Reframing a four-hotel brand portfolio around a single visitor intent — three hero explorations and one full landing page, desktop to mobile.
Three hero explorations and one full landing page (desktop to mobile) for Winklerhotels, a four-property collection of design hotels in South Tyrol. Maintain the brand. Optimize for conversion.
Winklerhotels' real site (winklerhotels.com) is a portal: four hotels, one navigation, built for a visitor who wants to compare before choosing. That's the right structure for a homepage. It is not the right structure for a page whose entire traffic arrives from a single ad, already holding one specific intent.
Scope was deliberately narrowed to one season (summer) and one traveller profile (leisure) to keep the single-intent premise honest.
Each gap below was read off the live site as a visitor arriving from a single ad would meet it — the hero first, everything else after a scroll. All three were re-checked against winklerhotels.com on 21 July 2026, and one of them is stated more narrowly here than in my original write-up, because the narrower version is the one that holds.
Visually strong, and nothing in the hero to act on — no booking field, no date picker, no ask. There is a persistent BOOK button in the header, so the site is not without a path entirely; but inside the page the next Book control appears 2,168px down, far below the fold. A visitor arriving with intent meets a photograph, and has to go looking for the thing they came for.
Winklerhotels holds a genuinely strong 9.2 "Wonderful" rating on Booking.com (30 reviews at the time of the analysis), surfaced nowhere on the site. A trust signal that already exists, unused.
The homepage optimizes for comparing four hotels, which is correct for brand-aware, browsing traffic. It actively works against single-intent, ad-driven visitors who need one path, not four.
Each hero targets a different visitor mindset at first contact, and answers a different gap.
Selected
Booking widget visible on load. Answers Gap 01 head-on — the missing call to action is simply gone.
Mood and photography lead, a soft CTA follows. Closest to the original site's own instinct, kept honest with an ask that wasn't there before.
Rating and guest quote foregrounded before any offer. Answers Gap 02 directly — trust stated before it's asked for.
I chose variant A because the brief asks to optimize for conversion, not brand awareness — and a landing page visitor already arrives with intent. It's the only variant that removes the missing-CTA problem entirely, and its directness set the pace for everything after it.
Booking widget visible on load, no scroll required to act. This is variant A, built out unchanged — the version marked Selected in the grid above, and the image at the top of this page.
One shared statement, then all four hotels differentiated in a single line each — visitors self-select before the page asks anything harder.
One room per hotel, real specs, its own CTA per card — the decision happens in small increments, not one big ask.
Shows what an OTA listing can't — riding, golf, the mountains — the case for booking direct.
The 9.2 rating and a guest quote, placed right before the final ask — the trust signal from Gap 02, put where it does the most work.
Same booking bar, repeated at the natural end of the journey, in a full-bleed section that signals "this is the destination."
Real addresses per hotel double as a quiet trust signal, not just navigation.
Hero to footer, one intent carried the whole way down — the booking bar at the top, every section earning the one below it, and the same ask again at the end. The scrollable version is one click away.
A design brief, answered in full and published as portfolio work. Winklerhotels is a real company; this page is an independent exercise, not commissioned by them and not endorsed by them. The Booking.com figure quoted in the analysis is dated where it appears, and will drift.
Designer spec sheet