Where they started
A single resort on a busy island, ranking for its own name and a handful of “[island] hotel” terms, and nothing else. Everything upstream of the booking decision belonged to OTAs and a couple of ad-heavy travel blogs. The site had exactly one page about the island itself: a 200-word “Explore the area” stub with a stock beach photo and no internal links out. Meanwhile the destination pulled tens of thousands of monthly searches for hikes, beaches, ferry times and day trips. The resort was invisible for all of it, despite sitting on the island.
What we did
- Built a destination guide hub: a pillar page plus 30-odd guides (“best beaches”, “things to do in the rainy season”, “getting around without a car”) targeting research-stage queries, each one linking through to the relevant room or package page.
- Wired every guide to a booking page so a reader planning a hike lands two clicks from the room nearest the trailhead. The guide hub is now the second-largest internal traffic source on the site.
- Earned links from travel and tourism sources with genuinely useful guides (a seasonal “when to visit” breakdown drew the bulk of the 24 links), lifting DR from 39 to 43.
- Structured the hub for depth, not breadth: tight internal linking between related guides let the cluster rank as a topic, which is why the curve compounds rather than spiking and fading.
This is a different bet from optimising the booking flow. The traffic is upper-funnel and OTA-immune: people research a destination long before they compare hotels, and the resort that answers those questions is the one already open in the tab when they book.