By Pablo Revale

Three Ways Hotel Marketing Tech Dies After the Kickoff

The app, the franchisee content tooling, and the CMS. Same failure wearing three different costumes.

July 31, 2026 · 7 Minutes

THE BUDGET WAS APPROVED. NOTHING CHANGED.

The pattern repeats across almost every marketing technology purchase in a hotel portfolio. The business case is approved, the vendor ships, the team is trained, and twelve months later the metric it was bought to move has not moved. HotelTechReport’s 2026 survey shows where hoteliers believe the value sits: 44% call CRM and guest marketing integrations critical to daily operations, 40% rank business intelligence dashboards, 39% point to guest messaging. Every one of those is an integration, not a feature. What you are buying is almost never the interface. It is whether the interface can read and write the state living in your PMS, your CRS, and your loyalty database. Three products make this concrete, because they are the three we get asked for most.

THE APP THAT GETS OPENED ONCE

An app whose only unique capability is mobile check in has no reason for a second session. The guest already has your confirmation email, and the email does not need 40 megabytes of storage. For an app to survive the first stay it has to do at least one thing the website cannot, and every candidate for that thing is an integration: the live folio so a guest sees charges before checkout, availability and rates so they can extend from the room, loyalty balance written back in real time, a room key. Push notifications are the clearest version of the failure. A push that cannot read reservation state is marketing spam, and it teaches the guest to disable notifications in week one. A push that knows the guest is in house, on day two of four, with unspent resort credit, is the product. The integration is not a phase two. It is the reason the app exists.

FORTY FRANCHISEES, FORTY VERSIONS OF YOUR BRAND

Distributed brands leak through the assets nobody controls. A franchisee needs a flyer on Thursday, the brand portal offers a 60 page PDF guideline and no editable file, so the flyer gets built in Canva with the 2019 logo. Policing that with a compliance email does not work, and it is not a discipline problem. It is a missing pipeline. What fixes it is a templating layer: locked design tokens for type, color, and logo placement, editable content slots, and export presets per channel, so the only thing a franchisee can change is the thing they should be changing. Travel advisors are the same problem with more revenue attached, because their job is not producing brochures, it is quoting. If quoting three room options across two properties takes eleven screens, they will sell whatever is easiest to quote, and that is a competitor. What a distributed seller needs is not a brochure. It is a quote they can build in front of a client without apologising.

THE CMS THAT STILL NEEDS A DEVELOPER

If changing a rate banner takes a ticket and a release, you do not have a CMS problem, you have a content model problem. It usually traces to one decision: the site was modeled as pages instead of components, so every editable region is coupled to a template only a developer can touch. The fix is structural. Model content as typed components with defined slots, expose them in a headless CMS with role based permissions, and decouple the front end so a content change is a publish and not a deploy. The multi property version is worse and more common. Twelve properties, twelve sites, one person editing, which makes every seasonal change twelve manual edits with at least one property permanently wrong. Hiring a second editor does not solve that. Making shared content inheritable, with property level overrides only where the difference is real, does.

NOBODY SCOPES THE MIGRATION

Here is the line item missing from most of these projects. The new platform is the easy half. Moving what already exists is the half that fails. Five hundred pages do not move themselves, and a migration that ignores URL parity hands back a site that is faster, better looking, and ranking for nothing, because redirects were treated as a launch week checklist instead of a design constraint. The same applies to anything carrying a session. If your booking path changes domain or structure, your analytics stop being comparable on the day you go live, which means the first thing your new site does is destroy your ability to prove it worked.

THE TRAINING IS THE DELIVERABLE

The second missing line item is the handover. If your marketing team cannot ship a promotional page without calling the agency, the platform did not buy you speed. It moved your bottleneck onto someone else’s calendar and put an hourly rate on it. Any implementation worth paying for ends with your team able to operate the thing, and with the decisions written down somewhere that outlives the people who made them.

THE PLATFORM ARRIVES EMPTY

The third missing item is whatever is supposed to go inside it. A CMS without content model discipline becomes a second website nobody maintains. An app with no offer calendar shows the same three tiles for a year. A template library with no photography for two thirds of the portfolio gets abandoned by month two, because a franchisee who cannot find an image of their own property will use one they took on a phone. Platform work creates content obligations, and those obligations land on a marketing team whose headcount was justified by the platform reducing work. Budget the production, or the tool gets measured as a failure when it was never given anything to display.

WHAT THIS LOOKED LIKE IN PRACTICE

Three of these ran with Unique Vacations. On the website, we migrated more than 500 pages to Strapi, coded around 100 reusable components, rebuilt the front end in Next.js on a property serving over a million monthly users, and trained their marketing team to ship content on their own. On mobile, two separate apps became one React Native codebase in under four months, live across 19 properties in 9 countries, now carrying about 16,000 daily interactions. On the advisor side, the Sandals travel advisor portal already handled roughly 60% of Unique Vacations’ sales, which makes it the least forgiving system in the company to touch. We rebuilt the codebase from scratch, ran the new portal in parallel with the old one, and swapped URLs at launch so reservations were never interrupted, moving a platform doing 600 million dollars without a maintenance window. We also replaced static navigation with a personalized dashboard and simplified quoting across multiple room options, which is the advisor workflow problem described above. For PRIOR, in bespoke luxury travel, the same discipline applied to a booking flow produced end to end booking in under three minutes.

IF ONE OF THESE THREE IS ON YOUR ROADMAP

Tell us which one, and what your stack looks like today. We will come back with what we would build, the integration points we expect to be the hard part, and what has to be true for your team to run it after we leave.