The Challenge
The Good Hotel Guide needed technical platform modernisation: its legacy WordPress infrastructure could not scale to meet growing commercial and editorial ambitions, for a decades-old trusted print-guide brand now needing to compete with searchable OTA-style platforms while retaining its earned trust and editorial identity. The engagement itself was structured atypically, scoped almost entirely through a long-standing personal relationship between the agency’s CEO and the client’s founder, with limited scope for direct discovery or user research with the client or its readers.
The Outcome
Delivered a technically scalable platform supporting thousands of pages and complex multi-contributor authoring workflows for editors, writers, hotels, and reviewers, with a genuinely sophisticated shared taxonomy across content types. Not all UX recommendations were implemented; the live site today still reflects at least one of the specific gaps originally raised, since search returns results only on submission with no predictive suggestions while typing, evidence the original diagnosis was accurate even without full authority to act on it.
The Approcah
Within that constraint, contributed the information architecture underpinning the platform: a shared taxonomy (Region, County, Area, City) spanning multiple distinct content types (hotel listings, offers, editor’s picks, and reader-contributed reviews), reflected in the site’s genuinely complex mega-menu navigation. Where direct research specific to this client wasn’t available, argued for specific UX decisions using calibrated evidence from concurrent, demographically similar research running on another live project, an imperfect but reasoned substitute for primary research, used transparently rather than presented as project-specific evidence.
Personally authored the site’s homepage trust messaging (‘Independent / Recommended / Trusted’), while raising at the time that generic trust claims, words every competitor site also uses, don’t interrupt a new visitor’s default assumptions about what kind of site this is, and argued for a short video or similar disruptive format that would force a moment of genuine re-evaluation rather than passive reassurance. Also raised the lack of predictive and contextual search results.
