Internal review — not published

Website revamp, v3

What changed, why, and the decisions that need your sign-off before any of it goes live.

Prepared for Sascha Lotz From Kam Ng October 2026 Status: draft for review

Open index.html in this folder to see the new homepage, then solutions.html. Everything works offline — no server needed. This document explains the thinking behind it; the pages themselves need no explanation to be read.

01Why we are rebuilding, not repainting

The old site was not badly made. The problem was that it described us in the order a company describes itself, rather than the order a customer arrives in. Five specific findings, all from the original pages:

FindingEvidence in the original site
The palette contradicted itselfThe stylesheet comments called #f0067b "copper". It renders as magenta. Magenta on plum, with gradient glows, is the default visual language of generated landing pages — it reads as "startup", not "engineers".
One claim, made six times"We connect you, you decide" appeared in the hero, the why-section, all six service cards, the network intro, the process intro and the CTA. 929 words, no new information per section.
No evidenceZero client names, zero project shapes, zero timelines, zero artefacts. The only numbers sat in a counter band with no context.
Content hidden behind JavaScriptEvery card was opacity: 0 until a scroll observer fired. No JS, no scroll, or a print — blank sections.
Placeholders were publicThree testimonial cards reading "Client name pending approval" were live on the homepage.

None of that is fixed by choosing a nicer font. So the rebuild had four rules.

02The four rules

  1. Problem before capability. A buyer arrives with something broken, not with a wish to browse services. The page should let them recognise their own situation first.
  2. Evidence over adjectives. Named people, real cities, counted suppliers, specific numbers. Where we have no evidence, we say so rather than write around it.
  3. One accent, one voice. Warm paper, graphite, and a single oxide accent drawn from the warm end of the logo. The pink now lives only inside the brand mark, on the dark masthead and footer.
  4. Every claim checkable. "Ask them yourself" beats "trusted partner". No claim on the page should require believing us.

03What changed — homepage

The structure moved from capabilities → industries → proof to problem → who → network → diagnose → where we've shipped → how we work → contact.

ThenNow
Hero with a city skyline and no proofSame hero and same headline, plus a proof strip under the buttons: 2 partners · 45 years · 4 contract manufacturers · 20+ sub-suppliers · Shenzhen–Heßdorf
"Why Eastbound Link" — five pain-point questions and a workbench photoA "Who you're talking to" block: two named people, their domains, their numbers, and one honest limit
Six service cardsA twelve-symptom diagnostic, grouped into the six service areas
Stats band reading 45+ / €72M / 20+ / 64Numbers attached to the person they belong to, not floating in a strip
"One connection to a wider network" — six card gridsA supplier ledger with real counts and cities, plus a schematic of where the sites are
The original homepage
Before — magenta accents, six equal cards, no names, no evidence
The v3 homepage
After — same content re-ordered around the customer's situation, with proof in the first screen

The diagnostic

Twelve symptoms, two per service area, written the way a customer says them rather than the way we name the service. Picking one produces a structured answer: most likely · check first · cost to ignore — and never asks for an email. The answers are ours to rewrite; the structure is the point.

Two symptoms selected, one diagnosis shown
Two symptoms selected, one diagnosis shown — the section never grows taller than this

The network

Four contract manufacturers and eleven named sub-suppliers, by capability and city. Presented as a ledger with a schematic, not a world map with glowing arcs — and with the line that matters: "Every site on this page has been visited by one of us. None of it was pulled from a supplier directory."

The supplier ledger and reach schematic
Real sites, real cities, no decoration that is not data

04What changed — solutions page

Six services, each on the same pattern: a headline, one line of context, a four-or-five stage band splitting what you do from what we do, then the detail.

The "you do / we do" split is the most important idea on that page. It says plainly that the decisions stay with the customer and the execution is where we plug in. For a firm with no case studies yet, that honesty is worth more than any claim of expertise.

The NPI section with a five-stage band
NPI — five stages, built from the roadmap you drafted
Sticky jump navigation
A sticky jump nav, because the page is now nine screens long

05What we deliberately did not do

06Decisions needed from you

1. Founding-client terms. I suggested offering the first two or three engagements at reduced rate in exchange for a reference and an anonymised write-up. It is the strongest honest lever we have — but it is a commercial decision, and we need to agree it together before it appears anywhere.
Recommendation: agree before launch, publish as a short block on the homepage.

2. The reference line. "Two suppliers and one former manager will take your call" is the most persuasive sentence on the site. It requires written consent from those three people before it ships.
Recommendation: ask for consent now; publish names only with permission.

3. The twelve symptoms. Six are mine, six were drafted earlier and are still generic. They should sound like our customers, not like a consultancy.
Recommendation: rewrite the six in your own words; I will keep the structure.

4. The stage bands for the five new services. The NPI five came from your roadmap. The other four are my reading of how we work — especially: do we sit in supplier award negotiations (Sourcing), do we run pre-shipment gate reviews (Quality), does "negotiation support" belong to us (PCBA), and is "you own the declaration" the right framing (Certification)?
Recommendation: correct anything that misrepresents what we actually do.

5. Publishing the network. Four CMs (unnamed) and eleven sub-suppliers by city and capability. Nothing confidential is disclosed — but confirm none of it is sensitive to a supplier relationship.
Recommendation: publish as-is; it is our strongest differentiator while we have no client list.

6. Going live. The revamp currently exists as a separate build. Making it live is a one-step change once you are happy.
Recommendation: review the two pages, then promote the homepage and solutions page together — they reference each other.

07Known gaps

08How to review this in ten minutes

  1. Open index.html. Read only the first screen. Does it tell you what we do and why to keep reading?
  2. Scroll to Who you're talking to. Is the "ask them yourself" line something we can stand behind?
  3. Click two symptoms in Pick what sounds familiar. Are the answers useful, or generic?
  4. Open solutions.html and read the NPI band. Does the you/we split describe our work accurately?
  5. Come back with the six decisions above.
Eastbound Link · v3 review package · internal draft · October 2026