracted down. Headline A adopted and the selector removed; the Gap section compressed to a band that leads with its own payoff; the four-card Loop reduced to Find → Fix → Verify → Monitor; the layoff rebuttal cut (naming the fear is what plants it); Who It's For merged into proof; and "and fix what you approve" removed from the Growth Check, which is the doorway, not the product. ~13% shorter than V4 and faster to scan.
V3 was built from the handoff brief alone. V4 goes back to all eight positioning documents. New: the category ("a business operating system — one layer underneath the tools you already have"), the emotional core from brand-foundation ("the owner becomes the integration layer"), the loop (what's wrong → what we'll do → whether it worked → keeping it that way), every fix tied to a number, who it's for (capacity, never headcount), and transparency named alongside control. Three headlines are switchable in the top bar — positioning is flagged IN FLUX, so feel them rather than argue about them.
Same visual language as V2, compressed and rebroadened. The full Problem section becomes a short resolution-gap band; the three examples now span growth, customer/operations and workflow so FluidIQ doesn't read as PPC software; customer control is promoted to a core product idea; the "17 accounts" metric is gone; the footer line goes active. The Growth Check CTA is held as provisional pending the offer review. Palette switcher in the top bar.
Six sections: hero → problem → how it works → see it work → you decide how involved → proof + CTA. Rebuilt 2026-08-10 after the positioning correction: FluidIQ acts, it is not a read-only diagnostic layer. Superseded by V3 on structure and copy — the visual system is unchanged.
The original nine-section page. Contains dead positioning — the "we never touch the layer that runs your business" boundary and mandatory review-first. Kept only so V2 can be compared against it. Do not use for copy.
The mechanism, the autonomy spectrum, ownership and permissions, and what happens when FluidIQ is wrong. ⚠️ Its written copy is still V1 and contains the dead read-only model.
The retainer. The only line of business that exists today.
Thirteen years, the origin, and the failure story.
The form, and what happens after you send it.
Two things are deliberately unresolved and visible in the mockup.
The palette. Use the switcher in the black bar on the Home mockup. The logo pack that actually exists is orange on near-black; the positioning doc proposes cyan on ink and marks itself “still TBD.” They can't both be right, and this has to settle before the design tokens are written.
The entry offer. On V3 the CTA block carries an
ENTRY OFFER — PROVISIONAL chip instead of a price. Growth Check is a placeholder:
the offer architecture is being reviewed separately, so nothing in that section should be
locked into the design system yet.