Skip to content

Conversion-focused landing page — self-initiated prototype · 2026

Alucover

Built in my own time as a prototype. It is not Alucover’s official website, was not commissioned, and is unrelated to my installation job there — that role involves no software work.

I install aluminium pool enclosures full-time. That means I know the product, the questions customers actually ask on site, and the reasons people hesitate — so I built the landing page I thought the product deserved. It is a single self-contained file: one page, three languages, every image embedded, no external requests, no dependencies.

3
Languages
158
Translation keys
1
Self-contained file
Alucover landing page hero showing an aluminium pool enclosure over a swimming pool, with a Polish headline reading 'Twój basen otwarty dziewięć miesięcy w roku'.
The hero leads with season length — the argument that decides the sale.

Objective

What the problem was

  • Turn what I hear from customers during installations into the structure of a page — lead with the season-length argument, because that is the objection that actually decides the sale.
  • Support Polish, English and German, since the product category sells across the border.
  • Ship something that loads fast on a phone in a garden with poor signal, which is where a prospect is most likely to look at it.

My role

Everything: the commercial argument and copy in three languages, the page design, the illustrations, the front-end build and the deployment.

What I built

  • A single-page site covering the seasonal argument, benefits, the product range, reasons to buy, the process, a gallery, FAQs and a quote request.
  • A trilingual Polish/English/German toggle driven by a translation table covering 158 keys per language.
  • Custom inline SVG line-drawings of each enclosure profile, with a human figure for scale, instead of photographs — they stay sharp, weigh nothing and read clearly at small sizes.
  • Every photograph embedded directly in the page, so it makes no external image requests at all.
  • Mobile behaviour handled explicitly: anchor scrolling that accounts for the fixed header, scroll locking behind the open menu, and safe-area insets for notched displays.

Screens

Product range section headed 'Trzy wysokości. Jeden standard wykonania.' with three cards, each showing a custom line drawing of an enclosure profile.
Product range. Each profile is a custom inline SVG rather than a photograph.
Benefits section with a grid of cards covering longer season, lower running costs, safety and property value, each with a small icon.
The benefits section, ordered by what customers actually ask about.
Alucover landing page on a narrow mobile viewport, with the hero headline and quote button stacked over the pool enclosure photograph.
At 390px — the viewport this page was really designed for.

Tools

Front end

  • HTML
  • CSS
  • JavaScript
  • Inline SVG
  • Embedded images

Approach

  • Single self-contained file
  • No dependencies
  • Three-language content table

Deployment

  • Netlify

Challenges

The decisions worth explaining

Three languages without a framework

Every translatable element carries a key, and one table holds all three languages. It keeps the copy in a single place rather than scattered through the markup, so adding German after the fact was filling a column rather than editing the page again.

Deciding what the page argues

The hardest part was not the code. The page leads with how much longer the swimming season gets, because that is the point at which customers on site stop treating the enclosure as a cover and start treating it as an extension of the house. Working with the product is what told me that; I would not have guessed it from a brief.

Mobile is where it will actually be read

A prospect looks at this on a phone, standing next to their pool. Anchor links that scrolled underneath the fixed header, a page that scrolled behind the open menu, and content lost under a notch were all real defects, and fixing them properly took a dedicated pass.

What I learned

  • Knowing the product physically changes the copy. The objections I answer on the page are ones I have answered in person.
  • Embedding assets and depending on nothing external is a legitimate engineering choice for a page this size, and it makes the result trivially portable.
  • A translation table is worth building before the second language, not after — retrofitting the first one is the expensive part.

Source code

The repository is private. Happy to share it on request.

Next project

Stravo Media

The independent venture the rest of this work sits under — positioning, website, sales process and the internal system that runs it.

Read on