You sell campaigns, content and strategy to large companies. Every one of your clients has a website. You don’t. In meetings, the portfolio PDF does the job, and does it well. The problem comes later, when someone writes “can you send me a link?” and there’s nothing to send.<\/p>\n\n\n\n
That’s how a creative agency with corporate clients came to us. It had looked at a few domains and none of them existed yet. No hosting, no CMS, not a single line written for the web. What it did have was its credentials PDF and a folder of campaign photos and videos. That was enough.<\/p>\n\n\n\n
The first decision wasn’t about design or technology. It was about content. Everything on the site would come from the portfolio PDF: the copy, the figures, the client logos, the project photos. Nothing invented. If the PDF didn’t say it, the site didn’t either.<\/p>\n\n\n\n
It sounds like a straitjacket. Far from it. Think of a design agency’s natural objection when it hands a site to another team: “we can do the design ourselves”. Of course they can. That’s why the copy, the brand and the choice of projects stayed in the client’s hands. What was left to sell was structure, motion and engineering, and nobody argued about that.<\/p>\n\n\n\n
Within a week, the client had three navigable versions of the same site:<\/p>\n\n\n\n
The client discarded the first intro as too basic, which was a good sign: they cared about craft. The third version was the one that got published.<\/p>\n\n\n\n
How do three directions come out of seven days? Through a technical decision that sounds boring: static pages, no framework and no build process. Every motion tweak was done by hand and showed up instantly in the browser.<\/p>\n\n\n\n
On the visual reference: the client admired the site of a European agency. We took its scrolling mechanics and never its palette. A color count of the original site corrected one of our versions, which had pushed a secondary tone onto whole surfaces.<\/p>\n\n\n\n
Every effect shipped with a fallback. On touch screens, the pinned carousel becomes normal scrolling with arrows. If the visitor has “reduce motion” turned on in their system, the animations switch off and the text reads the same. And without JavaScript, the content is still there.<\/p>\n\n\n\n
The question came with the proposal and came back later: what about WordPress? The answer was “not yet”. With the design still moving, a CMS would have made every change slower. There was also a simpler reason: we didn’t know what the client would want to edit. Any panel built at that point would have been a guess.<\/p>\n\n\n\n
Instead, the static site was published on the client’s domain and left to live for a while.<\/p>\n\n\n\n
The first maintenance requests arrived the way they always do: four WhatsApp messages of one line each. We turned them into four tasks, each with its cause and its evidence. One example: the menu didn’t navigate to the sections because the menu layer stopped the smooth scrolling and, when it closed, canceled the jump. It was fixed by closing the menu before scrolling, and tested on a phone and on a computer.<\/p>\n\n\n\n
With those requests, and the ones that followed, it became clear what really changed: copy, photos, projects, contact details. That was the moment to move to a CMS.<\/p>\n\n\n\n
We set a strict condition for the move: the new site had to look identical to the static one. The WordPress theme wraps the original markup as is, without reinterpreting it, and a custom plugin handles the logic. The acceptance test was a pixel comparison between the two sites, on desktop and on mobile. It had to come out with no differences, and it did.<\/p>\n\n\n\n
What about the block editor? Or a page builder? We looked at both and ruled them out: either one would have fought the scrolling choreography that had been tuned by hand.<\/p>\n\n\n\n